A Study on the Efficient Wireless Sensor Networks

for Operation Monitoring and Control in Smart Grid

Applications Md Monirojjaman Monshi, Osama A. Mohammed

Department of Electrical and Computer Engineering, Florida International University

Email: mmons021@fiu.edu, mohammed@fiu.edu

Abstract— Low cost and collaborative nature of wireless sensor

network makes it promising reliable, secure and self-healing two-

way communication for monitoring and controlling of the

equipment in the smart grid. But for time critical applications

like grid fault isolation and recovery, demand response, wide

area situational awareness demand low latency enabled faster

communication network infrastructure. The grid can be really

smart enough when we can control the grid in real time without

any potential blackout. To meet this challenge this paper

evaluates the performance of IEEE standard 802.15.4 wireless

sensor networks with varying network sizes. NS2 (Network

Simulator 2) has been employed to evaluate the performance of

the wireless sensor networks using various ad-hoc routing

protocols. The paper then analyzes the results and finally reaches

to a conclusion about the potential areas where wireless sensor

network can be deployed for efficient operation monitoring and

control of the smart grid.

Keywords— Smart grid, Wireless Sensors Networks (WSNs), Ad-

hoc routing protocols, latency, packet delivery ratio, IEEE 802.15.4

I. INTRODUCTION

Installing large wired communication system for monitoring

the power grid costs time and money. Because it is required to

set up additional equipments and cables in already cluttered

power grid facilities. Moreover, whenever any fault occurs in

the system, communication becomes difficult, sometimes even

impossible. Only wireless sensor network can ease this

problem for the grid. Low cost wireless sensor has paved the

way for grid automation, real time monitoring and remote

control of system elements such as primary and secondary sub

stations, power lines, capacitor banks, feeder switches, fault

indications and other physical facilities. Wireless Sensor

Networks (WSNs) with its affordable low cost and numerous

features enable utilities to monitor its remote facilities any

time with applications such as SCADA.

According to the recent Annual Energy Outlook report of

the U.S. Energy Information Administration, residential

electricity demand is forecasted to increase by 24% within the

following several decades [1], while the global electricity

consumption trend is also reported to be increasing

continuously [2]. The negative impacts of rising consumption

are becoming more evident with the diminishing fossil fuels

and accumulating greenhouse gases. Moreover, the mismatch

between demand and supply and lack of automation and

monitoring tools have already caused major blackouts

worldwide. Apparently, the traditional power grid has shown

signs of inefficient operation and has been experiencing

difficulties in meeting the requirements of the 21st century. As

a result, the Energy Independence and Security Act of 2007

gave a start for the smart grid implementation in the United

States. U.S. Department of Energy proposed the requirements

on the smart grid communication which essentially needs to

be two-way flow of electricity with the two-way flow of

information to revolutionize electricity generation and

delivery [3].

The need for having smart grid smart communication

infrastructure is insufficient for precise operation, monitoring

and control. In addition to the overstressed situation, the

existing power grid also suffers from the lack of pervasive and

effective communications, monitoring, fault diagnostics, and

automation, which further increase the possibility of region-

wide system breakdown due to the cascading effect initiated

by a single fault [4].Specially latency is a great issue for fault

monitoring and control of the grid to avoid potential black out

from cascading effect. Cascading effect can be avoided if the

fault can be identified in real time. But the existing

communication infrastructure with high delay hinders the real-

time control of the grid. Latency requirement of the network

for smart grid is shown in the Table I below [3].

TABLE I

SMART GRID FUNCTIONALITIES AND COMMUNICATIONS NEEDS

Application Bandwidth Latency

AMI 10-100 kbps/node,

500 kbps for backhaul

2-15 sec

Demand Response 14kbps- 100 kbps

per node/device

500 ms - several

minutes

Wide Area Situational

Awareness

600-1500 kbps 20 ms-200 ms

Distribution Energy

Resources and

Storage

9.6-56 kbps 20 ms-15 sec

Electric

Transportation

9.6-56 kbps, 100 kbps is a good target

2 sec-5 min

Distribution Grid

Management

9.6-100 kbps 100 ms-2 sec

In this paper, we focus on simulation of an ad-hoc wireless

network to evaluate the performance and latency of the WSNs

in order to establish their suitability for a typical set of

monitoring and supervision functionalities required by urban-

978-1-4799-0053-4/13/$31.00 ©2013 IEEE

scale Smart Grids applications. In the WSNs, sensor nodes use

IEEE 802.15.4 standard for communication.

