IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A...

26
1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information. This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 1 M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh Ghavimi, Student Member, IEEE, and Hsiao-Hwa Chen, Fellow, IEEE Abstract—Machine to machine (M2M) communication is an emerging technology to provide ubiquitous connectivity among devices without human intervention. The cellular networks are considered as a ready-to-use infrastructure to implement M2M communications. However, M2M communications over cellular pose significant challenges to cellular networks due to different data transactions, diverse applications, and a large number of connections. To support such a large number of devices, M2M system architecture should be extremely power and spectrum efficient. In this paper, we provide a comprehensive survey on M2M communications in the context of third generation partnership project (3GPP) long term evolution (LTE) and long term evolution-advanced (LTE-A). More specifically, this paper presents architectural enhancements for providing M2M services in 3GPP LTE/LTE-A networks and reviews the features and requirements of M2M applications. In addition, the signal over- heads and various quality-of-service (QoS) requirements in M2M communications deserve our attention too. We address M2M challenges over 3GPP LTE/LTE-A and also identify the issues on diverse random access overload control to avoid congestion caused by random channel access of M2M devices. Different application scenarios are considered to illustrate futuristic M2M applications. Finally, we present possible enabling technologies and point out the directions for M2M communications research. Index Terms—M2M communication, 3GPP, LTE, LTE- Advanced, architecture, random access I. INTRODUCTION Machine-to-machine (M2M) communications refer to the ways enabling automated applications that provide connec- tivity among machines or devices without any human in- tervention. The M2M communications may involve a large number of devices in a wide range of application domains, thus forming so-called Internet of Things (IoT). Cellular systems are expected to play a significant role in the successful deployment of M2M communications. Indeed, mobile cellular communications feature several advantages, such as global standard infrastructure, cost-effective connectivity, easy instal- lation and maintenance, especially for a short-term deployment of M2M applications. Several reports appeared in the literature to predict a con- siderable market growth for both M2M devices and M2M Fayezeh Ghavimi (email: [email protected]) and Hsiao- Hwa Chen (email: [email protected]) are with the Department of Engineering Science, National Cheng Kung University, 1 Da-Hsueh Road, Tainan City, 70101, Taiwan. This work was partially supported by Taiwan Ministry of Science & Technology research grant: NSC 102-2221-E-006-008-MY3. The paper was submitted on February 14, 2014, revised on July 7, 2014, and accepted on September 4, 2014. connectivity segments. For example, over the next a few years, the number of smart-metering devices per cell in a typical urban environment is estimated to be in an order of tens of thousands [1]. The M2M applications may include a large number of smart meters, health monitoring devices, and intelligent transportation terminals that must be efficiently connected via communication links [2]. In order to take full advantages of the opportunities created by a global M2M market over cellular networks, 3GPP and the Institute of Electrical and Electronics Engineering (IEEE) standardization bodies have initiated their working groups for facilitating such applications through various releases of their standards [3]-[4]. The 3GPP LTE and LTE-A offer higher capacity and more flexible radio resource management (RRM) schemes than many other packet access data technologies. In LTE-A, stations can be configured as evolved universal terrestrial radio access (E-UTRA) NodeBs (eNBs) in macrocells or picocells, home eNBs (HeNBs) in femtocells [5]-[7], and relay nodes (RNs) in relay networks to provide comprehensive wireless access in both outdoor and indoor environments. Via attaching to those stations, higher-layer connections among all M2M devices can be provided. However, LTE and LTE-A were designed basically for wideband applications only; while in M2M communications, transactions at each M2M device are usually dominated by a small amount of data, leading to an inefficient utilization of LTE and LTE-A technologies. Therefore, to support a large number of M2M devices, important issues such as energy efficiency and short latency have to be addressed in M2M communications. Notably, the efforts have been made by 3GPP to overcome the shortcomings of LTE/LTE-A with its provision to support M2M communications [8], where its initial studies on M2M communications were focused on the functional architecture, service requirements, and applications [3], [8]-[9]. With regard to service requirements, M2M applications are very much different from human-to-human (H2H) communi- cations (e.g., typical applications of mobile phones), since M2M services have their unique characteristics [3], [10]. Furthermore, QoS requirements of different types of M2M services may vary widely, and these service requirements then require special architectural designs. With an architecture in place, numerous challenges remain when implementing RRM for M2M communications in LTE-A cellular networks. For example, time and frequency resources are to be shared between H2H users and M2M devices, thus inevitably causing co-channel interference among them [11]. Such co-channel

Transcript of IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A...

Page 1: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 1

M2M Communications in 3GPP LTE/LTE-ANetworks: Architectures, Service Requirements,

Challenges, and ApplicationsFayezeh Ghavimi, Student Member, IEEE, and Hsiao-Hwa Chen, Fellow, IEEE

Abstract—Machine to machine (M2M) communication is anemerging technology to provide ubiquitous connectivity amongdevices without human intervention. The cellular networks areconsidered as a ready-to-use infrastructure to implement M2Mcommunications. However, M2M communications over cellularpose significant challenges to cellular networks due to differentdata transactions, diverse applications, and a large number ofconnections. To support such a large number of devices, M2Msystem architecture should be extremely power and spectrumefficient. In this paper, we provide a comprehensive surveyon M2M communications in the context of third generationpartnership project (3GPP) long term evolution (LTE) and longterm evolution-advanced (LTE-A). More specifically, this paperpresents architectural enhancements for providing M2M servicesin 3GPP LTE/LTE-A networks and reviews the features andrequirements of M2M applications. In addition, the signal over-heads and various quality-of-service (QoS) requirements in M2Mcommunications deserve our attention too. We address M2Mchallenges over 3GPP LTE/LTE-A and also identify the issueson diverse random access overload control to avoid congestioncaused by random channel access of M2M devices. Differentapplication scenarios are considered to illustrate futuristic M2Mapplications. Finally, we present possible enabling technologiesand point out the directions for M2M communications research.

Index Terms—M2M communication, 3GPP, LTE, LTE-Advanced, architecture, random access

I. INTRODUCTION

Machine-to-machine (M2M) communications refer to theways enabling automated applications that provide connec-tivity among machines or devices without any human in-tervention. The M2M communications may involve a largenumber of devices in a wide range of application domains,thus forming so-called Internet of Things (IoT). Cellularsystems are expected to play a significant role in the successfuldeployment of M2M communications. Indeed, mobile cellularcommunications feature several advantages, such as globalstandard infrastructure, cost-effective connectivity, easy instal-lation and maintenance, especially for a short-term deploymentof M2M applications.

Several reports appeared in the literature to predict a con-siderable market growth for both M2M devices and M2M

Fayezeh Ghavimi (email: [email protected]) and Hsiao-Hwa Chen (email: [email protected]) are with the Department ofEngineering Science, National Cheng Kung University, 1 Da-Hsueh Road,Tainan City, 70101, Taiwan.

This work was partially supported by Taiwan Ministry of Science &Technology research grant: NSC 102-2221-E-006-008-MY3.

The paper was submitted on February 14, 2014, revised on July 7, 2014,and accepted on September 4, 2014.

connectivity segments. For example, over the next a fewyears, the number of smart-metering devices per cell in atypical urban environment is estimated to be in an order oftens of thousands [1]. The M2M applications may includea large number of smart meters, health monitoring devices,and intelligent transportation terminals that must be efficientlyconnected via communication links [2]. In order to take fulladvantages of the opportunities created by a global M2Mmarket over cellular networks, 3GPP and the Institute ofElectrical and Electronics Engineering (IEEE) standardizationbodies have initiated their working groups for facilitating suchapplications through various releases of their standards [3]-[4].

The 3GPP LTE and LTE-A offer higher capacity and moreflexible radio resource management (RRM) schemes thanmany other packet access data technologies. In LTE-A, stationscan be configured as evolved universal terrestrial radio access(E-UTRA) NodeBs (eNBs) in macrocells or picocells, homeeNBs (HeNBs) in femtocells [5]-[7], and relay nodes (RNs)in relay networks to provide comprehensive wireless access inboth outdoor and indoor environments. Via attaching to thosestations, higher-layer connections among all M2M devicescan be provided. However, LTE and LTE-A were designedbasically for wideband applications only; while in M2Mcommunications, transactions at each M2M device are usuallydominated by a small amount of data, leading to an inefficientutilization of LTE and LTE-A technologies. Therefore, tosupport a large number of M2M devices, important issues suchas energy efficiency and short latency have to be addressed inM2M communications. Notably, the efforts have been madeby 3GPP to overcome the shortcomings of LTE/LTE-A withits provision to support M2M communications [8], where itsinitial studies on M2M communications were focused on thefunctional architecture, service requirements, and applications[3], [8]-[9].

With regard to service requirements, M2M applications arevery much different from human-to-human (H2H) communi-cations (e.g., typical applications of mobile phones), sinceM2M services have their unique characteristics [3], [10].Furthermore, QoS requirements of different types of M2Mservices may vary widely, and these service requirementsthen require special architectural designs. With an architecturein place, numerous challenges remain when implementingRRM for M2M communications in LTE-A cellular networks.For example, time and frequency resources are to be sharedbetween H2H users and M2M devices, thus inevitably causingco-channel interference among them [11]. Such co-channel

Page 2: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 2

interference is responsible for degraded performance of anLTE-A cellular network supporting M2M communications. Tooptimally allocate the physical resource blocks (PRBs) to theuser equipments (UEs) and/or M2M devices, the schedularshould exploit channel and traffic dynamics on a fast timescale, ideally per transmssion time interval (TTI) [12]. There-fore, it is necessary to investigate the ways, in which howH2H users and M2M devices can efficiently share availableradio resources, to mitigate the co-channel interference andthus improve network efficiency. In this paper, we intend topresent some architectural enhancements needed to accom-plish the M2M service requirements. In addition, M2M servicerequirements and features are to be illustrated in detail.

To deploy M2M communications successfully in 3GPPLTE/LTE-A cellular networks, several major challenges needto be tackled. One of the most important issues in enablingM2M in LTE/LTE-A networks is congestion and systemoverload problem. The LTE/LTE-A networks were designedmainly to handle H2H communications, where the amountof uplink (UL) traffic is normally lower than the downlink(DL) traffic. In contrast, M2M applications may produce moretraffic data in UL channels than the data over DL channels.Congestion due to concurrent transmit messages from a largenumber of M2M devices can be overwhelming, thus impactingon the operations of a whole mobile network. In the contextof M2M communications, signaling congestion may occurdue to a malfunction in an M2M server (e.g., M2M devicesrepeatedly try to connect to the same remote server, whichis down) or an application (e.g., synchronized operation of aparticular M2M application). The congestion can also occurdue to concurrent attempts from a large number of M2Mdevices to attach/connect to the network [3]. The investigationsin 3GPP in the literature indicated that both M2M devices andUE may suffer uninterrupted collisions at a random accesschannel (RACH) when a large number of M2M devices areactive. This challenge attracted a significant attention, andvarious possible solutions have been proposed by the 3GPP[13].

Several review papers in literature [14]-[17] discuss M2Mcommunications in the context of emerging wireless technolo-gies. In [14], the authors describe the technological scenarioof M2M communications consisting of wireless infrastructureto cloud and related technologies toward practical realization.Moreover, [15] presents a survey on home M2M networks andexamines the typical architectures of home M2M networksalong with discussing the performance tradeoffs in existingdesigns. Furthermore, [16] presents a survey of existing M2Mservice platforms and explores the various research issues andchallenges involved in enabling an M2M service platform.In addition, the authors in [17] describe machine type com-munications in 3GPP networks and provide a summary ofthe solutions agreed within 3GPP for congestion control andnetwork overload avoidance.

Hence to the best of our knowledge, a comprehensivesurvey on M2M communications with its focus on LTE/LTE-Asystems is not available in the literature. Therefore, the mainpurpose of this paper is to provide a review on the studiesappeared in the literature, helping the readers to understand-

ing what has been investigated (architecture, technologies,requirements, challenges, and proposed solutions) and whatstill remains to be addressed. In addition, this paper willreveal an evolutionary path of the M2M communications forfuturistic research.

The reminder of this paper is outlined as follows. In SectionII, we discuss architectural enhancement of LTE/LTE-A withregard to the M2M communications. The M2M communi-cation standardization activities, service requirements, andfeatures are the subject of Section III. In Section IV, theM2M challenges over 3GPP LTE/LTE-A are studied whilethe principal applications of the M2M communications willbe addressed in Section V. Section VI lists the open researchissues on M2M communications via discussing relevant topicssuch as traffic characterization, routing, heterogeneity, security,etc., followed by the conclusions given in Section VII.

II. M2M NETWORK ARCHITECTURE

Different from normal mobile network terminals, M2Mdevices carry many unique characteristic features from theperspective of mobile operators. Therefore, it is necessary toseek optimized networking solutions in particular for M2Mapplications over mobile networks. To provide global integra-tion among diverse solutions in the M2M applications, it is im-portant to design a standard end-to-end M2M communicationnetwork architecture. This section provides an overview onM2M network architecture and identifies related M2M R&Dactivities reported in the literature.

A. M2M Access Methods

M2M devices can be either stationary (e.g., power metersin homes, machines in factory, etc.) or mobile (e.g., fleetmanagement devices in trucks). The access network connectsM2M devices to the infrastructure using either wired (i.e.,cable, xDSL, and optical fiber) or wireless links. Wireless ac-cess methods can be either capillary/short range (i.e., WLAN,ZigBee, and IEEE 802.15.4x, etc.) or cellular (i.e., GSM,GPRS, EDGE, 3G, LTE-A, WiMAX, etc.). Although the wiredsolutions can provide high reliability, high rate, short delay,and high security, it may not be appropriate for the M2Mcommunication applications due to its cost ineffectiveness,and lack of scalability/mobility support. Alternatively, wire-less capillary solutions, mainly used for shared short rangelinks/networks, are rather cheap to roll out, and generally scal-able. However, small coverage, low rate, weak security, severeinterference, and lack of universal infrastructure/coverage poserestriction on its applications to M2M communications. Onthe other hand, wireless cellular offers excellent coverage,mobility/roaming support, good security, and ready-to-useinfrastructure, making M2M over cellular a promising solutionfor M2M communications. Therefore, in this article, our focusis on the M2M communications based on 3GPP LTE/LTE-Amobile networks.

B. 3GPP Network Architecture

In this part, we describe the 3GPP network architectureto provide a comprehensive survey and more specifically to

Page 3: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 3

reveal an evolutionary path from the non-LTE to the LTE/LTE-A technologies helping the readers to understanding what isthe similarity for the element nodes in non-LTE and LTE/LTE-A as well as a description of the functionality of them. Anoverview of the evolved packet core (EPC), the legacy packetand circuit switched elements, 3GPP RANs, and the mostsignificant interfaces are illustrated in Fig. 1. Furthermore,Fig. 1 shows the most important EPC nodes in LTE/LTE-A networks and also, the corresponding UMTS terrestrialradio access network (UTRAN) nodes, namely serving GPRSsupport node (SGSN), gateway GPRS support node (GGSN),media gateway (MGW), and mobile switching center (MSC)in the non-LTE network.

UTRAN

GERAN

E-UTRAN

MGW MSC

SGSN GGSN

GSMA

GPRS-GB

Lu-PS

Lu-CS

S1-MME

S1-U

S3

PS

Core

CS

Core

MME

S-GW

P-GW

EPC

S4

S11

S5

Mc

Gn

Fig. 1. An overview of EPC for 3GPP accesses.

The SGSN and mobility management entity (MME) receivedevice trigger from MTC-IWF; encapsulates device triggerinformation in non-access stratum (NAS) message sent tothe UE/M2M device; receives device trigger delivery suc-cess/failure status to MTC-IWF. Furthermore, SGSN performssecurity functions, access control and location tracking. Itplays the role of MME and serving gateway (S-GW) in theEPC.

The GGSN or packet data network gateway (P-GW) maysupport the following functionality. Based on access pointname (APN) configuration and unavailability of MSISDNand external identifier(s) in the GGSN/P-GW either queriesan MTC accounting, authorization, and authentication (AAA)server for retrieval of external identifier(s) based on IMSI orroutes RADIUS/Diameter requests for AAA servers in externalpacket data network (PDN). The GGSN function is similar tothe P-GW in the EPC.

The MSC server controls circuit-mode services. The MSC ismostly associated with communications switching functions,such as call set-up, release, and routing. It also performs a hostof other duties, including routing SMS messages, conferencecalls, fax, and service billing as well as interfacing with

other networks, such as the public switched telephone network(PSTN).

The MGW was introduced to bridge among different trans-mission technologies and to add service to end-user connec-tions. The MGW uses open interfaces to connect to differenttypes of node in the core network and external networks.

The requirements and major elements of the EPC architec-ture were characterized in 3GPP Release 8, which will playan important role in the implementation of the next generationM2M networks [20]. Along with the 3GPP LTE that appliesmore to the radio access technology, there is also an evolutionof the core network known as system architecture evolution(SAE). These two major parts lead to the characterization ofthe EPC, evolved UTRAN (E-UTRAN), and E-UTRA, eachof which corresponds to the core network (CN), RAN, and airinterface of the whole system, respectively [5].

