Android Based Body Area Network for theEvaluation of Medical Parameters

Matthias Wagner1, Benjamin Kuch2, Carlos Cabrera1, Peter Enoksson3, Arne Sieber41 FH Frankfurt am Main - University of Applied Sciences, Germany

2 Scuola Superiore SantAnna - RETIS Lab, Pisa, Italy3 Chalmers University of Technology, Gothenburg, Sweden

4 Institute of Micro and Nano Technology (IMEGO AB), Gothenburg, Sweden

AbstractThe telemedical system focuses on the measurementand evaluation of vital parameters, e.g. ECG, heart rate,heart rate variability, pulse oximetry, plethysmography andfall detection. Based on two different designs of a (Wireless)Body Area Network connected to an Android smartphone theReal-Time system features several capabilities: Data acquisitionin the (W)BAN plus the use of the smartphone sensors, patientlocalization, data storage, analysis and visualization on thesmartphone, data transmission and emergency communicationwith first responders and a clinical server. In the first ZigBeebased approach smart and energy efficient sensor nodes acquirephysiological parameters, perform signal processing and dataanalysis and transmit measurement values to a coordinatornode. In the second design sensors are connected via cable toan embedded system. In both approaches data are transferredvia Bluetooth to an Android based smartphone.

Several challenges are discussed: Measuring, analysing andvisualizing medical parameters characterize the system as safetycritical, requiring special development procedures and adherenceto safety standards. Reliability of wireless data transmissionhas to be optimized. Handling medical data requires securitymeasures on each level of the system hierarchy.

I. INTRODUCTIONMonitoring and recording of physiological parameters

of patients outside the clinical environment is becomingincreasingly important in research as well in appliedphysiology and medicine in general. Environmentalphysiology as one important example, is a scientific discipline[1] that gained significance with the continuing advances intechnology exposing humans to greater extremes and extremeenvironmental conditions. For instance extreme sports likeendurance running, climbing and high altitude mountaineeringare more popular also in recreational settings. Presently verylittle is known about acute adaptation mechanisms, andespecially about long term changes in physiological function,e.g. in professionals which are regularly exposed to extremeenvironmental conditions such as divers, astronauts or pilots.How age and gender influence these adaptations is also largelyunknown. A sound understanding of human physiology insuch environments is however the basis for being able to giverecommendations and draft guidelines on how and to whatextent exposures to extreme environments can be toleratedin a safe way with minimized health risks considering short,medium as well as long term effects. Technology is advancing

but knowledge of physiology is lagging behind thus there isan urgent need for rapid advances in these research topics.Environmental Physiology is the area of life science thatdescribes the human physiological and behavioral changes,in particular, acute responses, adaptations, habituation,acclimation and acclimatization. The main obstacle to theassessment of physiological changes in extreme environmentis the fact that most findings have been collected duringlaboratory conditions with frequently bulky instruments. Infield measurements outside laboratory are not feasible, simplyas suitable instrumentation that can withstand such harshenvironments was not available. Thus research is far awayfrom the real field conditions when different environmentalfactors could act synergistically or vice versa an individualresponse to environmental stimuli is complex. Moreover,in the general context of the reductionistic (cellular andmolecular) wave that has swept over biomedical researchduring the last 20-30 years, many fundamental physiologicalmechanisms still continue to be poorly understood [2].Applying laboratory measurement equipment in field is notenough, as measurement equipment and the measurementsetup are also subject to cross interferences caused byenvironmental stressors. Instrumentation developed especiallyfor application in extreme environments is the key forrapid advances in environmental physiology. Physiologicalparameters of paramount interest are among others: O2saturation (SO2), arterial blood pressure, heart rate andvariability, breathing parameters, etc. Diving is one exampleof an extreme environment and as such a research fieldof environmental physiology. Medical concerns about suchactivities derive from two major shortcomings: scantyknowledge of diving physiology and lack of monitoringof vital parameters during diving. Both deficiencies arevirtually related to the absence of instrumentation suitable forunderwater measurements of simple but crucial physiologicalparameters. First prototypes were developed in the past byour team. [3], [4]

Another important medical field, where autonomous datarecording and real time wireless data transmission is ofbenefit, is in preventive medicine and post treatment/surveyof patients with cardiac heart failure (HF) pathology and/or