Zigbee is used in this paper, since it is a strong candidate

for smart grid energy management applications. Zigbee is a

short-range, low-data rate, energy-efficient wireless

technology that is based on the IEEE 802.15.4 standard.

Zigbee utilizes 16 channels in the 2.4 GHz ISM band

worldwide, 13 channels in the 915MHzband in North America,

and one channel in the 868 MHz band in Europe. The

supported data rates are 250 kbps, 100 kbps (available in

IEEE 802.15.4-2006), 40 kbps, and 20 kbps, and its range

varies between 30m–90m indoors. Zigbee can support up to

64,000 nodes (devices). Zigbee certified devices can work for

several years without the need for battery replacement due to

the low duty cycle mechanism.

The performance evaluation of WSNs has been carried out

in the NS2 (Network Simulator 2). Different ad-hoc routing

protocols have been employed to asses which routing

protocols perform better in terms of latency, packet delivery

ratio and throughput. The results obtained show that the

application of WSNs based communication services exhibits a

set of intrinsic advantages, particularly useful in Smart Grids

control with optimum performance and enabling real-time

monitoring.

The remainder of the paper is organized as follows. In

section II related works have been presented. Section III

introduces different kinds of ad-hoc routing protocols for

IEEE standard 802.15.4. In section IV an overview of NS-2

architecture is shown and section V consists of evaluation

criteria and parameters for the simulation in NS-2.Section VI

summarizes simulation results and finally section VII

concludes the paper.

II. RELATED WORKS

In the literature, WSNs performance evaluations have

been studied in several works. In [5], the authors evaluate the

performance of in home energy management (iHEM), then

compared iHEM to optimization-based residential energy

management (OREM).They also evaluate packet delivery

ratio, latency and jitter in their proposed Wireless Sensor

Home Area Network (WSHAN) for iHEM.

In [6], the authors uses IEEE 802.11s standard for

evaluating wireless multigate mesh routing to improve latency

and throughput performance. They used neighbour nodes to

reach one of the gateways in the network in multi hop fashion

to reduce latency and throughput. But average end to end

delay increases significantly when bit rate increases.

In [7], wireless Link-Quality Estimation has been done in

smart grid environments using different estimators such as

Window Mean with Exponentially Weighted Moving Average

(WMEWMA), Expected Transmission Count (ETX), and

four-bit link quality estimation methods employed to find the

best one. The criteria for performance evaluation were packet

delivery ratio, average number of packet retransmissions,

average number of parent changes, average number of hops

and average communication delay.

In the paper [8], the role of pervasive and cooperative

sensor networks in smart grids communication has been

evaluated. The authors analyzed the performance of IEEE

802.15.4 based WSNs. They used qualnet developer platform

simulation environment. They observed that latency increases

due to rain, number of hops, end to end delay, and end nodes

delay to coordinator (number of active connections effectively

supported by the network). They proposed ad-hoc network in

order to improve the network performances in terms of data

latency, number of simultaneous and active connections.

III. ROUTING PROTOCOLS FOR WSNS

Large number of sensor nodes in the smart grid makes it

inevitable to employ self-healing, reconfigurable network for

faster response and control of the grid. Low cost WSNs pave

the way in this purpose but latency is the great problem [10].

To alleviate the latency problem choice of routing protocol is

one of the prominent solutions. Because IP-based global

addressing scheme is not possible as wireless networks have

to accommodate large number of sensor nodes [12]. Large

number of traffic from sensor nodes to a specific sink,

transmission power, on-board energy, processing capacity,

storage limitations of sensor nodes require careful resource

management.