Some frequently used acronyms in this paper are listed inTable I. In the following, we provide an overview of the E-UTRAN architecture, the main EPC node functionalities, andfunctionalities defined for LTE-A systems, respectively.

TABLE ILIST OF FREQUENTLY USED ACRONYMS.

AAA Accounting, authorization, and authenticationAPI Application program interfaceAPN Access point nameAS Application serverCN Core networkeNB Evolved node base stationEPC Evolved packet core

E-UTRAN Evolved universal terrestrial radio access networkGGSN Gateway GPRS support nodeHeNB Home evolved node base stationHLR Home location register

HPLMN Home public land mobile networkHSS Home subscriber serverIMSI International mobile subscriber identity

IP Internet protocolMAC Medium access controlMGW Media gatewayMME Mobility management entityMSC Mobile switching center

MSISDN Mobile station integrated service digital networkNAS Non-access stratum

PDCP Packet data convergence protocolPDU Packet data unit

P-GW Packet data network gatewayPHY Physical layer

PRACH Physical random access channelRACH Random access channelRAN Radio access networkRLC Radio link controlRRC Radio resource controlRN Relay nodeSCS Service capability server

SGSN Serving GPRS support nodeS-GW Serving gatewaySME Short message entity

SMS-SC Short message service-service center

1) LTE-A E-UTRAN Overview: The architecture of the E-UTRAN for LTE-A is shown in Fig. 2. As mentioned earlier,an LTE-A network comprises two parts, i.e., the EPC andthe RAN, where the former is known as CN, and the latterconsists of base stations (BSs) that are referred to as evolved

Page 4: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 4

node base stations (eNBs) [5]. The EPC is responsible foroverall control of mobile devices and establishment of InternetProtocol (IP) packet flows. The eNB is responsible for wirelesscommunications and radio access, and provides an air interfacewith user plane and control plane protocol terminations towardthe UE and M2M devices. Each of the eNBs serves one orseveral E-UTRAN cells, and the interface interconnecting theeNBs is called the X2 interface. Besides, the eNB is connectedto the EPC through the S1 interface. In addition, HeNBsthat are the eNBs for indoor coverage improvement can beconnected to the EPC directly or via a gateway that caters foradditional support for a large number of HeNBs. Furthermore,the 3GPP LTE-A encompasses relay nodes and sophisticatedrelaying strategies for network performance augmentation. Theaim of this new technology is to offer large coverage, high datarate, and better QoS performance and fairness for differentusers.

eNB

RN

RN

HeNB

HeNB GW

MME/S-GW/P-GW

eNB

EPC

X2

S1 S1E-UTRAN

Fig. 2. LTE-A E-UTRAN architecture.

As mentioned earlier, the eNBs provides the E-UTRANwith the user and control plane termination protocols. Fig. 3gives a graphical overview of both protocol stacks. In the userplane, the protocols include packet data convergence protocol(PDCP), radio link control (RLC), medium access control(MAC), and physical layer (PHY) protocols. The control planestack additionally includes the radio resource control (RRC)protocols.

The main functionalities carried out in each layer aresummarized as follows [5], [21]-[24].

• NAS: The NAS is the highest stratum of the controlplane between UE/M2M and core network at the radiointerface. This layer is used to support the continuousconnection of UE/M2M as it moves, and also to managethe establishment of communication sessions to maintainIP connectivity between the UE/M2M and an P-GW.Furthermore, the NAS is a protocol for messages passedbetween the UE/M2M and core network. The NAS mes-sages include update or attach messages, authenticationmessages, service requests, and so forth. In addition,

PHY

MAC

RLC

PDCP

RRC

NAS

S-GWMME

eNB

User

Plane

Control

Plane

Non-Access

Stratum

(NAS)

Access

Stratum

(AS)

Fig. 3. User and control plane protocol stacks.

the NAS control protocol performs bearer context acti-vation/deactivation, registration, and location registrationmanagement.

• RRC: The RRC protocol layer handles the control planesignaling between the UE/M2M and eNB. The main ser-vices and functions of the RRC sublayer include broad-cast of system information related to the NAS and AS.Furthermore, establishment, modification, and release ofRRC connections are performed in this protocol layer.Initial security activation (i.e., initial configuration of ASintegrity protection and AS ciphering), RRC connectionmobility including intra-frequency and inter-frequencyhandovers, and specification of RRC context informationare the other important tasks of RRC sublayer. Moreover,this sublayer performs QoS control functions, UE/M2Mdevice measurement configuration and reporting. In addi-tion, the RRC transfers dedicated NAS information andnon-3GPP dedicated information.

• PDCP: This layer performs IP header compression anddecompression using ROHC protocol (the current ver-sion is FFS) at the transmitting and receiving entities,respectively. Furthermore, the PDCP transfers user planeor RRC data, and this function is used for conveying dataamong users of PDCP services. Maintenance of PDCPsequence numbers for radio bearers and in-sequencedelivery of upper layer packet data units (PDUs) at HOare other functions of PDCP layer. In addition, duplicatedetection of lower layer session data units (SDUs), ci-phering and deciphering of user plane data and controlplane data, and integrity protection of control plane dataare performed in this layer.

• RLC: The RLC protocol layer exists in UE/M2M andeNB. It is part of LTE/LTE-A air interface controland user planes. This layer transfers upper layer PDU

Page 5: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 5

and performs error correction through automatic repeatrequest (ARQ). Moreover, the RLC protocol layer isused for concatenation, segmentation, and reassembly ofRLC SDUs. In addition, re-segmentation and reorderingof RLC data PDUs, RLC re-establishment, and errordetection and recovery are the other functions of thisprotocol layer.

• MAC: The MAC protocol is responsible for regulatingaccess to the shared medium. Furthermore, the choice ofMAC protocol has a direct bearing on the reliability andefficiency of network transmissions. Responsibilities ofMAC layer include multiplexing/demultiplexing of RLCPDUs, scheduling information reporting, error correctionthrough hybrid ARQ (HARQ), logical channel prioritiza-tion, and transporting format selection.

2) Evolved Packet Core Overview: The EPC is a flat all IP-based core network that can be accessed through 3GPP radioaccess (e.g., WCDMA, HSPA, and LTE/LTE-A) and non-3GPP radio access (e.g., WiMAX and WLAN), to efficientlyaccess to various services such as the ones provided in IPmultimedia subsystem (IMS). The access flexibility to the EPCis attractive for operators since it enables them to modernizetheir core data networks to support a wide variety of accesstypes using a common core network. The following textdescribes the main components of the EPC along with theirfunctionalities.

• Mobility Management Entity (MME): The MME is akey control plane element for the LTE/LTE-A accessnetwork. It is responsible for managing security functions(authentication, authorization, and NAS signaling), roam-ing, handover, and handling idle mode user equipment. Itis also involved in choosing the S-GW and packet datanetwork gateway (P-GW) for an UE/M2M device at aninitial attach. The S1-MME interface connects the EPCwith the eNBs.

• Serving Gateway (S-GW): The S-GW resides in the userplane, where it routes and forwards packets to and fromthe eNBs and packet data network gateway (P-GW).It is also a mobility anchor point for both local inter-eNB handover and inter-3GPP mobility. The S-GW isconnected to the eNB through S1-U interface and tothe P-GW through S5 interface. Each UE/M2M deviceis associated to a unique S-GW, which will be hostingseveral functions.

• Packet Data Network Gateway (P-GW): The P-GW pro-vides connectivity from the UE/M2M device to an PDNby assigning an IP address from the PDN to the UE/M2Mdevice. Moreover, P-GW provides security connectionbetween UEs/M2M devices by using Internet protocolsecurity (IPSec) tunnels between UEs/M2M devices con-nected from an untrusted non-3GPP access network withthe EPC.

As mentioned earlier, this system is considered as ”flat”since from a user-plane point of view there are only theeNBs and the gateways. This leads to a reduced complexitycompared to previous architectures.

C. M2M Communications over 3GPP LTE/LTE-A Networks

The 3GPP system provides services for M2M communi-cations1, including various architectural enhancements (e.g.,control plane device triggering), transport, and subscribermanagement. Different deployment paradigms foreseen forM2M communications between the M2M applications and the3GPP LTE/LTE-A networks are discussed in the text followed[25].

The most straightforward deployment paradigm is the directmodel, where the application server (AS) connects directly toan operator network in order to communicate with the M2Mdevices without using the services of any external servicecapability server (SCS), as shown in the left-most stack ofFig. 4 (or Fig. 4A).

M2M

device

M2M

device

M2M

device

Core

network

+RAN

SCS SCSM2M

network

domain

M2M

application

domain

A. Direct model B. Indirect model C. Hybrid model

-------- Control plane

User plane

Core

network

+RAN

Core

network

+RAN

Fig. 4. Deployment scenarios for M2M communications over 3GPPLTE/LTE-A operator network.

The second deployment paradigm is the indirect model,in which the AS connects indirectly to an operator networkthrough the services of an SCS in order to utilize additionalvalue added services for M2M (e.g., control plane devicetriggering). The SCS can be either

1) M2M service provider controlled, that is deployed out-side the operator domain. The SCS is an entity that mayinclude value added services for M2M communications,performing user plane and/or control plane communica-tion with the M2M device; or

2) the 3GPP LTE/LTE-A network operator controlled andconsidered as an internal network function. In thiscase, security and privacy protection for communicationsbetween the 3GPP LTE/LTE-A network and the SCS isoptional for being trusted.

Yet another deployment paradigm is the hybrid model,where the AS uses the direct model and indirect modelsimultaneously in order to directly connect to an operatornetwork to perform direct user plane communications with

1M2M communication is also known as machine-type communications(MTC) in 3GPP.

Page 6: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 6

the M2M devices while also using an SCS. From the 3GPPLTE/LTE-A network perspective, the direct user plane commu-nications from AS and any value added control plane relatedcommunications from the SCS are independent and have nocorrelation to each other even though they may be serving thesame M2M applications hosted by the AS.

MTC server

MTC

Application User

MME/S-GW/P-GW

S1S1

S1 S1S1

S1S1

S1

eNB

eNB

eNB

eNB

UE UE

UE

MTC

device

MTC

deviceMTC

device

MTC

device

MTC

device

MTC

device

UE

MTC-GW

MTC-GW

MTC-GW

MTC-GW

HeNB

HeNB

HeNB

HeNBRN

RN

RN

RN

UE

Operator domain A

Operator domain B

Fig. 5. Communication scenarios with MTC devices communicating with theMTC server.

As shown in Fig. 5, two communication scenarios can beenvisioned. One scenario considers communications betweenthe MTC devices and one or more MTC servers in the M2Mapplication domain. In this scenario, an M2M user (e.g., apower plant in the smart grid, or indoor health monitoring athome, etc.) can manage a massive number of M2M devicesthrough M2M server(s). The M2M servers are catered by anoperator, who offers an application program interface (API)for M2M users to access the M2M servers. The M2M serversand the 3GPP LTE/LTE-A infrastructure can be under the sameoperator domain (i.e., the operator domains A and B in Fig.5 can be the same).

To provide communications between M2M devices andM2M server(s), the public land mobile network (PLMN)enables transactions between an M2M device and an M2Mserver. Furthermore, the PLMN should provide authenticationand authorization for an M2M device before the M2M devicecan communicate with the M2M server [3].

An alternative scenario is depicted in Fig. 6, in which thereis a peer-to-peer model, and M2M devices are communicatingdirectly among themselves without M2M server(s). Communi-cations among M2M devices can be provided within the sameoperator domain or among different ones. Inter-M2M devicecommunications can be either via the mobile network or inad-hoc mode.

D. Service Capability Server (SCS)As mentioned earlier, the SCS connects to the 3GPP

LTE/LTE-A network via MTC-IWF in HPLMN to commu-nicate with M2M devices used for M2M communications.

The SCS provides an API to allow different ASs to use thecapabilities of SCS. An SCS may be controlled by the operatorof HPLMN or by an MTC service provider.

The SCS uses subscription database to authorize connec-tions on Tsp reference point2, and to locate the SCS servingnode so that control and data could be routed towards theSCS [26]. The SCS subscription identifier may be permanentsubscriber data and can be used for the following purposes:authentication and charging on the Tsp reference point, charg-ing for the SMS messages that may be sent towards the SCS,or charging for the data that may be sent to SMS-SC.

The format of the SCS subscription ID can be an interna-tional mobile subscriber identity (IMSI). Temporary subscrip-tion identifiers may be established for security purposes in amanner similar to establishing a temporary IMSI (T-IMSI) fora 3GPP LTE/LTE-A UE.

The MTC devices use SCS public identifier to send SMSmessages and/or IP packets towards the SCS. The SCS publicidentifier may be permanent subscriber data and can be usedfor the following purposes: identification on the Tsp referencepoint, charging on the Tsp reference point, charging for SMSmessages that may be sent towards the SCS (e.g., instead ofthe SCS subscription ID), or charging for data that may besent to the SMS-SC.

The SCS public identifier can be used as a field in triggermessage interactions on the Tsp reference point. The SCSpublic identifier may be an MSISDN. In this case, a specialrange of the MSISDN is allocated for the SCS so that corenetwork node can identify when traffic is destined for an M2Mdevice or an SCS. The format of the SCS public identifiermay be the format of a fully qualified domain name (FQDN),a mobile station integrated services directory (MSISD), anIP address, or an alpha-numeric format. The SCS can havemultiple public identifiers.

An SCS connects to the SCS serving node for control planecommunications (e.g., including short message exchange). Theother core network nodes use this information to determine thenext hop destination of the control messages in order to reach aparticular SCS. The SCS serving node may be a core networknode. Furthermore, the SCS serving node can be an MTC-IWF, an MSG, an MME, an SGSN, or an S-GW. An SCSserving node identifier may be temporary subscriber data. Inaddition, the serving node can be the primary node used forrouting control information towards an SCS. Also, it can bean IP address or an ISDN address.

The SCS trigger quota can be permanent subscriber data,indicating the number of triggers that an SCS is allowed torequest per time period. Furthermore, the SCS trigger quotadefines the number of successful triggers that an SCS initiatesper unit of time.

E. 3GPP LTE/LTE-A Architecture Reference Model for M2MFig. 7 depicts a typical architecture for M2M devices used

for M2M connecting to the 3GPP LTE/LTE-A radio accessnetworks.

2Tsp is a 3GPP standardized interface to facilitate value-added servicesmotivated by M2M communications (e.g., control plane device triggering)and provided by an SCS.

Page 7: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 7

MME/S-GW/P-GW

S1S1

S1 S1S1

S1S1

S1

eNB

eNB

eNB

eNB

UE UE

UE

MTC

device

MTC

deviceMTC

deviceMTC

device

MTC

device

MTC

device

UE

MTC-GW

MTC-GW

MTC-GW

MTC-GW

HeNB

HeNB

HeNB

HeNBRN

RN

RN

RN

UE

MME/S-GW/P-GW

S1S1

S1 S1S1

S1S1

S1

eNB

eNB

eNB

eNB

UE UE

UE

MTC

device

MTC

deviceMTC

deviceMTC

device

MTC

device

MTC

device

UE

MTC-GW

MTC-GW

MTC-GW

MTC-GW

HeNB

HeNB

HeNB

HeNBRN

RN

RN

RN

UE

Fig. 6. Communication scenarios of MTC devices communicating with each other without intermediate MTC server.

To support indirect and hybrid models of M2M communi-cations, one or more instances of an MTC-IWF reside in thehome public land mobile network (HPLMN). The MTC-IWFis a functional entity that hides the internal PLMN networktopology and relays or translates signaling protocols used overTsp to invoke specific functionality in the PLMN. An MTC-IWF may be a standalone entity or a functional entity ofanother network element [27].

The SCS connects to the 3GPP LTE/LTE-A network viathe MTC-IWF in the HPLMN to communicate with M2Mdevices used for M2M communications. The SCS offerscapabilities for use by one or multiple M2M applications. AnM2M device can host one or multiple M2M applications. Thecorresponding M2M applications in the external networks arehosted on one or multiple ASs. The interface between SCSand AS is not standardized by 3GPP, but other standardsdevelopment organizations (SDOs), such as the ETSI TCM2M, are expected to standardize the API. It is important tonotice that the development of M2M API should be drawn upfor all devices of the identified application areas. Furthermore,a uniform protocol view compatible with the current IP suite,will provide protocols at different levels and will be the basisof device interoperability. The development of interfaces willallow all devices, generally developed with a precise servicein mind, to embrace a greater variety of applications andto enable proactive communications of devices which aretransparent to the users.

Tsms is a non-standardized interface that encompasses var-ious proprietary short message service-service center (SMS-SC) to short message entity (SME) interfaces [28]. Tsms canbe used to send a trigger to an M2M device encapsulatedin a mobile terminated-SMS (MT-SMS) as an over-the-topapplication by any network entity (e.g., SCS) acting as anSME.

As further development of the M2M architecture takesplace, further reference points are added. In Fig. 7, blue col-ored arrow reference points are the new reference points addedto facilitate M2M communications over 3GPP LTE/LTE-A

systems.The T4 interface is used by the MTC-IWF to route a device

trigger as an MT-SMS to the SMS-SC in the HPLMN. T5a/b/cinterfaces provide optimized paths for device trigger deliveryand small data service to the M2M devices. The MTC-IWF uses S6m interface to interrogate the home subscriberserver (HSS)/home location register (HLR) for mapping anexternal identifier or a mobile station integrated services digitalnetwork (MSISDN) to the international mobile subscriberidentity (IMSI), authorizing a device trigger to a particularM2M device, and retrieving serving node information. TheMTC AAA uses an S6n interface to interrogate HSS/HLR formapping IMSI to external identifier(s) and vice versa.

III. SERVICE REQUIREMENTS AND FEATURES OF M2MCOMMUNICATIONS OVER 3GPP LTE/LTE-A

All kinds of applications can be involved in M2M commu-nications and it becomes massive in terms of diversity acrossthe applications. However, not all M2M applications havethe same characteristics [3]. This implies that every systemoptimization may not be suitable for every M2M applicationwith regard to the variety of requirements. In order to copewith this heterogeneity of requirements, the 3GPP has defineda number of features [3] (i.e., particular characteristic featuresassociated with certain applications), for which the networkneeds to be optimized. In this section, first some informationrelated to the standardization activities are provided. Then,we give information pertaining to the service requirements.Finally, the categories of features for M2M communicationsare specified in the last part of this section.

A. Standardization Activities for M2M Communications

Recently, 3GPP, European Telecommunications StandardsInstitute (ETSI), Open Mobile Alliance (OMA), China Com-munications Standards Association (CCSA), and the Alliancefor Telecommunications Industry Solution (ATIS) have startedstandardization processes on the M2M communications. The

Page 8: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 8

MSC

MME

CDF/

CGF

IP-SM-GW

MTC

application

UE

RAN

HPLMN

VPLMN

S5

T5b

T5c

Tsms

S6n

S6mRf/Ga

GdSGd

T4

E

Gi/SGi

Gi/SGi

Tsp

S4

S3

1

2----------- Control plane

User plane

Service

Capability

Server

(SCS)GGSN/P-

GW

MTC-IWF

SMS-SC/

GMSC/

IWMSC

HSSMTC

AAA

SGSN

S-GW

1

2

2

Indirect Model

Direct Model

Hybrid Model 1

T5a

Fig. 7. 3GPP LTE/LTE-A architecture reference model for M2M communications.

activities of 3GPP is concentrated on the M2M communica-tions that can be supported by mobile cellular networks. Incontrast, ETSI addresses the issues on M2M service archi-tecture, its components, and the interactions between threedomains, i.e., M2M device domain, communication networkdomain, and M2M application domain.

• 3GPP standardization group: In order to take potentialadvantages of M2M communications over cellular net-works, 3GPP system architecture working group 2 (SA2)aims to use 3GPP network and system progress thatsupport M2M in evolved packet system (EPS) [8]. In [18],the first study on M2M was initiated without specifyingsystem characteristics. The 3GPP SA2 defined 3GPP net-work system improvements in Release 10 to enable M2Mcommunications in UMTS and LTE-A core networks.The objective is to optimize the system design that canmitigate M2M signaling congestion and network overloadproblems. For Release 10 and beyond, the focus is mainlyon studying the impacts of standardized system networkimprovements on the architecture. The purpose of thesestudies is to provide essential network enablers for M2Mservices and to distinguish 3GPP network enhancementsrequired to support a large number of M2M devices inthe 3GPP network domain.

• ETSI standardization group: The ETSI technical commit-tee (TC) M2M standardization intends to provide an end-to-end overview of M2M standardization, which concen-trates on the service middleware layer that is independentof the underlying access network and transmission tech-nologies. The goal of the ETSI TC M2M is to support awide range of M2M applications and needed functions(e.g., functional architecture and interface standardiza-tion) to be shared by different M2M applications.

• OneM2M: The aim of oneM2M is to meet the criticalneeds for designing a common M2M service layer, whichcan be easily embedded within different hardware andsoftware to connect a large number of devices with M2Mapplication servers. Furthermore, oneM2M will developglobally agreed-upon M2M end-to-end specifications andarchitecture principles across multiple M2M applications[19].

B. M2M Service Requirements

In this part we identify the service requirements for M2Mapplications.

1) General Service Requirements: Here, general require-ments for the M2M systems are provided. However, there isno need for all particular M2M systems or components of these

Page 9: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 9

systems to implement every requirement. The followings arethe M2M general service requirements [3]:

• Enable the network operator to identify which individ-ual M2M features are subscribed by a particular M2Msubscriber.

• Provide a mechanism to activate or deactivate M2Mfeatures for the M2M subscribers.

• Identify which individual M2M features are activated fora particular M2M subscriber by the network operator.

• Provide a mechanism for the network operator to controlthe addition or removal of individual M2M features andalso restrict activation of M2M features.

• Provide a mechanism to reduce peaks in data and sig-naling traffic when a large number of M2M devicessimultaneously attempt data transmissions.

• Provide a mechanism to restrict downlink data trafficand also limit access towards a specific APN when thenetwork is overloaded.

• An M2M device may support the extended access barring(EAB) mechanism.

• An M2M device supporting the EAB mechanism shouldbe able to be configured for EAB by the HPLMN.

• The HPLMN should be able to configure EAB on anM2M device that supports it.

• Provide mechanisms to efficiently maintain connectivityfor a large number of M2M devices.

• The system should provide mechanisms to lower powerconsumption of M2M devices.

2) M2M Device Triggering: Device triggering is one of thekey requirements for a 3GPP LTE/LTE-A network. Motivatedby the network, triggered devices should perform certainapplication-related tasks. For devices that do not have IPaddresses (e.g., 2/3G devices), it is obvious that these devicescannot be attached in the packet switch (PS) domain inorder to be reached by the network. Since the majority ofM2M applications are data applications, it is necessary for anapplication server to reach the device in the PS domain. Thisrequires a device to be allocated an IP address. Therefore,device triggering is related to the devices that are not reachableby the AS or the SCS.

To address this requirement, control plane device triggeringis defined as the mechanism [27] to trigger a device to performspecific applications. To this end, the AS first determines theMTC-IWF that serves the M2M device. Then, AS queriesthe MTC-IWF for the IP address assigned to the M2Mdevice by sending a trigger request message. The MTC-IWFinitiates procedures for triggering the M2M device. Then,MTC-IWF passes the device trigger request to the PSDN,which communicates with the RAN. The device trigger requestmessage contains information that allows the network to routethe message to an appropriate device and also allows thedevice to route the message to an appropriate application[25]. The information destined to the application, along withthe information to route it, is referred to as the triggerpayload. An M2M device needs to be able to distinguish amobile terminated (MT) message carrying device triggeringinformation from any other type of messages.

Device triggering is subscription-based. The informationprovided by the subscription determines whether an M2Mdevice is allowed to be triggered by a specific SCS. Tspprovides connectivity for the MTC-IWF to connect to one ormore SCSs and receive the device trigger request from theSCS.

3) M2M Identifier: A large number of M2M services arecurrently deployed over circuit-switched (CS) GSM architec-ture and therefore use E.164 MSISDNs, although such servicesdo not require dialable numbers. On the other hand, thereis a concern over the numbering requirements and shortageof E.164 MSISDNs for new M2M services. Therefore, 3GPParchitecture has been enhanced to allow delivering communi-cation services using an alternate identifier, which is called anexternal identifier. More information about identifiers relevantfor the 3GPP network are specified in [29].

M2M identifiers can be categorized into:1) Internal identifiers, which is the identity that the entities

within the 3GPP system use for addressing an M2Mdevice.

2) External identifiers, which is the identity used fromoutside the 3GPP system, by which an M2M device isknown to the M2M server.

The IMSI is used as an internal identifier within the 3GPPsystems. For the external identifier, a subscription used forM2M communications has one IMSI and may have one orseveral external identifier(s) that are stored in the HSS. Theexternal identifier is globally unique and has two components:the domain identifier used to identify where services providedby the network can be accessed (e.g., MTC-IWF providedservices); and the local identifier that is used to derive or obtainthe IMSI. The local identifier should be unique within theapplication domain.

4) Addressing: In M2M communications, each terminal isconsidered as a mobile subscriber and must have a uniqueIMSI. The current structure of the IMSI allows a networkoperator to theoretically support up to 1 billion subscribers,assuming 9 digits Mobile Subscriber Identification Number(MSIN). This number, however, includes both H2H UEs andM2M terminals. The structure of the IMSI is shown in Fig. 8[30].

MCC

3 digits

MNC

2 or 3 digits

MSIN

9 or 10 digits

MCC: Mobile Country Code

MNC: Mobile Network Code

MSIN: Mobile Subscriber Identification Number

Fig. 8. The structure of the IMSI.

In addition, each mobile station in a cellular network musthave at least one assigned MSISDN. The structure of theMSISDN is depicted in Fig. 9 [30]. The current structure ofthe MSISDN, assuming a 9 digit subscriber number (SN), cantheoretically support up to 1 billion subscribers. This number,again, includes both H2H UEs and M2M terminals.

Page 10: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 10

CC NDC SN

CC: Country Code NDC: National Destination Code SN: Subscriber Number

Fig. 9. The structure of the MSISDN.

On the other hand, the growth in M2M communications isprojected to reach over 50 billion devices connected to theInternet by 2020 [31]. Therefore, the development of M2Mapplications will have an impact on national numbering planssince devices need to be uniquely addressed in order to com-municate with them, or rather to enable them to communicatewith each other. Thus, the 3GPP studied the problems in [32]and concluded that IMSI is the limiting factor of addressingand may not be suitable for M2M applications, which mayneed to make use of IP addresses. Therefore, 3GPP networksshould contain mechanisms to connect with IP-based devices.

The IPv4 protocol identifies each node through a 4-byteaddress. Due to the large number of devices in M2M com-munications, it is well known that the number of availableIPv4 addresses is decreasing rapidly and will soon reach zero.Therefore, other addressing policies should be utilized. Tosolve this problem, IPv6 addressing [33] has been proposed.The IPv6 address is a 128-bit identifier which should beenough to identify any device in M2M communications in3GPP LTE/LTE-A networks. In fact, its nearly infinite addressspace enables a future with ever more ubiquitous computation.Issues such as connectivity, interoperability, and compatibilitywith M2M communication networks must be provided. In thiscontext, IETF IPv6 provides a set of protocols over low-powerwireless personal area networks (6LoWPAN) [34] that can beutilized to integrate resource-limited devices into IPv6 net-works by simplifying IPv6 including compressing addresses,removing options considered rarely used, simplifying packetprocessing, etc.

C. Features of M2M Communications

To facilitate system optimization, the 3GPP defines 14features [3] in M2M communications. These features are listedas follows.

1) Low Mobility: The low mobility feature is suitable forM2M devices that do not move, move infrequently, or moveonly within a certain area [3], [35]. This feature enablesthe network operator to be able to simplify and reduce thefrequency of mobility management procedures [35].

2) Time Controlled: This feature is suitable for those M2Mapplications that can tolerate to transmit and receive data dur-ing defined time intervals and can therefore avoid unnecessarysignaling outside these time intervals. The network operatormay allow such applications to send/receive data and signalingoutside these defined time intervals but charge differently forsuch traffic. To make use of time controlled M2M feature,the network operator should reject access requests per M2Mdevice during a defined forbidden time interval. Furthermore,

the local network should be able to alter the access grant timeinterval based on local criteria (e.g., daily traffic load, timezones, etc.). The forbidden time interval should not be altered.It is assumed that an access grant time interval will not overlapwith a forbidden time interval.

3) Time Tolerant: The time tolerant feature is suitable forM2M devices that can delay their data transfer. The purposeof this functionality is to allow the network operator to preventM2M devices that are time tolerant from accessing the network(e.g., in case of radio access network overload).

4) Packet Switched (PS) Only: The another M2M featureis packet switched only, which is intended to provide PS-onlysubscriptions with or without assigning an MSISDN [3]. Re-mote M2M device triggering will be supported with or withoutassigning an MSISDN. Remote M2M device configurationwill still be supported for subscription without an MSISDN.

5) Mobile Originated Only: This feature is suitable for usetogether with M2M devices that only utilize mobile originatedcommunications. This is intended for applications where itis possible to reduce the frequency of mobility managementprocedures per M2M device; the network should be able toprovide a mechanism for the network operator to dynamicallyconfigure the M2M devices to perform mobility managementprocedures only at the time of mobile originated communica-tions.

6) Small Data Transmission: This feature is suitable foruse with M2M devices that transmit small amounts of dataand can thus ensure minimal network impact (e.g., signalingoverhead, network resources).

7) Infrequent Mobile Terminated: This M2M feature, i.e.,infrequent mobile terminated, is suitable for M2M devices thatmainly utilize mobile originated communications and thus thenetwork operator is able to reduce the frequency of mobilitycontrol information per M2M device.

8) M2M Monitoring: This is suitable for monitoring thestate of the M2M devices and possible events that occur inthe network. This is a feature vital to all M2M applications toguarantee that the deployed devices are operational.

9) Priority Alarm Message (PAM): The priority alarmmessage M2M feature is suitable for use with M2M devicesthat issue a priority alarm in the event of theft, vandalism, orother needs for immediate attention. In addition, this featureis used in applications, which require attention but are nottoo critical. An example is the detection of a leak, whichrequires some valves or switches to be closed. In addition, theM2M devices may issue a priority alarm even when it cannotuse normal services for some reasons (e.g., access time notallowed, roaming not allowed).

10) Secure Connection: The secure connection M2M fea-ture is suitable for M2M devices that require a secure con-nection between the M2M devices and M2M server(s). Thisfeature applies even when some of the devices are roaming.

11) Location Specific Trigger: This feature is intendedfor applications where the M2M devices are known to bein a particular area and thus the M2M device triggering isperformed by using the location information.

12) Network-provided Destination for Uplink: This featureis suitable for use with M2M applications that require all data

Page 11: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 11

from an M2M device to be directed to a network provideddestination IP address. For uplink M2M communications, thenetwork should use a destination IP address.

13) Infrequent Transmission: This feature is intended forM2M devices with long periods between two subsequent datatransmissions. The network should provide resource only whena transmission occurs.

14) Group-based Policing and Addressing: This feature issuitable for applications with an M2M group where devicesshould be optimized to handle in groups for tasks. The networkoperator may use group-based policing feature to perform acombined QoS policing. Group-based addressing M2M featureis suitable for applications with an M2M group, in which thenetwork operator should optimize the message volume whenM2M devices need to receive the same message.

IV. CHALLENGES OF M2M COMMUNICATIONS OVER3GPP LTE/LTE-A NETWORKS

In M2M communications, the necessity for supporting alarge number of M2M devices is a challenging issue. Toprovide ubiquitous wireless connections for M2M devices,3GPP LTE-A introduces a heterogeneous network (HetNet) asa special network architecture for this purpose [5], [36]-[37].The HetNet comprises four parts: conventional macrocellsformed by eNBs of E-UTRA, picocells formed by smalltransmission power eNBs deployed underlay macrocells toshare traffic loads of macrocells, femtocells formed by HeNBsto enhance signal strength in indoor environment, and RNsdeployed in coverage edges of macrocells (see Fig. 5).

Higher layer connections among all above stations can beprovided by 3GPP LTE/LTE-A infrastructure. On the otherhand, in the HetNet, interference arises between macrocellsand small cells, leading to the degradation of network enhance-ment. However, by applying recent solutions [38] for picocells,[38]-[40] for femtocells, and [41]-[42] for RNs, interferenceproblems can be effectively mitigated. Consequently, ubiqui-tous connections among all M2M devices can be providedby attaching to these stations. However, it does not ensurea successful implementation of M2M communications in the3GPP LTE/LTE-A and therefore some challenges are stillthere.

One major challenge lies in the air interface. In order tomeet requirements defined by International Mobile Telecom-munications Advanced (IMT-Advanced), the air interferencein LTE/LTE-A has been designed for broadband applications,while most M2M applications transmit and receive smallamounts of data, leading to an unreasonably low ratio betweenpayload and required control information due to the use ofnon-optimized transmission protocols. In addition, the otherimportant aspects, such as the need for low-energy and low-latency devices, have to be considered for M2M communi-cations. Therefore, efforts have been made by 3GPP underthe umbrella of MTC study and work items to begin thestandardization process for the air interface of M2M communi-cations [8], [43]. Besides, in order to support a large numberof M2M devices, the efforts were also made to address theissues, such as vast diversity of M2M service characteristics,

the need for enhancing energy efficiency, and coexistencewith current communication systems. Some solutions havebeen proposed by using cooperative techniques among stations[44]-[48], and a group based operation of M2M devices, tobe discussed in the text followed, has been regarded as apromising solution [3], [49]-[51] to support device-to-device(D2D) communications in the future.