The aforementioned limitations of sensor nodes demand a

careful implementation of WSNs. To prepare efficient WSNs

different routing algorithms have been coined to improve

latency, packet delivery ratio and traffic congestion. Ad-hoc

routing protocol can improve the latency of WSNs for

extreme environmental condition, large traffic handling, self-

healing [9, 10, 11]. There are two kinds of ad-hoc routing

protocols: table-driven protocol maintains routing table of the

whole network and on-demand protocols keep routes when it

is required. There is another type of protocol which is the

combination of table driven and on-demand protocol.

Destination Sequence Distance Vector (DSDV):

DSDV is a proactive routing protocol which maintains a

routing table containing all routes between all source

destination pairs. DSDV is a hop-by-hop distance vector

routing protocol where each node maintains routing

information in the form of look up table which broadcasted

periodically.

Dynamic Source Routing (DSR):

DSR uses source routing where each data packet header

contains complete route of the nodes through which it passes.

So intermediate nodes do not need to maintain updated

routing table for the packets they forward as the packets

already have the information for routing decisions.

Ad-hoc Distance Vector (AODV):

AODV is the combination of two routing protocols DSR

and DSDV. It uses route discovery and route maintenance

from DSR as well as hop-by-hop routing, sequence numbers

and periodic beacons from DSDV.

AOMDV protocol:

An extension to AODV is ad-hoc on-demand Multipath

Distance Vector (AOMDV) routing protocol which is for

computing multiple loop-free and link disjoint paths. For each

destination, along with the respective hop counts it contains a

list of the routing entries of the next-hops. Same sequence

number is allocated to all next hops. This helps for keeping

track of a route. A node maintains the assigned hop count,

which is the maximum hop count for all the paths at each node.

Loop freedom is assured for a node by accepting another path

to destination if it has a less number of hop counts than the

assigned for that destination. AOMDV allows intermediate

nodes to reply to RREQs, while still selecting disjoint paths.

During route discovery, its message overhead is high, due to

increased flooding. Since it is a multipath routing protocol, the

destination replies to the multiple RREQs those results are in

longer overhead [13].

IV. NETWORK SIMULATION MODEL (NS2)

NS2 simulator deal with transport layer protocols, routing

protocols, different types of queues, link layer mechanisms.

This simulator gives the essence of dealing with practical

network structure components such as routers, switches,

bridges, wired and wireless nodes. Once simulation completed

NS2 produces trace files which keep the record of every event

in the simulation line by line. NS2 consists of C++

programming language, Tcl and Object Tcl interface allowing

the user to put inputs in the simulation script of the network

model. User can define network with different topologies,

standards, protocols consisting wired and wireless nodes,

routers, bridges, switches, links and shared media.

Figure -1: Layered structure of NS2

Layered structure of NS2 is presented here in the Figure -

1. The event schedulers, traffic pattern, network components

are implemented in C++ and can be merged to Tcl script. Core

structure of network is built on C++ code and Tcl is on the top

of it to make the simulation handling much easier to carry out. The overview of the network can be seen then upon the Tcl

level. AWK script is then used to measure the performance

evaluation of the simulated network from the trace files. These

are the things which are combined in NS2 software.

V. EVALUATION OF PROPOSED WSNS FOR

DIFFERENT AD-HOC ROUTING PROTOCOLS

Performance Metrics used for evaluation of routing

protocols are [9, 10, 11, 12]:

Packet Delivery Ratio

Measures the percentage of total number of data packets

received out of total number of data packets sent.

Routing Over Head

The total number of routing packets transmitted during

simulation. Routing overhead is important as it measures the

scalability of a protocol, the degree to which it will function in

congested or low bandwidth environments.

End-to-End Delay of Data Packets

This metric measure the average time it takes to route a

data packet from the source node to the destination node. The

lower the end-to-end delay the better the application

performance. If the value of end-to-end delay is high then it

means the protocol performance is not good due to the

network congestion.

TABLE II

SIMULATION PARAMETERS

Simulation Dimension 500*500

Number of Nodes Vary according to experiment

Propagation Two Ray Ground

Routing queue Drop tail

Mac protocol 802.15.4

Antenna Omni Directional

Traffic type cbr

Wireless Sensor Node

Figure- 2 Multi-hop ad-hoc WSN

Throughput

Total data traffic in bits/sec successfully received and

forwarded to the higher layer. Throughput shows protocol’s

successful deliveries for a time; this means that the higher

throughput, the better will be the protocol performance. It

measures the amount of data received by the destination node

within certain period of time. In multi-hop environment, the

throughput is computed at the final destination as the

intermediate nodes are responsible of relaying the packets.

The proposed wireless architecture consisting 25, 16, 9

nodes with 90 meters spacing are arranged in grid pattern. A

network consisting of 25 wireless sensor nodes is depicted in

the Figure -2.

VI. RESULTS

Four ad-hoc routing protocols DSDV, DSR, AODV and

AOMDV have been used for the performance evaluation with

varying node sizes 25,16 and 9 in grid pattern where last one

node is used as sink node for all nodes. In this simulation only

two criteria latency and packet delivery ratio have been used

for performance measurement. Packet size has been chosen

256 bytes and packet interval was introduced to emulate

bandwidth of the system. Here, the packet inter-arrival time

has been chosen from 0.02s to 1s meaning bandwidth of the

system will vary from 2kbps to 100kbps.

Figure-3.1: Packet interval vs. latency for 25 grid connected nodes

Variation of latency is presented with different packet

inter-arrival time in the Figure- 3.1.Actaully packet inter-

arrival was introduced to simulate the variation of network

speed and to observe the characteristics of it with applying

different ad-hoc routing protocols. It can be observed that

when packet inter-arrival time is 0.02s (that means the system

speed is about 100 kbps) all of the routing protocols showed

high latency as high as around 0.2 s for AOMDV except DSR

which showed as low as 8 ms. As the packet arrival time

increases all of the routing protocols showed lowered latency

for the same system. That means when the system generates

less traffic it can perform better. The packet delivery ratio is

presented in the figure 3.2 and it can be observed that the

system’s packet delivery ratio is best for DSR with varying

traffic generation rate which is around 90%.

Performance indication of WSN is also observed for

different network sizes such for 16 and 9 wireless nodes

system which is presented in the figures 4.1, 4.2, 5.1, 5.2.

It also observed that Packet delivery ratio and latency

improved much for smaller networks as evident from the

figures available for 16 and 9 nodes network. Such as for a

WSN with 9 nodes shows 100% packet delivery ratio and

latency of 4 ms which evident from figure 5.2 and 5.1

respectively.

Figure -3.2: Packet interval vs. Packet delivery ratio for 25 grid connected

nodes

Figure- 4.1: Packet interval vs. latency for 16 grid connected nodes

Figure- 4.2: Packet interval vs. packet delivery ratio for 16 grid connected nodes

Figure- 5.1: Packet interval vs. latency for 9 grid connected nodes

Figure- 5.2: Packet interval vs. packet delivery ratio for 9 grid connected

nodes

With analysis of the graph it can be infer that for large or

small wireless sensor network only DSR routing protocol

showing better performance than any other ad-hoc routing

protocols like DSDV, AODV and AOMDV in terms of

improved latency and higher packet delivery ratio.

VII. CONCLUSION

Network simulator tool has been employed to analyze the

performance of ad-hoc routing protocols using the above

simulation parameters. It can be concluded that WSNs

perform better with comparatively low speed communication

(2kbps~100kbps) which makes it suitable for smart grid

applications like AMI, distributed automation, distributed

energy resources and storage, electric transportation,

distribution grid management according to the communication

requirements of the smart grid as indicated by the Table-I [3].