A. Group-based Operations of M2M Devices

The primary goal of grouping a number of M2M devicesis to alleviate the signaling congestion on the air interfaceby reducing communication loads between an M2M deviceand 3GPP E-UTRA and EPC. Moreover, one of the mostimportant requirements in cellular M2M communications is toreduce power consumption [52]. This requirement can be metby employing group based operation, where a group headercollects requests, uplink data packets, and status informationfrom M2M devices in the group, and then forwards such trafficto a station of 3GPP LTE/LTE-A. Furthermore, downlink datapackets and control messages can be relayed by the groupheader from a station of 3GPP LTE/LTE-A to M2M devices inthe group. For M2M communications, devices can be groupedlogically based on service requirements or based on physicallocations of M2M devices.

One of the major applications of M2M communications isto gather measurement data from M2M devices. To logicallygroup M2M devices, it should be mentioned that the traffic ofsuch application typically has the characteristics of periodicalpacket arrivals, small data transmissions, and some givenjitter constraints. Therefore, to develop practical schedulingschemes that support a large number of M2M devices withsmall data to meet the corresponding jitter constraints is achallenging issue. To tackle this challenge, M2M devices withsimilar characteristics can be merged into a group logically.Hence, the resources for these M2M devices can be scheduledon the basis of groups [51].

To support physically grouped M2M devices, leveraging onthe presence of more capable nodes to help others in reliablydelivering their data is a good approach in a heterogeneousM2M communications in 3GPP LTE/LTE-A networks. How-ever, the challenging task is how to place a number of suchnodes in an environment so as to improve the overall networkperformance. It is foreseen that M2M communications willadd some features to current cellular systems, such as WiMAX1.0 system based on IEEE 802.16-2009 [53] or WiMAX 2.0system based on IEEE 802.16m [54].

B. Device-to-Device Communications

In a cellular network, direct communications between mo-bile devices are not permitted. Traffic should be routed viaa core network even if a source and a destination are veryclose to each other. However, by allowing two physically closeusers to communicate directly, instead of being relayed by acore network, Device-to-Device (D2D) communication mayachieve lower power consumption, less transmission delay,and less load distribution of data servers for locally process-able M2M traffic. To enable direct communications among

Page 12: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 12

M2M devices, a new communication scheme (e.g., a newair interface standard with a new radio frame structure) willbe defined in 3GPP Release 12, to establish communicationsbetween end devices [55]-[57]. The D2D communicationscan improve the efficiency [58]-[62] by exploiting the highchannel quality of short range D2D links. Furthermore, byreducing transmission power, the battery life of M2M devicescan be significantly prolonged [63]-[65]. The other advantagesof D2D communications over cellular networks include moreefficient resource (e.g., spectrum) utilization because of directrouting of D2D traffic [63]-[68], and improved content deliv-ery performance by using inter-recipient transmissions [69]-[72]. Therefore, the 3GPP intends to provide the interfaces andprotocols for direct packets exchanges among M2M devicesto enable communications for M2M devices in LTE/LTE-Anetworks.

C. Cognitive M2M Communications

It is expected that a large number of M2M devices willbe deployed to support various applications. Even thoughthe signaling congestion can be potentially alleviated by thegroup based M2M devices, if the number of M2M devicesgrows rapidly, this problem may not be easy to solve. In thissituation, it may be needed to deploy more E-UTRA stationsto dilute traffic of UEs and M2M devices. To tackle signalingcongestion, E-UTRA stations for UEs and E-UTRA stationsfor M2M devices may need to work together. However,under this circumstance, interference between conventionalH2H communications and M2M communications turns outto be a challenging issue. A centralized coordination mayhelp to reduce the interference. However, due to the factthat centralized coordination creates significant signaling over-head and management burden, this scheme was not widelyadopted. Therefore, for interference mitigation between H2Hand M2M communications, an appropriate method is to em-ploy a distributed resource management, and a promisingsolution known as cognitive M2M communications [74]-[75]is particularly useful.

To support wireless transmissions for a large number ofdevices, the M2M communications can work based on arandom access channel (RACH). The advantage of using theRACH is that the devices can compete and access an avail-able channel for wireless transmission independently withoutcoordination and centralized control. Furthermore, the RACHmechanism has low communication and signaling overheads.This mechanism is found suitable for M2M communications asthe data to be transmitted from M2M devices is usually smallin amount. However, the RACH mechanism was designed tooperate on the shared channels and the number of availableshared channels is limited, and they also have to be sharedwith H2H communications. Therefore, an appropriate methodto utilize the spectrum for M2M communications is requiredto support various wireless applications.

Cognitive radio (CR) has been introduced to improve thespectrum utilization and transmission efficiency. In cognitiveradio, unlicensed users (i.e., secondary users) are allowedto access the spectrum allocated to the licensed users (i.e.,

primary users) [73], as long as the transmissions of thelicensed users are not interfered. Therefore, cognitive radio isa promising technique for M2M communications by allowingthe devices to opportunistically access the channel, when sucha channel currently is not used by the primary users. It isexpected that cognitive radios can be an effective solution forthe practical implementation of M2M networks [74].

D. Resource Allocation with QoS ProvisioningQuality of service (QoS) provisioning is one of the most

important requirements and challenging issues in M2M com-munications. M2M communications feature no or little humanintervention, low power, high reliability, and low complexity.The lack of power supply is always a challenge that limits theperformance of wireless communication devices, for both UEsand M2M devices. In H2H communications, battery can easilybe changed in a handset. However, in M2M communications,saving energy for devices is more important than increasingthe data throughput since M2M devices may be deployed indangerous or non-reachable places. Consequently, the batteryin an M2M device should be used for a relatively longtime. Furthermore, a typical M2M communication networkmay consist of a large number of devices. To allocate radioresources efficiently while ensuring QoS requirement for re-liable communications is an essential and challenging issue[76]-[77]. In addition to energy consumption and reliability,complexity is another consideration in the design of M2Mcommunications. Sophisticated algorithms should be avoidedin M2M communication devices that should be simple andyet effective, which may not be the same as those in H2Hcommunications.

For M2M devices, some applications (e.g., traffic control,robotic networks, and e-health) need mobility support [78].Some other applications (e.g., data traffic from meters in smartgrid or navigation systems) require strict timing constraints,and catastrophes may occur if timing constraints are violated[51]. Therefore, in M2M communications, providing diverseand strict QoS guarantees is one of the most importantand challenging issues [76]. Such diverse QoS requirementsparticularly need for appropriate resource allocation that canbe applied to M2M communications in LTE and LTE-Acellular networks. In [79], joint massive access control andresource allocation schemes was proposed which performmachine node grouping, coordinator selection, and coordinatorresource allocation, and also determine the proper number ofgroups under a 2-hop transmission protocol, to minimize totalenergy consumption in both flat and frequency-selective fadingchannel.

Two major methods can be considered for radio resourceallocation between M2M and H2H communications [80].M2M and H2H communications can access the same radioresources via orthogonal channels. Although this scheme issimple, it leads to a low spectral efficiency. Another methodis to use a shared resource allocation scheme. In this way,M2M devices can reuse the radio resources allocated to H2Hcommunications to achieve a higher spectral efficiency. How-ever, this may increase the interference level in comparisonwith orthogonal channel allocation.

Page 13: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 13

The minimum resource unit for downlink and uplink trans-missions is referred to as a resource block (RB). One RBnormally consists of 12 subcarriers (180 kHz) in the frequencydomain and one subframe (1 ms) in the time domain [5].When an M2M device has packets to transmit, it performsrandom access (RA) using the physical random access channel(PRACH) during an allowable time slot, called an access granttime interval (AGTI) or RA opportunity (i.e., RA-slots). InM2M communications, generally small amounts of data needto be transmitted. Although the data size is small, when a largenumber of M2M devices try to communicate over the samechannel, the devices should contend to access the shared radiochannels, causing the network overload problem.

On the other hand, in collocated H2H and M2M communi-cations networks (as shown in Fig. 10), various types of linksexist and they are listed as follows:

• The eNB-to-UE link;• The eNB-to-M2M device link;• The eNB-to-M2M gateway link;• The M2M gateway-to-M2M device link; and• The M2M device-to-M2M device link.

1

2

5

4

3

M2M

device

UE

M2M

device

M2M

device

M2M

device

M2M

gateway

Fig. 10. Illustration of transmission links in 3GPP LTE/LTE-A networks withM2M communications.

When radio resources are shared among these links, inter-ference arises and poses a big challenge. Thus, it is neededto efficiently partition the radio resources in such networks[80]. The purpose of radio resource partition is to apply therestrictions to the radio resource management between H2Hand M2M devices. Given the characteristics of links, therestrictions can be either on the transmit power or in the formof restrictions on available radio resources. Such restrictionsimprove the signal-to-interference-plus-noise ratio (SINR), andconsequently to the cell edge performance and coverage.

E. Random Access Channel Congestion

In LTE/LTE-A systems, random access procedure [13], [81]is generally performed when an M2M device turns on anddoes not have uplink radio resources assigned to send userdata or control data (e.g., a channel measurement report)to the eNB. Furthermore, random access procedure is used

by the M2M device in order to perform handover fromone eNB to another eNB, or to acquire the uplink timingsynchronization. When the number of UE/M2M devices isan acceptable value, random access provides efficient requestdelivery. However, the number of M2M devices in a cell isexpected to be much larger than the number of UEs. Whena large number of M2M devices try to access the networksimultaneously, it leads to a low RA success rate, and thus bothM2M devices and UEs may suffer continuous collisions at thePRACH [17], [82]. This may cause packet losses, extra energyconsumption, waste of radio resources, and unexpected delays.The channel can be further overloaded when the M2M devicesrepeat their access attempts after collisions. Thus, effectiveoverload control mechanisms are required for RA-based M2Mcommunications. In [83], the feasibility of semi-persistentscheduling for voice over IP (VoIP) by random access wasinvestigated and its performance in terms of throughput ofrandom access and traffic channels, and random access delaywas evaluated. Furthermore, best effort application for randomaccess in wireless multimedia network [84] and distributedrandom access scheduling exploiting the time-varying natureof fading channels for multimedia traffic in multihop wirelessnetwork [85] have been studied. In the next part of this section,we review the existing mechanisms for controlling PRACHoverload to support M2M communications in LTE/LTE-Anetworks.

To support M2M communications in LTE/LTE-A, the fol-lowing solutions have been proposed for controlling PRACHoverload [45]-[46].

1) Backoff Scheme: The backoff scheme is used to delay therandom access (RA) attempts of H2H and M2M devicesseparately. In this scheme, the backoff time for the H2Hdevices is set to a small value (e.g., the maximum backoffduration 20 ms); while the backoff time for the M2Mdevices is set to a large one (i.e., an upper limit for theretransmission intervals, which can be as long as 960ms). This scheme is effective in low channel overload,and thus it can alleviate collisions in these situations.However, it cannot solve the congestion problem inheavy overload situations when a massive number ofM2M devices initiate RA at the same time. Seo andLeung [86] studied the uniform backoff in LTE relativeto the exponential backoff in IEEE 802.16 WiMAX.Furthermore, these authors investigated a multipacketreception (MPR) slotted ALOHA system using binaryexponential backoff (BEB) algorithm with infinite buffersin the mobile terminals [87].

2) Slotted Access Scheme: In this scheme, each M2M deviceis allowed to perform RA only in its dedicated accessslot. At other times, the M2M devices are in sleep mode.The M2M devices can calculate the allowable access slotsthrough its ID and RA-cycle. The eNB broadcasts theRA-cycle, which is an integer number multiple of a radioframe. The number of unique RA-slots is proportional tothe RA-cycle length and the number of RA-slots withina radio frame. PRACH will be overloaded when thenumber of M2M devices in a cell is greater than the total

Page 14: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 14

number of unique RA-slots. In this case, several M2Mdevices share the same RA-slot and collisions may occur.Increasing the RA-cycle can reduce collision but createsunacceptable delay for an RA request. The impact of thenumber of transmission attempts on the throughput anddelay of the slotted ALOHA based preamble contentionin the LTE-A random access system investigated in [88].

3) Access Class Barring (ACB) Scheme: The ACB-basedscheme was originally designed for the access controlof devices. In this scheme, an eNB broadcasts an accessprobability (AP) and access class (AC) barring time. Inthe ACB, there are 16 ACs. AC 0-9 represents normaldevice, AC 10 represents an emergency call, and AC11-15 represents specific high-priority services. When adevice initiates RA, the device randomly draws a valuebetween zero and one, and compares this with AP. If thenumber is less than AP, the device proceeds to the randomaccess procedure. Otherwise, the device is barred for anAC barring duration. The ACB scheme can deal withexcessive PRACH overload by setting an extremely smallvalue of AP. However, a small AP leads to unacceptabledelay for some devices.From the Release 10 and onwards, the existing ACBscheme has been extended to allow one or more new ACsfor M2M devices, and an individual access barring factorcan be assigned for each of the classes. Furthermore,3GPP also proposes an extended access barring (EAB)scheme [89], in which when EAB is activated, the devicesbelonging to certain ACs (i.e., delay-tolerant devices) arenot allowed to perform RA. However, without cooper-ations among eNBs, devices within dense area suffersevere access delays. To facilitate devices escaping fromcontinuous congestions, [48] proposed the cooperativeACB for global stabilization and access load sharing toeliminate substantial defects in the ordinary ACB, thus,significantly improving access delays.

4) Pull-based Scheme: The pull-based scheme is a central-ized control mechanism, in which the M2M server re-quests the eNB to page the intended M2M devices. Uponreceiving a paging signal from the eNB, the M2M devicewill initiate RA. In this scheme, the eNB can control thenumber of devices to be paged by taking into accountthe PRACH load and resource availability. However, thisscheme requires extra control channel resources to pagea huge number of M2M devices.

5) Dynamic PRACH Resource Allocation Scheme: In thisscheme, the eNBs can dynamically allocate PRACH re-sources based on PRACH overload condition and overalltraffic load. When a subframe is used for the PRACH, partof that subframe cannot be used for data transmission.Therefore, to meet a given QoS requirement, a certainnumber of subframes should be used for the PRACH.Although dynamic allocation of PRACH resources canbe applied in most cases, the efficiency of this scheme islimited by the availability of additional resources. A self-optimizing algorithm was proposed in [90], where theeNBs can automatically increase or decrease the numberof RA-slots based on channel traffic load.

F. Reliable Data Transmission

In M2M communications in 3GPP LTE/LTE-A, each trans-mission of an M2M device may only carry a small amount ofdata due to the small data transmission feature. Therefore, ahigh peak data rate transmission scheme may not be necessaryfor M2M devices. Instead, reliable transmission (i.e., low biterror rate, and low latency) are essential. Network coding hasbeen shown to provide an effective means for efficient reliabledata dissemination and to require little coordination amongnodes. Furthermore, random data combination is a lightweight,yet effective, mechanism to provide adequate reliability anderror control with little overhead. These paradigms have beenshown to greatly improve the performance of dissemination inhomogeneous networks, but extension of these techniques toheterogeneous scenarios like M2M communications in 3GPPLTE/LTE-A has not yet been addressed. Finally, for denselydeployed nodes with limited individual capabilities in suchnetworks it makes sense to look into distributed processingparadigms for decoding.

G. Energy Management

Energy management ranging from harvesting, conservation,to consumption is a major issue in the M2M communicationcontext in 3GPP LTE/LTE-A networks. Reducing power con-sumption is one of the major challenges in M2M communi-cations. In the literature, various energy-efficient MAC pro-tocols [15], [91]-[92] exist that can be implemented in M2Msystem to save energy. However, the development of novelsolutions that maximize energy efficiency is essential. Networkprotocols will have to deal with inherent characteristics ofM2M communications such as long sleep cycles, energy andprocessing power constraints, time-varying radio propagationenvironments, and topologies varying with node mobility.

In this regard, current technology is inadequate, and ex-isting processing power are too low to meet future require-ments. Therefore, the development of novel, more efficient,and compact energy storage sources such as fuel cells andprinted/polymer batteries are paramount. Furthermore, devel-oping new energy generation devices coupling energy trans-mission methods or energy harvesting using energy conver-sion, as well as extremely low-power circuitry and energyefficient architectures and protocols will be the key factorsfor roll out of autonomous and smart M2M communications in3GPP LTE/LTE-A networks. In order to realize the decouplingof M2M applications and services, novel energy efficientservice discovery mechanisms must be designed to minimizehuman intervention during configuration and managementphases [93].

H. Self-Management Capabilities

In order to support the expected huge scale of M2M com-munications in 3GPP LTE/LTE-A, devices will need to self-manage without external intervention [94]. Due to multi-pathfading, path loss, and shadowing phenomena in radio channels,self-management learning is essential when an M2M systemencounters with such dynamic and unstable environment.

Page 15: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 15

Therefore, when trying to apply M2M communications in3GPP LTE/LTE-A networks, we have to face the exponentialgrowth in complexity that the connection of large number ofdevices will bring, which will call for context awareness, self-organization, self-management, self-optimization, self-healingand self-protection capabilities. In a wireless context, theincreased topology dynamics due to channel fluctuations andpossible device mobility, as well as the loss of signaling andcontrol messages, make these issues all the more challengingtask.

M2M communications encompass a huge sensor network,with immense amounts of sensor data from various sensors,meters, appliances and electrical vehicles. Data mining andpredictive analytics are essential for efficient and optimizedoperation of such network. A key question is how to analyzeand process these data in an efficient and timely manner.Various machine-learning techniques can be used in this regardfor data analysis and processing.

V. M2M COMMUNICATIONS APPLICATIONS

The emergence of low-power and low-cost sensors andactuator nodes (such as radio frequency identification, or RFIDtags), which are capable of communicating wirelessly usingstandardized interfaces and protocols, and increased comput-ing capability make it possible to develop a large numberof applications for M2M communications in 3GPP LTE/LTE-A. These M2M applications significantly improve the qualityof our life at home, at work, in traveling, etc. These M2Mapplications were proposed based on the possibility for a largenumber M2M devices to communicate with each other and toconvey the information they perceive from the surroundingswhere a wide range of applications are deployed. These M2Mapplications can be classified into the following categories:

• e-Health;• Smart environment (home, office, and plant);• Intelligent transportation;• Security and public safety; and• Other futuristic applications.Some of the applications that we are talking about are rather

straightforward or close to our current living style, and someothers may be futuristic such that we can only imagine at themoment, as the technologies are not available yet and we arenot ready for their deployment (see Fig. 11). In the followingsubsections, we will discuss all of them.

A. e-Health

Various advantages of the M2M communications can beuseful to the healthcare applications, and those M2M appli-cations can include tracking and monitoring of patients anddrugs, identification and authentication of patients in hospitals,automatic medical data collection and retrieving [95].

1) Tracking and Monitoring: Tracking is a function aim-ing at the identification of a person or object in motion.This includes real-time position tracking, such as the caseof tracking a patient or tracking the motion of organ (ora segment of an organ) in a patient. In terms of physicalassets, M2M communications can also be used in continuous

inventory/stock tracking (e.g., for goods availability mainte-nance), and substance tracking to prevent left-ins during asurgery. In the case of monitoring [96], the M2M applicationsin e-health enable remote monitoring of patient health andfitness conditions via M2M sensor nodes, alerting serviceswhen elderly people fall, and triggering alarms when criticalconditions are detected. The M2M communications can alsohelp in remote medical treatments or operations.

2) Identification and Authentication: Identification and au-thentication in healthcare are needed in a variety of forms, in-cluding patient identification to reduce the risk of wrong treat-ments to patients (in terms of drug/dosage/time/procedure),real-time based electronic medical record/data maintenance,and privacy protection against possible medical data intru-sion/leakage. Identification and authentication are most com-monly used to grant security access (e.g., to restricted areasand containers).

3) Data Collection: Automatic data collection and transferis required to reduce patient processing/treatment time and toimplement medical treatment automation (including medicaldata retrieving), medical care service and procedures auditing,and medical inventory management. Depending on the typesof M2M applications, data collection may proceed in differentways [9]. For example, important and vital information in e-health should be delivered as soon as detected, whereas theenergy consumption data at home may be collected only oncea time. The M2M system should support different ways ofdata delivering/reporting requested by the M2M applicationsas listed below [9]:

• A periodic reporting with the time period being definedby the M2M applications;

• An on-demand reporting with two possible modes, onebeing instantaneous collecting and reporting of data, theother being reporting of data that were pre-recorded at aspecific time period;

• A scheduled reporting; or• An event-driven reporting.4) Sensing: Sensor devices enable many functions related

to patients in healthcare, in particular on diagnosing patientconditions, providing real-time information on patients’ bio-logical data [95]. A body area network (BAN) of sensors istypically deployed around a patient to record his/her biologicalparameters, such as blood pressure, body temperature, heartrate, weight, etc.

In order to enable the M2M applications for e-health andto acquire the information on a patient’s health, the BAN ofM2M sensors have to be used. For this reason, the patient ormonitored person typically wears one or more M2M sensordevices that record health indicators (e.g., body pressure, heartrate, etc.). Due to the strict constraints on form factor andbattery consumption of these M2M sensors, it is expectedthat they require to forward the collected data with someshort range technology to a device that can act as an M2Maggregator of the collected information and an M2M gateway.Then, the LTE/LTE-A as an access network connects theM2M gateway to the M2M core network. Through the M2Mcore network, the M2M gateway is connected to the M2Mserver that stores and possibly reacts to the collected data and

Page 16: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 16

M2M communications

applications

e-Health Smart environmentIntelligent

transportation

Security and public

safety

Other futuristic

applications

Tracking and

monitoring

Smart homes,

offices, and shopsLogistics services

Remote

surveillance

Information-

ambient society

Identification and

authentication

Data collection

Sensing

Smart lighting

Smart industrial

plants

Smart water

supply

M2M assisted

driving

Fleet management

e-Ticketing and

passenger services

Smart parking

Personal tracking

Public

infrastructure

protection

Robotic

applications

Environment

monitoring

Green

environment

Smart car

counting

Journey time

estimation

Fig. 11. M2M applications and relevant scenarios.

subsequently the M2M application user (i.e., healthcare remotemonitoring center). In this scenario, the gateway could be afixed device such as a PC or a mobile device like a cell phoneor a standalone device carried on a keychain or worn aroundthe patient’s wrist or neck.

B. Smart Environment

With recent advances in wireless communications, intel-ligent systems, sensor networks, the quality of human lifehas been improved significantly in every aspect. A futuristicsmart city based on M2M communication technologies, firstproposed by IBM as one of its most important strategies,may generate an enormous amount of information, and it iscapable of collecting, managing, and taking advantage of theinformation to implement automations in our daily life. Mak-ing better decisions based on real-time information leads tosignificantly reduced living costs, and more efficient utilizationof natural resources. To this end, the M2M communicationscan be used every where around us, at homes, in offices, inindustrial plants, and every corner of cities, to realize a smartenvironment.

1) Smart Homes, Offices, and Shops: We are living in anenvironment surrounded by various electronic appliances, suchas lights, air conditioners, heaters, refrigerators, microwave

ovens, and cookers. Sensors and actuators of M2M com-munications are installed in these devices to make a moreefficient utilization of energy and also to make our life morecomfortable. Heating and cooling in homes can be adjustedto the weather conditions to maintain a desirable temperature.The lighting in rooms can be adapted to the time of a dayand to the number of occupants inside the rooms. Domesticincidents, like fire, a fall of elderly people, or burglary can bedetected with appropriate M2M monitoring and alarm systemsassociated with the M2M devices in place. Energy savingcan be improved by automatically switching off the electricaldevices when not in use. Consequently, the power consumptioncosts can be reduced by using electric appliances only whenthe energy price is cheaper, a function that can be implementedwith the help of smart grid technology [104].

Smart cities, as a new concept in urban planning, haveattracted a lot of attention recently. Connected by M2Mcommunication technologies, the people living in the urbanareas will be able to enjoy the life style of the smart citiesin the future [105]. A lot of new business models will becreated due to the existence of the smart cities in the yearsto come. Imagine a scenario that people are going shoppingin a smart city, where advertisements can be delivered to acustomer based on his/her particular taste or hobby, tellinghim/her about a store around the corner that is selling the items

Page 17: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 17

the customer is just looking for with a significant discountavailable.

Once people enter the stores, M2M communication infras-tructure can provide unique and innovative communicationchannels and everything is connected, including beverage cool-ers and freezers, and people can sense and thus be aware of thedrinks they want in the beverage coolers and freezers with theirlocation information displayed in the M2M mobile terminalsattached to the glasses or watches. The M2M technologycan also optimize inventory, provide automatic updates onmaintenance needs, or even handle payment services. Thisallows retailers to cut down the cost, while ensuring thecustomers’ satisfaction.

2) Smart Lighting: Another M2M application is to imple-ment smart lighting systems for the homes, offices and streets.Smart lighting can also attribute to a significant improvementon energy saving in the cities around the world. Due to therapid growth of the urban population, at present about half ofthe world population live in cities. This trend will continueto escalate with an estimation that by 2050 about 70% ofpeople will live in cities and the number of mega-cities withtheir populations more than 10 million will increase. Thisundoubtedly poses a new challenge to the city management,intelligent building, and environment protection, especially inenergy management efficiency. Therefore, a highly efficientillumination systems in streets of a city is extremely importantto reduce carbon emission, which has be put into the agendafor many big cities in the world. Apart from energy savingthrough improving lighting systems, smart lighting couldcontribute to another 40% of electrical energy saving throughthe implementation of advanced lighting management systemsbased on M2M technology [106]. More specifically speaking,the power consumption reduction policy in Europe has beenextended from cities to the street lighting systems in order toreach the desirable energy efficiency for Europe in 2020, andthe low-carbon European economy by 2050 [107]. Therefore,M2M technological innovations, such as remote street lightcontrol that allows the M2M user applications such as thecity lighting control managers to monitor and control streetlights by smart phones, turning them on or off automaticallydepending on local illumination levels and traffic intensity willsoon be widespread.

3) Smart Industrial Plants: Industrial M2M communica-tions will enhance intelligence in control systems to improvethe automation in industrial plants by exchanging and gath-ering information among sensors, actuators, and RFID tagsin M2M communications associated with the products. TheseM2M devices can monitor vibration in an industrial machinery,and a warning can be signaled or even the whole productionprocess can be made stop if it exceeds a specific threshold.Once such an emergency event is triggered, the M2M devicesimmediately connect to the M2M controller server and trans-mit this event-related information to the M2M server throughthe 3GPP LTE/LTE-A core network [97]. If the robots detectan emergency shutdown event, M2M controller server willstop their working immediately. The M2M application user(e.g., plant manager) can also see the status of the EnterpriseResource Planning (ERP) orders, the production progress, the

M2M device status, as well as a global view on all factories.It can also predict the consequence of device malfunctions inthe production lines based on the information stored in theM2M controller server.

4) Smart Water Supply: Nowadays, demand for watercontinues to grow rapidly, and worldwide water usage isincreasing at a rate twice as fast as the population growth.Many corporations rely on water for their critical functionsfrom management to manufacture. However, most people donot know how much water is wasted. As a matter of fact,a large percentage of world’s water disappears from agingand leaky piping systems, costing a huge amount of moneyevery year. To combat this problem, smart cities must be ableto monitor water supply closely to ensure that there is anadequate water supply to residence and business. Smart citiesequipped with M2M sensors can accurately monitor waterpiping systems and discover water leakage at the very firstmoment. These M2M sensors measure pipe flow data regularlyand propagate alerts and transmit an emergency message to theM2M controller server via the 3GPP LTE/LTE-A core networkif water usage is beyond an estimated normal range. This M2Mcommunication capability allows a smart city to determine thelocations of leaking pipes to prevent a waste of precious waterresource.

5) Green Environment: Managing electric devices to max-imize energy efficiency is one of the most important issuesto establish the green cities. Recently, smart grid has beenreceived a considerable attention as an intelligent solution tomanage electric power consumption. The main concept ofsmart grid is to employ intelligent communication networksto meet the pressing demands for efficiency improvementon power generation, distribution, and consumption sectorswith the help of M2M communications. In addition, smartgrid works based on an environmentally friendly infrastructurein order to keep its energy wastage as low as possiblefor minimizing CO2 emission. To realize this goal, it is ofgreat importance to equip smart grid with the abilities toautonomously collect data from various sectors of the gridsystems, analyze data on energy usage in real time, and self-configure its operating parameters to achieve the goals.

For the aforementioned M2M applications can be con-sidered the following structure. The M2M devices (e.g., ahome appliance, smart lighting system, water piping system,etc.) are connected to a smart meter and their informationis measured and collected by the smart meter. Due to theresource-constrained sensors associated in the M2M devices,wireless communications technologies based on Zigbee canbe established among M2M devices and a smart meter. Tocollect data packets from smart meters to the M2M gatewayshort-range communication technologies (e.g., WiFi) could beutilized. The received packets are stored in the buffer of theM2M gateway. Different types of data packets with differentQoS requirements can be stored in different buffers. The 3GPPLTE/LTE-A transceiver of the M2M gateway receives a head-of-queue packet from the buffer and transmits it to a 3GPPLTE/LTE-A eNB. The 3GPP LTE/LTE-A eNB is in charge ofbandwidth allocation for the data transmission of each M2Mgateway. After the data packets sent from M2M gateway are

Page 18: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 18

received by the eNB , they are then forwarded to the M2Mcontrol center. The M2M server is located at the M2M controlcenter for processing and storage of the received data. Thisdata is used to monitor, control, and command for the M2Mdevices.

C. Intelligent Transportation

With an increasing number of vehicles on the road, thetransportation and logistics services represent another big mar-ket for M2M communication technology. Advanced vehicles(e.g., cars, trains, trucks, buses, motor-bikes, and containerlorries) equipped with M2M sensors, actuators, and processingpower, become M2M communication entities. Furthermore,roads and transported goods use M2M sensors and tags thatcan also send valuable information to traffic M2M controlcenters and transportation companies to route the traffic,monitor the status of the transported goods, seamlessly trackthe physical locations of fleet vehicles, and deliver updatedschedule information to customers. More M2M applications inthe transportation and logistics services are discussed below.

1) Logistics Services: Goods supply chain can work ina more efficiency way as M2M communications providepossibility to track the status of goods in real-time via theM2M sensors associated with them. The M2M logistics en-ables total surveillance on the status of goods, raw materials,products, transportation, storage, sale of products, and after-sales services by keeping an eye on temperature, humidity,light, and weight, etc. If the status has some problem, theM2M devices can automatically send an alert to the M2Mserver via the 3GPP LTE/LTE-A core network. Furthermore,it is also possible to track the inventory in a warehouse sothat stockholders and enterprises can respond to the marketdynamics and to decide when to refill and when to go on-sale. Therefore, this can significantly reduce the space ofwarehouse, the waiting time of customers, and the numberof the employees to save the operational costs for businessentities [98].

2) M2M Assisted Driving: Intelligent transport systemsbased on M2M technologies along with the roads equippedwith M2M sensors and actuators can help to optimize andcontrol the traffic flows, and vehicle navigation/safety, toreduce the costs and carbon emissions. A sleepy drivercan be alerted and warned by the driving behavior M2Mmonitoring systems to avoid possible traffic accidents. M2Mcommunication systems can also automatically call for helpwhen they detect an accident, and they can alert people ifthey detect hazardous materials on vehicles. Furthermore, thegovernmental authorities can overview road traffic patternsfor traffic route planning purposes. Moreover, the informationabout the movement of vehicles transporting goods togetherwith the information about the types and status of the goodscan be used to predict the delivery time of goods and the timewhen the traffic peaks will arrive or end.

3) Fleet Management: Today, a large number of containercargo ships are traveling through international waters. Thesecontainer cargo delivery services may risk theft, physicaldamage, delivery delays, piracy, and even ship sinking. M2M

technology provides solutions that are being used in fleetmanagement to acquire a better control where cargos can berapidly delivered across different continents. The M2M appli-cations enable the tracking of vehicles and cargo containers tocollect the data on locations, fuel consumption, temperature,and humidity, in order to increase fleet safety, reduce theaccident rates, and increase the productivity of a fleet company.With more precise information, greater control, better resourcemanagement, and higher cost effectiveness, a fleet business canbe able to maintain its competitiveness with the help of M2Mtechnology.

4) e-Ticketing and Passenger Services: Ticketing systemsof traditional public transport systems are based mainly onmanual systems, and in some cases semi-automatic and/orautomatic systems are used for fare collection. In most cases,they involve tedious, time-consuming, and stressful labors dueto the need of human interventions. As a better choice, thee-ticketing model can be utilized, which comprises of a NearField Communications (NFC) enabled device as a M2M sensornode for scanning passenger’s identity at the entrance/exit ofthe stations. Once a mobile phone with NFC capability isscanned in the pay station, the code number of the stationis sent to the M2M transportation service provider throughthe 3GPP LTE/LTE-A core network. Based on the tariff tableand distance traveled, the fare is calculated and sent to themobile M2M service provider, which deducts the money fromthe passenger’s account. Furthermore, the information abouttransportation services (e.g., cost, schedule update, number ofpassengers, and available services) can be saved in an M2MNFC tag. As a matter of fact, the customers can get thisinformation by hovering their mobile phones over the M2MNFC readers. Mobile ticketing can enhance the effectivenessof ticketing, save the costs for transportation service providers,and increase convenience of passenger.

5) Smart Parking: Nowadays the car is the most ubiquitousmeans of transportation of human beings. Driving a car inurban environments has, however, significantly deterioratedliving conditions. This is mainly caused by long searchingtimes which causes frayed nerves, significant pollution, re-duced working time, financial loss and much more [99].M2M application in smart parking is a proven, robust andcost-effective way to ensure that road users know exactlywhere unoccupied car parking spaces are. Worldsensing [99]provides a cutting edge wireless smart parking technologynamed Fastprk that is based on a robust package of M2Msensors embedded into the tarmac so that it enables drivers tofind parking quickly and efficiently.