And necessary routing protocol can be DSR ad-hoc routing

protocol to operate WSN better. It is

also observed that WSNs meet the latency (700ms~4ms)

requirement of the smart grid communication for time critical

applications but they work better for short range

communication (30m~90m). Wireless sensor network can be

used to collect data and sending control information to the

device in distributed manner for the last mile two-way

communication which makes it a promising option for the

smart grid communication. In future the performance of the

WSNs can be analyzed with hardware implementation in a

real smart grid environment as Florida International

University has state of the art smart grid test bed facility. The

performance of WSN then can be compared with that of the

existing wired communication infrastructure. WSNs can also

be implemented for operation monitoring of the smart grid

and control of smart capacitor bank that can monitor and

control capacitor banks remotely, Volt/VAR Control,

Substation circuit breaker trip coil status and control, State

Estimation, Fault Detection/Isolation.

REFERENCES

[1] J. J. Conti, P. D. Holtberg, J. A. Beamon, A. M. Schaal, G. E.Sweetnam, and A. S. Kydes, Annual energy outlook with

projections to 2035, report of U.S. Energy Information Administration

(EIA), Apr. 2010 [Online]. Available: http://www.eia.doe.gov

[2] T. F. Garrity, “Getting smart,” IEEE Power Energy Mag., vol. 6, no. 2,pp. 38–45, Apr.–Mar. 2008.

[3] Report on “Communications Requirements Of Smart Grid Technologies”, US Department of Energy, Oct. 2010

[online].Available: http://www.smartgrid.gov/

[4] Vehbi C. Gungor, Bin Lu and Gerhard P. Hancke, “Opportunities and

Challenges of Wireless Sensor Networks in Smart Grid”, IEEE

Transactions on Industrial Electronics, vol. 57, no. 10, pp. 3557-3564, Oct. 2010

[5] Melike Erol-Kantarci and Hussein T. Mouftah," Wireless Sensor

Networks for Cost-Efficient Residential Energy Management in the Smart Grid", IEEE Transactions on Smart Grid, vol. 2, no. 2,pp. 314-

325, June 2011

[6] Hamid Gharavi,Bin Hu, "Multigate Communication Network for Smart Grid", Proceedings of the IEEE, Volume: 99 , Issue: 6, Page(s): 1028 -

1045 June 2011

[7] V. C. Gungor and M. K. Korkmaz, “Wireless Link-Quality Estimation in Smart Grid Environments” , Hindawi Publishing Corporation,

International Journal of Distributed Sensor Networks, vol.2012,

Article ID 214068,10pages, doi:10.1155/2012/214068 [8] Silvia Ullo,Alfredo Vaccaro,Giovanni Velotto, “ The role of Pervasive

and Cooperative Sensor Networks in Smart Grids Communication”,

MELECON 2010 - 2010 15th IEEE Mediterranean Electrotechnical Conference, Page(s): 443 - 447, Date of Conference: 26-28 April 2010

[9] D.D.Chaudhary, Pranav Pawar, Dr. L.M. Waghmare, "Comparison and

Performance Evaluation of Wireless Sensor Network with different Routing Protocols", International Conference on Information and

Electronics Engineering, 2011, IPCSIT vol.6 (2011) © (2011) IACSIT

Press, Singapore [10] Siti Rahayu, Abdul Aziz, Nor Adora Endut, Shapina Abdullahand Mior

Norazman Mior Daud, “Performance Evaluation of AODV, DSR and

Dymo Routing Protocol In Manet” Conference on scientific & Social research, Cssr 08’09, 14-15 March 2009.

[11] Woon, W.T.H.; Wan, T.C.; , "Performance Evaluation of IEEE

802.15.4 Ad Hoc Wireless Sensor Networks: Simulation Approach," Systems, Man and Cybernetics, 2006. SMC '06. IEEE International

Conference on , vol.2, no., pp.1443-1448, 8-11 Oct. 2006, doi:

10.1109/ICSMC.2006.384920

[12] Poulomi Goswami, Dr.A.D.Jadhav, "Evaluating the Performance of

Routing Protocols in. Wireless Sensor Networks", International

Journal of Computer Technology and Electronics Engineering (IJCTEE), Vol: 2, Issue: 3, June 2012

[13] M.K.Marina and S.R.Das, “On-Demand multipath distance vector

routing in ad hoc networks” in: Proceedings of the 9th IEEE International Conference on Network Protocols (ICNP), 2001.