Fastprk not only can reduce the frustration experiencedwhen trying to locate a parking spot, it will also allowdrivers to save time, fuel, and associated costs. In addition,it allows the city council to monitor and manage the parkingspaces, and getting real time information. The system relieson embedded M2M sensors in each parking bay in the street.When a car parks over the M2M sensor, it is detected andM2M sensor relays that information in a wireless way tothe 3GPP LTE/LTE-A’s gateway. Then the gateway sends theinformation to the 3GPP LTE/LTE-A core network. Finally,the core network can send the information via the Internet to

Page 19: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 19

the M2M database server in real time. The occupancy is theninstantly reported to users via apps and illuminated panels inthe street.

6) Smart Car Counting: It is expected that a large numberof people will live in the cities in near future, and in the next20 years the urban population will grow from 3.5 billion to 5billion people. Therefore, it is undeniable that the private caras a means for transportation will predominate in one formor another for the next decades to come. This undoubtedlyposes new challenges in terms of city management, city trafficcontrol, and intelligent transportation. It would be necessaryto establish data collection station that provides accuratedetection of vehicles for measuring traffic flow.

In this regard, Sensefield [100] offers an end-to-end solutionfor traffic management. Wireless M2M sensors installed on thepavement detect vehicles and measure their speed and length.This information is transmitted to a nearby Data ProcessingStation (DSP) that rolls as a M2M gateway and providesdiverse connectivity and serves as a local hub. Through the3GPP LTE/LTE-A core network, DSP can be able to transmitcollected data to the M2M server. Then, the M2M control cen-ter utilizes this data to manage and monitor the infrastructureand analysis the traffic data to ease traffic flow and to smoothtransportation in the city region.

7) Journey Time Estimation: Currently available real-timetraffic information is practically non-existing and insufficientboth for road operators and for travelers. The root of theproblem can be found in the data collection systems, whichare expensive, inefficient, and inadequate. In addition to thelimitations in data collection, most traffic information andmanagement solutions are unable to provide a complete in-tegration in real time of data collection, aggregation, anddissemination.

Bitcarrier [101] offers a solution for traffic information andmanagement in any kind of road. This solution consists ofthree main elements that are as follows.

1) A network of M2M sensors auditing the Bluetooth andWiFi public frequencies of mobile devices;

2) A network of M2M servers hosting the databases;3) An online web client displaying all results regarding

speed, travel times and incidents.Bitcarrier M2M sensor is able to audit the signals emitted by

GPS navigators, hands free car kits and cell phones embarkedin vehicles, provided that the Bluetooth and WiFi sensorsof the mobile devices are active. Furthermore, this solutionensures total privacy protection for travelers. The M2M sen-sors collect anonymous data which is further encrypted beforebeing sent to the M2M server database. As original data isencrypted and destroyed, it is impossible to link a particulardevice with a user.

D. Security and Public Safety

Security is one of the most serious concerns for privateresidential, as well as commercial and public locations. Theconcerns over security and safety are attracting a lot ofattention. A great demand for effective security systems makesthe M2M communication technology a perfect choice for the

simplification and automation of security and public safetymonitoring and management. The M2M technologies pro-vide cost-effective, rapid, and flexible deployment for remotesurveillance, remote burglar alarms, personal tracking, andpublic infrastructure protection.

1) Remote Surveillance: Remote surveillance is one of thepublic safety applications. Remote surveillance systems can beapplied to monitor any open areas, valuable assets, people oreven pets for appropriate protection, where M2M sensors invideo cameras are used to transmit signals either continuouslyor at a fixed interval. The M2M applications can help todetect possible risky situations, trigger proper actions, alertauthorities, and keep an eye open to all suspicious activitiesand incidents.

More specifically, the M2M applications can let the userknow if some objects are moved to/from a restricted area (e.g.,home or office), report to an unauthorized entity, and providethe exact locations of the incidental events. In this case, theevent has to be notified immediately to the owner, authorities,and/or to the security companies. The M2M communicationscan simplify the designs of alarm management and quick re-porting systems. Connected M2M sensors via 3GPP LTE/LTE-A core network can provide a detailed road map of a thief,and thus increase the security for home residents, personnel,assets, and properties.

2) Personal Tracking: Personal tracking devices integratedwith the M2M technology, allow users to be informed wheretheir relatives/friends are on a real time basis, and they canbe warned in case of troubles/risks or when they request forany assistance. In this application [9], persons, assets, and/oranimals are equipped with portable M2M devices, each ofwhich contains a M2M communication module, together withan optional GPS unit, which sends location information auto-matically or on an on-demand basis to an M2M applicationserver via 3GPP LTE/LTE-A core network, then, the corenetwork can send the information via the internet to the M2Mdatabase server in real time, which can monitor the statuswhile also being able to track and trace the persons, objects,or animals.

3) Public Infrastructure Protection: Every government hasa wide range of infrastructure, such as roads, bridges, tunnels,buildings, cables, pipes, which should be maintained andmonitored. The applications of M2M technology enable thegovernment agencies to enhance their operational efficiencies,and to cut the costs for infrastructure maintenance. TheM2M communications can be used to efficiently monitor thecondition of public infrastructure equipped with M2M sensorsor M2M RFID tags and even simplify daily maintenanceby automating some routing tasks, including remote parkingmanagement, selective activation of street lights, or remotesurveillance of public spaces.

E. Information-Ambient Society

The M2M communications play an important role in digitiz-ing ”everything”, which means that it interprets the ”feeling”or ”intuition” in our real life using digital data [108]. Thedigitized world is creating a sheer volume of data at a rapid

Page 20: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 20

pace, which requires numerous new approaches to delivery thatamount of data effectively. The emerging Big Data technologycan significantly improve the efficiency to process those digitaldata. In fact, the Big Data has a great potential to materialize”everything” in the near future. Furthermore, both M2M andBig Data technologies will be needed, and possibly be mergedas an important platform for the construction of futuristicintelligent societies, such as a smart city or a smart community,etc. In order to enhance our society in terms of its intelligenceand innovation level, it is required to construct a comprehen-sive service platform. Based on the argument made by RobertMetcalf that the value of a network increases exponentiallyas devices connect. Therefore, more M2M devices must belinked to each other and the information from these M2Mdevices must be collected, not individually but all at once.To achieve this, service platform must be shifted from thatof a conventional vertically integrated platform system (orindividually optimized service) to a horizontally integratedplatform system [108].

As depicted in Fig. 12, the horizontally integrated M2Mservice shares the data collected from various M2M devicesand utilizes them for various services. Furthermore, the M2Mapplication users can construct an M2M system with lessinvestment cost than for those systems that are constructedindividually since instead of accumulating the required in-formation to construct a desired system, employment of ahorizontally integrated M2M platform enables the availabilityof the information for efficient use by various services.

M2M

device

M2M

device

M2M

device

M2M

device

Gateway Gateway

M2M platform

Network

M2M

application

M2M

application

Fig. 12. Conceptual diagram of the horizontally integrated M2M service.

Integrating M2M with Big Data, we create a new infor-mation based society, namely ”Information-ambient society”.The ”Information-ambient society” will evolve even furtherand eventually it will provide us with a society, in whichmachines detect various conditions by using their sensors. Itwill have its capability to understand what the human beingis thinking in an autonomic manner. This is a salient featureof the ”Information-ambient society”.

F. Robotic Applications

Robotics are able to improve the quality of our life, tosave costs, and to minimize the resource wastage. In thefuture, robots will be highly intelligent, and networked withother robots and human beings. They will be used not onlyto clean and guard our homes, but also to assist elderlyand handicapped people, perform surgery, conduct dangeroustasks, such as identifying and disabling improvised explosivedevices (IED), fire-fighting, and hazardous site inspection.These robots are machines controlled by machines (M2M)with their ability to sense, reason, and communicate in a realtime basis. They will certainly become very powerful tools inour life in the future.

In particular, the M2M based robot controlled cars ordriverless/unmanned vehicles will become more and morepopular in the near future. They could help to reduce trafficaccident risk to human-beings. Furthermore, these robot baseddriverless vehicles will be widely used in cargo transportationto reduce the traffic accidents caused by manned vehicles dueto the tiredness of the drivers. Some studies carried out by theIEEE reveal that, by 2040, driverless cars will account for upto 75 percent of cars on the roads worldwide [109].

G. Environment Monitoring

Environmental monitoring is essential to verify environ-mental stress, understand ecological patterns, and evaluatethe effectiveness of environmental protection policies andprograms. Environmental monitoring includes to monitor air,water, soil, animals, and plants. M2M communications pro-vide the solutions that can automatically take samples andconvey the monitored results to the government agencies incharge, and this has become an extremely important part ofenvironmental monitoring programs in various countries of theworld. Furthermore, the quality of fruits, vegetables, meat, anddairy products is vital to the health of their consumers. Foodsfrom production to consumers have to go through variousstages and be transported over thousands of kilometers beforereaching their consumers. During the transportation processes,the preservation status (e.g., temperature, humidity, and light)need to be monitored closely with the help of appropriateM2M sensors. The M2M sensors can precisely measure thesevariations and send the related information to the M2M servervia 3GPP LTE/LTE-A core network immediately wheneverneeded. The advancement of pervasive computing and sensortechnologies offers an effective solution for monitoring theenvironment and in-danger animals/plants [102]-[103].

VI. OPEN RESEARCH ISSUES

A. Traffic Characteristics

The characteristics of M2M communication traffic are dif-ferent from those of H2H network traffic. M2M traffic en-compasses specific traffic patterns due to its special functions(e.g., data collection and monitoring) and requirements (e.g.,strictly real-time based traffic), whereas H2H traffic follows acertain data volume, session length, and interaction frequency.Traffic characterization is an important issue for designing and

Page 21: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 21

optimization of network infrastructure. It is well known thattraffic characteristics in wireless sensor networks depend verymuch on the application scenarios [110]. It is not a problemas the issues of interest are focused on the traffic flow insidethe wireless sensor network itself. Complications arise whensensor nodes become part of the overall M2M communicationnetworks. In this case, the M2M communications will betraversed by a large amount of data generated by sensornetworks deployed for heterogeneous purposes, thus withextremely different traffic characteristics.

M2M applications may generate different traffic patternssuch as streaming, periodic, and event-driven signals [16]. Inaddition, the M2M data could have varying sizes and band-width requirements. In the case of video monitoring devices,data having a size of megabytes could be normally expected.In the case of sensor data (e.g., temperature and humidity),the amount of data per transmitted packet is usually small, andthe measured data is reported in periodical intervals. Althoughthose intervals may range from several minutes to hours [111],the aggregation of multiple M2M devices may form a notice-able dense node distribution scenario. Furthermore, allocatinga single PRB to an M2M device that transmits only smalldata could significantly degrade the spectral efficiency. Incases of emergency event-driven traffic like fire and flooding,networks may have to deal with simultaneous transmissions ofemergency data. This can severely impair the overall networkperformance and may blockage resources for other regularusers. M2M traffic characterization is also required to caterfor QoS guarantee for various M2M applications. The matterof resource allocation for LTE/LTE-A stations for the supportof QoS provisioning for M2M devices is also a challengingproblem that is still subject to further research.

The straightforward employment of the existing LTE/LTE-A protocols may not satisfy the requirements of M2M com-munications due to the large-bandwidth and low-latency linksused in LTE/LTE-A networks. Therefore, a new concept ofthe transport layer is required for M2M communicationswith regard to the use of LTE/LTE-A networks. TransmissionControl Protocol (TCP) utilized at the transport layer is knownas inadequate for M2M traffic due to the following reasons:

• Connection setup: most of the communications in M2Mdeal with the exchanges of a small amount of data, andthus the setup phase accounts for a noticeable portion ofthe session time which is unnecessary.

• Congestion control: one of the major goals of TCPis to perform end-to-end congestion control. In M2Mcommunications in 3GPP LTE/LTE-A, this may causeperformance degradation problems since the communi-cations are performed by utilizing wireless medium. Inaddition, if the amount of data to be exchanged is verysmall, TCP congestion control would be useless.

• Data buffering: TCP requires data to be stored in amemory buffer. Management of such a buffer may benot efficient regarding to the resource-constrained M2Mdevices.

• Real time applications: TCP was not originally designedfor real-time applications and it is not adequate for M2Mwireless communication networks.

Therefore, an enhanced congestion control mechanism is re-quired to improve the performance of TCP over LTE/LTE-Abefore it can be suitable for applications in M2M communi-cations.

The REST architecture is another approach which consistsof clients and servers. The REST uses the GET, PUT, POST,and DELETE protocols to access resources. However, theprotocols are not appropriate for resource-constrained de-vices in M2M communications. To meet the requirementsof resource-constraint M2M devices, Internet EngineeringTask Force (IETF) has standardized constrained applicationprotocol (CoAP). The CoAP involves very low HypertextTransfer Protocol (HTTP) overhead and supports multicastand asynchronous message exchanges over a user datagramprotocol (UDP) particularly suitable for M2M applications.However, there are still some concerns regarding to the CoAPapplications that require for further considerations. A primaryissue is the creation of an intuitive network that directlyincludes device data without the need for a cross-proxy.Another concern is the ways to support for CoAP’s security.

In the past, various solutions have been proposed for themobility management. However, their validity in the M2Mcommunications should be proven in terms of their scalabilityand adaptability before being applied to such a heterogeneousnetwork.

Another issue regards to the ways in which addresses areobtained. In the M2M communications, the Object NameService (ONS) associates a reference to a description of anidentifier to be translated into a Uniform Resource Locator(URL), identifying where the information about the objectresides. In the M2M applications, the ONS should operatein both directions, i.e., it should be able to associate thedescription of the object specified to a given identifier, andvice versa. Inverting the function is not an easy task andneeds an appropriate Object Code Mapping Service (OCMS).Desired features for OCMS were investigated in [112], wherea peer-to-peer (P2P) approach was proposed in order toenhance scalability. However, design and evaluation of OCMSin heterogeneous M2M communications are still open issuesthat require for further considerations.

B. Routing Protocols

The sensor networks [113] as a primitive form of M2Mcommunications are utilized for sensing and gathering appli-cation based on low-rate, low bandwidth, and delay tolerantdata collection processes. While current research considersfor more sophisticated applications such as scientific, mil-itary, healthcare, and environmental monitoring researches,where each M2M device performs various tasks ranging fromsensing, decision making, and action executing. Therefore,the communication framework for the sensor nodes in M2Mcommunication encounters various difficulties in satisfyingthe different technical requirements of these applications.Furthermore, the aforementioned applications have differentQoS requirements (e.g., delay, throughput, reliability, band-width, and latency) and traffic characteristics. To extract morerealistic and precise information of fast changing events in the

Page 22: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 22

real world and also to deal with them in a responsive manner,the abilities of sensor nodes in M2M devices should besignificantly enhanced. Wireless Multimedia Sensor Network-ing (WMSN), as a powerful and intelligent class of sensor-based distributed systems, is gaining more popularity since itscapability of ubiquitously retrieving multimedia informationto support a large number of both non-real time and real-timeapplications [114]. However, routing to satisfy the stringentQoS requirements of multimedia transmission in a resource-constrained M2M communication imposes new challenges.

Despite the availability of various routing protocols, theproblems still remain and other challenges are emerging withregard to the growing demands for M2M applications [114]-[116]. These challenges include mobility issues of sensors andsinks, multiple sources and sinks, dynamic hole bypassing,cross-layer awareness, multi-channel access, resource con-strained QoS guarantee, and secure routing. Available rout-ing protocols proposed for the resource-constrained networksconcentrate only on power consumption with the assumptionthat data traffic has no or loose QoS requirements. Therefore,to provide various QoS requirements for M2M applicationsand to be power efficient, routing techniques are needed to besignificantly improved or re-invented.

C. HeterogeneityOne of the main requirements for the M2M communications

in 3GPP LTE/LTE-A network to be successful is its capa-bility to integrate many types of devices, technologies, andservices. At the device domain, this involves vast variety offeatures in terms of data communication capabilities (e.g., datarates, latency, and reliability), flexibility in handling differenttechnologies, availability of energy, computational and storagepower, etc. As to services, the system should be able to supportvery diverse applications, whose characteristics and require-ments may be extremely different, in terms of bandwidth,reliability, latency, etc. These heterogeneity properties of theoverall system make the design of communication protocols avery challenging task. There are several other challenges andopen research topics to be investigated in future, which arelisted as follows.

1) Spectrum management: In wireless M2M communica-tions, spectrum scarcity is a serious issue for most applications.Heterogeneous LTE/LTE-A network is a new trend in telecom-munications, which could significantly enhance spectrum effi-ciency, power saving, and signal strength and coverage area.However, with the deployment of M2M communications basedon the LTE/LTE-A, significant spectrum sharing problems mayarise resulting in a low spectrum efficiency. Therefore, theimprovement of spectrum efficiency under a spectrum sharingenvironment for M2M communications should be carefullyconsidered.

2) Opportunity access: In this way, cognitive radio tech-nique is used to detect spectrum holes and utilize them fordynamic access. It is flexible for supporting various systemsincluding M2M applications in LTE/LTE-A. However, theopportunity access requires complex technologies for detectingthe white spectrum space and efficient radio resource manage-ment protocols without interfering to the primary users.

3) Connectivity: Another key area of investigation is howto provide communications capabilities to the various devicesinvolved in 3GPP LTE/LTE-A M2M communications. Issuessuch as communications energy consumption, antenna design,interoperability of different technologies (e.g., via cognitiveradio capabilities), adaptive techniques for a dynamic envi-ronment in the face of possibly heavily constrained resources,etc. will have to be addressed. Furthermore, it should bementioned that a system that is too connected becomes hardto manage (e.g., due to the excessive interference), thus, itwill be important to understand what needs to be connectedso as to provide the necessary communications capabilities forM2M devices.

Moreover, the 3GPP LTE/LTE-A standards provide ubiqui-tous wireless access by attaching to eNBs through single-hoplinks in H2H communications. However, distinct characteris-tics and a large number of devices in M2M communicationsmay create some other challenges if compared with H2H com-munications. Therefore, in this scenario, utilizing single-hopmay not be an appropriate solution and multi-hop connectionsmay be needed instead, and thus further investigations arerequired.

D. Security

Security is an important issue for successful applicationsof M2M communications. The M2M is vulnerable to attacksfor several reasons. First, for most of the time, M2M nodesare normally unattended, and thus it is easy to be physicallyattacked. Second, due to the limited capabilities of M2Mnodes in terms of their energy and computing resources theycannot implement very complex algorithms to support security.Furthermore, very often a fraction of M2M nodes switchinto sleeping mode, which makes the attacks undetectable bysystem supervisors. Finally, eavesdropping could be relativelyeasy since the M2M communications are performed in wire-less channels.

More specifically, the major problems regarding to theM2M communications security include authentication and dataintegrity. Authentication is a prerequisite for secure M2Mcommunications, and it requires appropriate authentication in-frastructures and servers allowing eNB to confirm the sensorydata from the M2M nodes through the exchange of messageswith other nodes. However, since the passive sensor nodescannot exchange too many messages with the authenticationservers, and thus such approaches may not be feasible in theM2M applications.

Data integrity must guarantee that illegal alteration of thesensory data can be detected. In M2M communications, thedata integrity requirement should be satisfied since illegalalteration may cause serious consequences, especially in life-critical M2M applications such as e-healthcare systems. Datacan be modified either by

1) adversaries while storing in the M2M node or,2) when it goes through the network.To protect data against the first type of attack, memory is

protected in most tag technologies and solutions have beenproposed for wireless sensor networks as well [117]. To protect

Page 23: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 23

data against the second type of attack, messages may be pro-tected according to the Keyed-Hash Message AuthenticationCode (HMAC) [118] that is based on a common secret keyshared between the tag and the destination of the message,which is used in combination with a hash function to provideauthentication. However, it should be mentioned that, thepassword length supported by most tag technologies is tooshort to support strong levels of protections. Furthermore,even if longer passwords are supported, still their managementremains a huge burden and challenging task especially whenentities belonging to the heterogeneous M2M networks.

The problem of data integrity was extensively investigatedin traditional communication systems and some preliminaryresults (e.g., [119]) are also available for sensor networks.However, new problems arise when sensory nodes are inte-grated into M2M communications, and thus security in M2Mcommunications remains to be an open issue.

In [120], the 3GPP Security Workgroup (SA3) has collectedcategories of vulnerabilities that are as following:

1) Physical attacks including the insertion of valid au-thentication tokens into a manipulated device, insertingand/or booting with fraudulent or modified software,and environmental/side-channel attacks, both before andafter in-field deployment. These possibilities requiretrusted ’validation’ of the integrity of the M2M device’ssoftware and data, including authentication tokens.

2) Compromise of credentials comprising brute force at-tacks on tokens and (weak) authentication algorithms,as well as malicious cloning of authentication tokens re-siding on the Machine Communication Identity Module(MCIM).

3) Configuration attacks such as fraudulent software up-date/configuration changes; misconfiguration by theowner, subscriber, or user; and misconfiguration or com-promise of the access control lists.

4) Protocol attacks directed against the device, which in-cluded man-in-the-middle attacks3 upon first networkaccess, denial-of-service (DoS) attacks, compromising adevice by exploiting weaknesses of active network ser-vices, and attacks on over-the-air management (OAM)and its traffic.

5) Attacks on the core network, the main threats to the mo-bile network operator (MNO), include impersonation ofdevices; traffic tunneling between impersonated devices;misconfiguration of the firewall in the modem, router,or gateway; DoS attacks against the core network; alsochanging the device’s authorized physical location in anunauthorized fashion or attacks on the network, using arogue device.

6) User data and identity privacy attacks include eavesdrop-ping device’s data sent over the E-UTRAN; masquerad-ing as another user/subscriber’s device; revealing user’snetwork ID or other confidential data to unauthorizedparties, etc.

3Man-in-the-middle attack considers the cases, in which a node is utilizedto identify something or someone and, accordingly, provide access to a certainservice or a certain area.

Some of the vulnerabilities that are more specifically gearedto the subscription aspects of the M2M device are exhaustiveand span the network, device, and user [120]. However, forspecial application, more additional consideration should beinvolved including the issues of liability identification thatrestrict user privacy to allow for identification of users whoseactions disrupt the operation of nodes or the transportationsystem.

Finally, all the solutions proposed to support security usesome cryptographic methodologies. Typical cryptographic al-gorithms spend large amount of resources in terms of energy atthe source and the destination which cannot be applied to theM2M communications in 3GPP LTE/LTE-A, given that M2Mdevices include elements (like tags and sensor nodes) that areresource-constrained in terms of energy, communications, andcomputation capabilities. Therefore, new solutions are requiredable to provide a satisfactory level of security includinglight symmetric key cryptographic and effective managementschemes regarding to the resource-constrained M2M networks.

VII. CONCLUSIONS

With a wide range of potential applications, the M2Mcommunications are emerging as an important networkingtechnology, which will become the infrastructure to implementthe IoT. To enable full automation in our daily life, it is neededto provide connections among all M2M devices. To implementthose ubiquitous connections, the existing 3GPP LTE/LTE-A networks have considered as a ready-to-use solution tofacilitate M2M communications. In this paper, we discussedthe architectural enhancements in 3GPP LTE/LTE-A networksfor M2M communications. We highlighted key architecturalchanges as well as the functionalities of 3GPP LTE/LTE-Anetwork elements to support various requirements of M2Mcommunications, such as device triggering, M2M identifier,addressing, etc. Then, the salient characteristic features ofM2M communications, and the issues for implementation ofM2M communications based on 3GPP LTE/LTE-A networkswere identified and discussed. Furthermore, we presented anoverview on the major challenges to implement M2M com-munications over the 3GPP LTE/LTE-A networks. In addition,typical M2M applications that can play a critical role in ourfuture life were illustrated. Finally, the open research issues onM2M communications were pointed out in order to stimulatemore research interests in the subjects.

REFERENCES

[1] Vodafone, ”RACH intensity of time controlled devices,” 3GPP, Tech.Rep., R2-102296, Apr. 2010.

[2] EXALTED Deliverable 2-1.(2011). Description of baseline referencesystems, scenarios, technical requirements and evaluation methodology[Online]. Available: http://www.ict-exalted.eu

[3] 3GPP TS 22.368 V11.5.0, “Service Requirements for Machine-TypeCommunications,” Sep. 2012.

[4] DRAFT Amendment to IEEE Standard for WirelessMAN-AdvancedAir Interface for Broadband Wireless Access Systems: Enhance-ments to Support Machine-to-Machine Applications, IEEE StandardP802.16p-11/0033, Oct. 2011.

[5] 3GPP TS 36.300 V11.2.0, ”Evolved Universal Terrestrial Radio Access(E-UTRA) and Evolved Universal Terrestrial Radio Access Network(EUTRAN), Overall Description,” June 2012.

Page 24: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 24

[6] IEEE Std 802.16-2009, “IEEE standard for local and metropolitan areanetworks part 16: air interface for broadband wireless access systems,”in revision of IEEE Std 802.16-2004, May 2009.

[7] V. Chandrasekhar, J. G. Andrews, and A. Gatherer, “Femtocell net-works: A survey,” IEEE Commun. Mag., vol. 46, no. 9, pp. 59-67,Sep. 2008.

[8] 3GPP, “System improvements for machine-type communications; (Re-lease 11), v.1.6.0,” TR 23.888, 2011-11.

[9] ETSI TS 102 689, v1.1.1, “Machine-to-Machine Communications(M2M): M2M Service Requirements” Aug. 2010.

[10] M. Dohler, T. Watteyne, and J. Alonso, “Machine-to-Machine: AnEmerging Communication Paradigm,” Tutorial, PIMRC 2010, 26 Sept.2010, Istanbul, Turkey; also at Globecom 2010, Miami 2010, USA.

[11] K. Doppler, M. Rinne, C. Wijting, C. B. Ribeiro, and K. Hugl,”Device-to-Device Communication as An Underlay to LTE-AdvancedNetworks,” IEEE Commun. Mag., vol. 47, no. 12, pp. 42-49, Dec.2009.

[12] A. G. Gotsis, A. S. Lioumpas, and A. Alexious, ”M2M Schedulingover LTE: Challenges and New Perspectives”, IEEE Vehicular Tech.Mag., vol. 7, no. 3, pp. 34-39, Sep. 2012.

[13] 3GPP TS 36.321 V10.5.0, “Medium Access Control (MAC) ProtocolSpecification,” Mar. 2012.

[14] K.-C. Chen and S.-Y. Lien, ”Machine-to-machine communications:Technologies and challenges”, Ad Hoc Networks, 2013.

[15] M. Chen, J. Wan, S. Gonzalez, X. Liao, and V. Leung, ”A Surveyof Recent Developments in Home M2M Networks”, IEEE CommunSurveys & Tutorials, vol. 12, no. 1, pp. 98-114, Feb. 2014.

[16] J. Kim, J. Lee, J. Kim, and J. Yun, ”M2M Service Platforms: Sur-vey, Issues, and Enabling Technologies”, IEEE Commun Surveys &Tutorials, vol. 16, no. 1, pp. 61-76, Feb. 2014.

[17] T. Taleb and A. Kunz, ”Machine Type Communications in 3GPPNetworks: Potential, Challenges, and Solutions”, IEEE Commun. Mag.50(3) (2012), pp. 178-184.

[18] 3GPP, “Facilitating Machine to Machine Communication in GSM andUMTS,” TR 22.868, Mar. 2007.

[19] K. Satoh, S. Miller, Z. Yang, A. Scrase, C. Blum, W. J. Kyu, and Y.Yukio. (2012) Leading ict standards development organizations launchonem2m. http://www.onem2m.org/press/oneM2M

[20] 3GPP, Overview of 3GPP release 8 v.0.1.1, Tech. Rep., June 2010.[21] 3GPP, TS 36.323 Packet Data Convergence Protocol (PDCP)

specification, Tech. Rep., December 2009 [online]. Available:http://www.3gpp.org/ftp/Specs/html-info/36323.htm.

[22] 3GPP, TS 36.322 Radio Link Control (RLC) protocolspecification, Tech. Rep., March 2010 [Online]. Available:http://www.3gpp.org/ftp/Specs/html-info/36322.htm.

[23] 3GPP, TS 36.321 Medium Access Control (MAC) protocolspecification, Tech. Rep., March 2010 [Online]. Available:http://www.3gpp.org/ftp/Specs/html-info/36321.htm.

[24] 3GPP, TS 36.331 Radio Resource Control (RRC), Tech. Rep.,March 2010 [Online]. Available: http://www.3gpp.org/ftp/Specs/html-info/36331.htm.

[25] ETSI TS 123 682, V11.4.0, “Architecture enhancements to facilitatecommunications with packet data networks and applications” Jun.2013.

[26] 3GPP TS 23.204 V11.1.0; 3rd Generation Partnership Project; Tech-nical Specification Group Services and System Aspects; Support ofShort Message Service (SMS) Over Generic 3GPP Internet Protocol(IP) Access; Stage 2; Release 11; Sep. 2011; 53 pages.

[27] 3GPP, “Architecture enhancements to facilitate communications withpacket data networks and applications” TS 23.682 V11.2.0, 2012.

[28] 3GPP TR 23.039, “Interface Protocols for the Connection of ShortMessage Service Centers (SMSSCs) to Short Message Entities(SMEs)”

[29] 3GPP TS 23.003 V10.5.0, ”Numbering, addressing and identification,”Mar. 2012.

[30] 3GPP TS 23.003 V10.0.0 (2010-12) Technical Specification 3rd Gener-ation Partnership Project; Technical Specification Group Core Networkand Terminals; Numbering, addressing and identification (Release 10).

[31] Ericsson white paper: “more than 50 billion connected devices”, 28423-3149 Uen, February 2011.

[32] 3GPP TR22.868 vb.0.0 (2007-03): Technical Report 3rd GenerationPartnership Project; Technical Specification Group Services and SystemAspects; Study on Facilitating Machine to Machine Communication in3GPP Systems; (Release 8).

[33] R. Hinden and S. Deering, ”IP Version 6 Addressing Architecture,”IETF, RFC 4291, February 2006. http://www.ietf.org/rfc/rfc4291.txt

[34] M. Gabriel, K.Nandakishore, H. Jonathan, and C. David. ”Transmissionof IPv6 Packets over IEEE 802.15.4 Networks.” IETF, RFC 4944,September 2007. http://www.ietf.org/rfc/rfc4944.txt

[35] E. Hossain, Z. Han, H. V. Poor, “Smart Grid Communications andNetworking”, Jun. 2012.

[36] J. Damnjanovic, Y. Montojo, T. Wei, T. Ji, M. Luo, M. Vajapeyam, T.Yoo, O. Song, D. Malladi, A survey on 3GPP heterogeneous networks,IEEE Wireless Communication 18 (3) (2011) 10-21.

[37] D. Lopez-Perez, I. Guvenc, G.D.L. Roche, M. Kountouris, T.Q.S.Quek, J. Zhang, Enhanced intercel interference coordination chal-lenges in heterogeneous networks, IEEE Wireless Communication18(3) (2011) 22-30

[38] X. Y. Wang, P.-H Ho, K.-C. Chen, Interference analysis and mitigationfor cognitive-empowered femtocells through stochastic dual control,IEEE Transaction on Wireless Communication 11 (6) (2012) 2065-2075.

[39] Y.-S. Liang, W.-H. Chung, G.-K. Ni, I.-Y. Chen, H. Zhang, S.-Y. Kuo,Resource allocation with interference avoidance in OFDMA femtocellnetworks, IEEE Transaction on Vehicular Technology 61 (5) (2012)2243-2255.

[40] Y. Sun, R.P. Jover, X. Wang, Uplink interference mitigation forOFDMA femtocell networks, IEEE Transaction on Wireless Commu-nication 11 (2) (2012) 614-625.

[41] O. Bulakci, S. Redana, B. Raaf, J. Hamalainen, Impact of power controloptimization on the system performance of relay based LTE-advancedheterogeneous networks, Journal of Communication & Networking 13(4) (2012) 345-359.

[42] R. Combes, Z. Altman, E. Altman, Self-organizing relays: dimension-ing, self-optimization, and learning, IEEE Transaction on Network andService Management, in press.

[43] ETSI MCC, ”R2-101881: Report of 3GPP TSG RAN WG2 Meeting68bis,” 3GPP TSG RAN WG2 Mtg. 68bis, Feb. 2010.

[44] K.S. Ko, M.J. Kim, K.Y. Bae, D.K. Sung, J.H. Kim, J.Y. Ahn, A novelrandom access for fixed-location machine-to-machine communicationin OFDMA based systems, IEEE Communications Letters 16 (9)(2012) 1428-1430.

[45] ZTE, R2-104662: MTC Simulation Results with Specific Solutions,3GPP TSG RAN WG2 Meeting 71, Aug. 2010.

[46] A. Laya, L. Alonso, and J. Zarate, ”Is the Random Access Channelof LTE and LTE-A Suitable for M2M Communications? A Survey ofAlternatives”, IEEE Communications Surveys & Tutorials, vol. 16, no.1, pp. 4-16, 2014.

[47] CATT, R2-100182: Access Control of MTC Devices, 3GPP TSG RANWG2 Meeting 68bis, Jan. 2010.

[48] S.-Y Lien, T.-H. Liao, C.-Y. Kao, K.-C. Chen, Cooperative access classbarring for machine-to-machine communication, IEEE Transaction onWireless Communication 11 (1) (2012) 27-32.

[49] S.-E. Wei, H.-Y Hsieh, and H.-J. Su, Enabling dense machine-to-machine communications through interference-controlled clustering, in:Proc. IEEE IWCMC, 2012.

[50] C.Y. Ho, C.-Y. Huang, Energy-saving massive access control andresource allocation schemes for M2M communications in OFDMAcellular networks, IEEE Communications Letters 1 (3) (2012) 209-211.

[51] S.-Y. Lien, K.-C. Chen, and Y. Lin, ”Toward ubiquitous massive ac-cesses in 3GPP machine-to-machine communications,” IEEE Commun.Mag., vol. 49, no.4, Apr. 2011, pp. 66-74.

[52] “Machine to Machine (M2M) Communication Study Report,” IEEE802.16ppc-10/0002r7, 2010.

[53] ”IEEE Standard for local and metropolitan area networks Part 16: AirInterface for Broadband Wireless Access Systems,” IEEE Std 802.16-2009 (Revision of IEEE Std 802.16-2004), May 2009.

[54] “IEEE 802.16m Draft Amendment to IEEE Standard for Local andMetropolitan Area Networks,” IEEE P802.16m/D7, July 2010.

[55] K. Doppler, M. Rinne, C. Wijting, C.B. Ribeiro, K. Hugl, Device-to-device communication as an underly to LTE-advanced networks, IEEECommun. Mag., vol 47, no. 12, pp. 42-49, 2009.

[56] 3GPP RAN WS, RWS-120003: LTE Release 12 and Beyond, in: 3GPPRAN WS on Rel-12 and onwards, June 2012.

[57] G. Fodor, E. Dahlman, G. Mildh, S. Parkvall, N. Reider, G. Miklos,Z. Turanyi, Design aspects of network assisted device-to-device com-munications, IEEE Commun. Mag., vol. 50, no. 3, pp. 170-177, 2012.

[58] R. Bhatia, L. Li, H. Luo, and R. Ramjee, ”ICAM: Integrated cellularand ad hoc multicast,” IEEE Trans. Mobile Comput., vol. 5, no. 8, pp.1004-1015, Aug. 2006.

[59] F. Hou, L. X. Cai, P. H. Ho, X. Shen, and J. Zhang, ”A cooperativemulticast scheduling scheme for multimedia services in IEEE 802.16networks,” IEEE Trans. Wireless Commun., vol. 8, no. 3, pp. 1508-1519, Mar. 2009.

Page 25: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 25

[60] Q. Zhang, F. H. P. Fitzek, and V. B. Iversen, ”Design and performanceevaluation of cooperative retransmission scheme for reliable multicastservices in cellular controlled P2P networks,” in Proc. IEEE PIMRC,Athens, Greece, Sep. 2007, pp. 1-5.

[61] S. C. Spinella, G. Araniti, A. Iera, and A. Molinaro, ”Integration of adhoc networks with infrastructured systems for multicast services pro-visioning,” in Proc. Int. Conf. Ultra Modern Telecommun. Workshops,St. Petersburg, Russia, Oct. 2009, pp. 1-6.

[62] F. H. P. Fitzek and M. Katz, Cognitive Wireless Networks: Concepts,Methodologies and Visions Inspiring the Age of Enlightenment ofWireless Communications. New York: Springer- Verlag, 2007.

[63] H. Xing and S. Hakola, ”The investigation of power control schemesfor a device-to-device communication integrated into OFDMA cellularsystem,” in Proc. IEEE PIMRC, Istanbul, Turkey, Sep. 2010, pp. 1775-1780.

[64] J. Lehtomaki, I. Suliman, J. Vartiainen, M. Bennis, and K. Umbayashi,”Direct communication between terminals in infrastructure based net-works,” in Proc. ICT-MobileSummit, Stockholm, Sweden, Jun. 2008,pp. 1-8.

[65] L. Al-Kanj and Z. Dawy, ”Optimized energy efficient content distri-bution over wireless networks with mobile-to-mobile cooperation,” inProc. IEEE 17th Int. Conf. Telecommun., Apr. 2010, pp. 471-475.

[66] B. Wang, L. Chen, X. Chen, X. Zhang, and D. Yang, ”Resource al-location optimization for device-to-device communication underlayingcellular networks,” in Proc. IEEE VTC, Budapest, Hungary, May 2011,pp. 1-6.

[67] C. H. Yu, O. Trikkonen, K. Doppler, and C. Ribeiro, ”Power optimiza-tion of device-to-device communication underlaying cellular commu-nication,” in Proc. IEEE ICC, Dresden, Germany, Jun. 2009, pp. 1-5.

[68] C. H. Yu, K. Doppler, C. Ribeiro, and O. Tirkkonen, ”Resource sharingoptimization for device-to-device communication underlaying cellularnetworks,” IEEE Trans. Wireless Commun., vol. 10, no. 8, pp. 2752-2763, Aug. 2011.

[69] L. Popova, T. Herpel, W. H. Gerstacker, and W. Koch, ”Cooperativemobile-to-mobile file dissemination in cellular networks within aunified radio interface,” Int. J. Comput. Netw., vol. 52, no. 6, pp. 1153-1165, Apr. 2008.

[70] A. A. Khalek and Z. Dawy, ”Energy-efficient cooperative video distri-bution with statistical QoS provisions over wireless networks,” IEEETrans. Mobile Comput., vol. 11, no. 7, pp. 1223-1236, Jul. 2012.

[71] B. Xing, K. Seada, and N. Venkatasubramanian, ”An experimentalstudy on Wi-Fi ad-hoc mode for mobile device-to-device video de-livery,” in Proc. IEEE INFOCOM, Rio de Janeiro, Brazil, Apr. 2009,pp. 1-6.

[72] L. Popova, T. Herpel, and W. Koch, ”Enhanced downlink capacity inUMTS supported by direct mobile-to-mobile data transfer,” in Proc.6th Int. Conf. IFIP-TC6 Netw., Atlanta, GA, May 2007, pp. 522-534.

[73] E. Hossain, D. Niyato, and Z. Han, ”Dynamic Spectrum Access andManagement in Cognitive Radio Networks,” Cambridge UniversityPress, 2009.

[74] Y. Zhang, R. Yu, M. Nekovee, L. Yi, S. Xie and S. Gjessing, ”CognitiveMachine-to-Machine Communications: Vision and Potential for theSmart Grid”, IEEE Network, vol. 26, no. 3, pp. 6-13, Jun. 2012.

[75] H.-K. Lee, D. Kim, Y. Hwang, S. Min, and S.-L. Kim, ”Feasibil-ity of Cognitive Machine-to-Machine Communication Using CellularBands”, IEEE Wireless Commun., vol. 20, no. 2, pp. 97-103, Apr.2013.

[76] Y. Zhang, R. Yu, S. Xie, W. Yao, Y. Xiao, and M. Guizani, ”HomeM2M networks: architectures, standards, and QoS improvement,” IEEECommun. Mag., vol. 49, no. 4, pp. 44-52, 2011.

[77] Z.M. Fadlullah, M.M. Zubair, N. Kato, A. Takeuchi, N. Iwasaki, andY. Nozaki, ”Toward Intelligent Machine-to-Machine Communicationsin Smart Grid,” IEEE Commun. Mag., vol. 49, no. 4, pp. 60-65, 2011.

[78] B.H. Lee and S.L. Kim, ”Mobility Control for Machine-to-MachineLTE Systems,” Wireless Conference 2011-Sustainable Wireless Tech-nologies (European Wireless), 11th European. VDE, pp. 1-5, 2011.

[79] C. Ho, and C. Y. Huang, ”Energy-Saving Massive Access Control andResource Allocation Schemes for M2M Communications in OFDMACellular Networks,” IEEE Wireless Commun. Lett., vol. 1, no. 3, pp.209-212, Jun. 2011.

[80] K. Zheng, F. Hu, W. Wang, W. Xiang, and M. Dohler, ”RadioResource Allocation in LTE-Advanced Cellular Networks with M2MCommunications,” IEEE Commun. Mag., vol. 50, no. 7, pp. 184-192,Jul. 2012.

[81] S. Sesia, I. Toufik, and M. Baker, LTE - The UMTS Long TermEvolution - From Theory to Practice, Wiley, 2009.

[82] M. Hasan, E. Hossain, and D. Niyato, ”Random Access for Machine-to-Machine Communication in LTE-Advanced Networks: Issues andApproaches”, IEEE Commun. Mag., vol. 51, no. 6, pp. 86-93, Jun.2013.

[83] J. Seo and V. Leung, ”Performance Modeling and Stability of Semi-Persistent Scheduling with Initial Random Access in LTE”, IEEETransactions on Wireless Communications, vol. 11, no. 12, pp. 4446-4456, Dec. 2012.

[84] M. H. Cheung, H. Mohsenian-Rad, V. Wong, and R. Schober, ”Utility-Optimal Access for Wireless MUltimedia Networks”, IEEE WirelessCommunications Letters, vol. 1, no. 4, pp. 340-343, Aug. 2012.

[85] C. Joo, ”On Random Access Scheduling for Multimedia Traffic in Mul-tihop Wireless Networks with Fading Channels”, IEEE Transactions onMobile Computing, vol. 12, no. 4, pp. 647-656, App. 2013.

[86] J. Seo and V. Leung, ”Design and Analysis of Backoff Algorithms forRandom Access Channels in UMTS-LTE and IEEE 802.16 Systems”,IEEE Transactions on Vehicular Technology, Vol. 60, no. 8, pp. 3975-3989, Oct. 2011.

[87] J. Seo and V. Leung, ”Approximate Queuing Performance of a Multi-packet Reception Slotted ALOHA System with an Exponential BackoffAlgorithm”, Fourth International Conference on Communications andNetworking, China, 2009, pp. 1-5.

[88] R. Tyagi, F. Aurzada, K.-D. Lee, S. kim, and M. Reisslein, ”Impactof Retransmission Limit on Preamble Contention in LTE-AdvancedNetwork” IEEE Systems J., pp. 1-14, Oct. 2013.

[89] 3GPP TR 37.868 V11.0.0, ”Study on RAN Improvements for MachineType Communications,” September 2011.

[90] A. Lo, Y. W. Law, and M. Jacobsson, ”Enhanced LTE-AdvancedRandom-Access Mechanism for Massive Machine-to-Machine (M2M)Communications,” Proc. 27th Meeting of Wireless World ResearchForum, Oct. 2011.

[91] P. Huang, L. Xiao, S. Soltani, M. Mutka, and N. Xi, ”The evolution ofMAC protocols in wireless sensor networks: A survey,” IEEE Commun.Surveys & Tutorials, vol. 15, no. 1, pp. 101-120, 2012.

[92] W. Ye, J. Heidemann, and D. Estrin, ”Medium access control with co-ordinated, adaptive sleeping for wireless sensor networks,” IEEE/ACMTrans. Netw., vol. 12, no. 3, pp. 493-506, Jun. 2004.

[93] B. Villaverde, R. Alberola, A. Jara, S. Fefor, S. Das, and D. Pesch,”Service Discovery Protocols for Constrained Machine-to-MachineCommunications,” IEEE Commun. Surveys & Tutorials, vol. 16, no.1, pp. 41-60, 2014.

[94] Z. Zhang, K. Long, J. Wang, and F. Dressler, ”On Swarm IntelligenceInspired Self-Organized Networking: Its Bionic Mechanisms, Design-ing Principles and Optimization Approaches,” IEEE Commun. Surveys& Tutorials, vol. 16, no. 1, pp. 513-537, 2014.

[95] L. Atzori, A. Iera, and G. Morabito, ”The Internet of Things: A survey,”Computer Networks, vol. 54, no. 15, pp. 2787-2805, X. 2010.

[96] ETSI TR 102 732, ”Machine to Machine Communications (M2M):Use cases of M2M applications for eHealth”, 2013.

[97] P. Spiess, S. Karnouskos, D. Savio, O. Baecker, L. Souza, and V. Trifa,SOA-based integration of the internet of things in enterprise services,in proceedings of IEEE ICWS 2009, Los Angeles, Ca, USA, Jul. 2009.

[98] S. Karpischek, F. Michahelles, F. Resatsch, E. Fleisch, Mobile salesassistant - an NFC-based product information system for retailers,in: Proceedings of the First International Workshop on Near FieldCommunications 2009, Hagenberg, Austria, February 2009.

[99] www.worldsensing.com[100] www.sensefield.com[101] www.bitcarrier.com[102] A. Ilic, T. Staake, E. Fleisch, Using sensor information to reduce the

carbon footprint of perishable goods, IEEE Pervasive Computing 8 (1)(2009) 22-29.

[103] A. Dada, F. Thiesse, Sensor applications in the supply chain: theexample of quality-based issuing of perishables, in: Proceedings ofInternet of Things 2008, Zurich, Switzerland, May 2008.

[104] D. Niyato, L. Xiao, and P. Wang, ”Machine-to-Machine Communica-tions for Home Energy Management System in Smart Grid”, IEEECommun. Mag., vol. 49, no. 4, pp. 53-59, Apr. 2011.

[105] m2m.vodafone.com[106] International Energy Agency, ”CO2 emissions from fuel combustion:

Highlights”, 2006.[107] European Commission, ”Energy efficiency 2020 horizon and low-

carbon energy by 2050”, Available: http://ec.europa.eu/news/energy[108] O. Shigeru, M2M and big data to realize the smart city, NEC Technical

Journal, vol. 7, no. 2, 2012.[109] LIBELIUM. Libelium Communications Distribuidas. 2013. Retrieved

May 14, 2013 from /http://www.libelium.com/.

Page 26: IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. … · M2M Communications in 3GPP LTE/LTE-A Networks: Architectures, Service Requirements, Challenges, and Applications Fayezeh

1553-877X (c) 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. Seehttp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/COMST.2014.2361626, IEEE Communications Surveys & Tutorials

IEEE COMMUNICATION SURVEYS AND TUTORIALS, VOL. XX, NO. Y, MONTH 2014 26

[110] I. Demirkol, F. Alagoz, H. Delic, C. Ersoy, Wireless sensor networksfor intrusion detection: packet traffic modeling, IEEE CommunicationLetters 10 (1) (2006) 22-24.

[111] M. Z. Shafiq, L. Ji, A. X. Liu, J. Pang, and J. Wang, ”A FirstLook at cellular Machine-to-Mchine Traffic: Large Scale Measure-ment and Characterization,” in Proc. of the 12th ACM SIGMET-RICS/PERFORMANCE Joint International Conference on Measure-ment and Modeling of Computer Systems, London, England, UK, 2012,pp. 65-76.

[112] V. Krylov, A. Logvinov, D. Ponomarev, EPC Object Code MappingService Software Architecture: Web Approach, MERA Networks Pub-lications, 2008.

[113] I. F. Akyildiz, W. Su, Y. Sankarasubramaniam, and E. Cayirci, ”Asurvey on sensor networks,” IEEE Commun. Mag., vol. 40, no. 8, Aug.2002. pp. 102-14.

[114] S. Ehsan and B. Hamdaoui, ”A Survey on Energy-Efficient RoutingTechniques with QoS Assurances for Wireless Multimedia SensorNetworks,” IEEE Commun. Surveys & Tutorials, Vol. 14, No. 2, 2012,pp. 265-278.

[115] N. A. Pantazis, S. A. Nikolidakis, and D. D. Vergados, ”Energy-Efficient Routing Protocols in Wireless Sensor Networks: A Survey,”IEEE Commun. Surveys & Tutorials, Vol. 15, No. 2, 2013, pp. 551-591.

[116] T. Watteyne, A. Molinaro, M. G. Richichi, and M. Dohler, ”FromMANET To IETF ROLL Standardization: A Paradigm Shift in WSNRouting Protocols,” IEEE Commun. Surveys & Tutorials, Vol. 13, No.4, 2011, pp. 688-707.

[117] R. Kumar, E. Kohler, M. Srivastava, Harbor, ”Software-based MemoryProtection for Sensor Nodes,” in: proceedings of IPSN 2007, Cam-bridge, MA, USA, April 2007.

[118] H. Krawczyk, M. Bellare, R. Canetti, ”HMAC: Keyed-Hashing forMessage Authentication,” IETF RFC 2104, February 1997.

[119] R. Acharya and K. Asha, Data integrity and intrusion detection inwireless sensor networks, in: Proceedings of IEEE ICON 2008, NewDelhi, India, December 2008.

[120] 3rd Generation Parnership Project (3GPP). (2007, May). Feasibility

study on remote management of USIM application on M2M equipment.3GPP Tech. Rep. 33.812, unpublished draft version 1.4.0.

Fayezeh Ghavimi ([email protected]) iscurrently a Ph.D. student in the Department of En-gineering Science, National Cheng Kung University,Taiwan. She obtained her M.Sc. degree in 2012 andher B.Sc. degree from the University of Tabriz, Iran,in 2007, all in electrical engineering. She received?Graduate Research Award? from the EngineeringScience Department of National Cheng Kung Uni-versity in 2012. Her research interests include wire-less communications, machine-to-machine commu-nications, QoS provision for supporting next gener-

ation wireless communications, and next generation CDMA networks. She isa graduate student member of IEEE.

Hsiao-Hwa Chen ([email protected]) is currentlya Distinguished Professor in the Department ofEngineering Science, National Cheng KungUniversity, Taiwan. He obtained his BSc andMSc degrees from Zhejiang University, China,and a PhD degree from the University of Oulu,Finland, in 1982, 1985 and 1991, respectively.He is the founding Editor-in-Chief of Wiley?sSecurity and Communication Networks Journal(http://www.interscience.wiley.com/security).Currently, he is also serving as the Editor-in-Chief

for IEEE Wireless Communications. He is a Fellow of IEEE, a Fellow ofIET, and an elected Member at Large of IEEE ComSoc.