10th International Workshop on Intelligent Solutions in Embedded Systems, 2012

~33

cardiovascular diseases. HF represents a diffused pathology inthe industrialized countries with high morbidity and mortalityrisk and a subsequent high request for hospitalization. Infact, in these countries, the first mortality cause is related tothe cardiovascular diseases: in 2005 the US AHRQ (Healthand Human Services Agency for Healthcare Research andQuality) reported the costs of the top ten morbid conditions:76 B$ for cardiovascular pathologies (48 B$ related to thenecessary hospitalization), 69 B$ for cancer, 39 B$ fordiabetes and 39 B$ for osteo-articular diseases. The incidenceand the prevalence of HF, as final status of all the chroniccardio-pathologies, are continuously growing. In the USAa 300% increase of HF in 15 years has been related withthe improving of cardiology knowledge and care, and to theaging of the population. It has been estimated that in 2020HF and atrial fibrillation will represent the most importanthospitalization causes. Furthermore it has been estimated thatin Italy 10% of the over 65 population is affected by HF[5]. The difficulty of management of HF is not only relatedto its diffusion, but also to the typical clinical course of thedisease characterized by frequent and cyclic haemodynamicinstability related to age, co-morbidity, cognitive decay,therapy effectiveness, ambient and social factors.Literature reports the clear and effective advantage ofspecific patient monitoring well tuned to the patients clinicalcondition. Reported results of a metanalysis on 5000 patientsby Alister [6] point out that specific assistance programs likea phone hotline to medical staff (for example a nurse) orby outpatient medical follow-up assure both a reduction ofre-hospitalization and also mortality.In daily practice, management of patients with HF relays onrelatively simple physical signs such as the ones mentionedabove plus body weight which however need to be carefullymonitored in order to optimize the efficiency of treatment, toprevent the relapse of HF and the need of hospitalization.A user wearable breathing recorder was already describedin [7]. There the acquired data are transmitted to a nearbyinstalled computer. Young [8] presents an oximeter withwireless data transfer. Both approaches may be useful forclinical applications, but for continuous monitoring of patientsoutside the clinical environment, a wearable and miniaturizedrecording system is necessary. Previously our team hasdeveloped a GSM based data logger, which was able torecord breathing rate and SO2 and transmit relevant data tothe clinical server [9].

Technical solutions enabling the field of Telemedicinepromise to mediate the impact of changing population statis-tics. Most important is the field of on-line monitoring andanalysis of vital parameters. Different kinds of wireless tech-nologies promise to ensure patient compliance. EspeciallyBody Area Networks (BAN) coupled with these wireless tech-nologies [10] allow the setup of a comprehensive telemedicalinfrastructure. Depending on the environment two differentsubsystems are central to a BAN:

1) An embedded system as a sensor platform.2) A Wireless Sensor Network (WSN) tailored to the

specific task. In this case the BAN is usually called aWireless BAN or WBAN.

With respect to WBANs ZigBee/IEEE 802.15.4 is widelyused in WSN. 2008 a standardization of WBANs near to thehuman body has been started [11], [12]. The IEEE 802.15Task Group 6 (BAN) is developing the communicationstandard optimized for low power devices and operation on,in or near the human body [13].Smartphones and Tablet PCs take over tasks of the traditionalPC, once again changing the hardware base of computerscience. Therefore it is a natural development to usesmartphones as mobile sinks for WSNs (e.g. [14], [15]), andintegrate them as central modules of a telecare system (cf.[16], [17]). In addition to a (W)BAN a typical smartphonefeatures around a dozen internal sensors, some of them maybe used in a medical application (e.g. [18]).

Telemedical Systems are by definition Safety CriticalSystems (SCS), whose development is governed by nationaland international safety standards (e.g. IEC 60601 basedstandards). In addition to functional safety concerns securityin the (W)BAN (cf. [19], [20], [21], [22]), between the(W)BAN and the smartphone (e.g. via Bluetooth [23]), itsoperating system [24] and a wider infrastructure is a seriousissue.

The main idea for the current project is to develop anAndroid OS based data collection platform, that can collectphysiological data from multiple sensors, perform signal pro-cessing and analyses, store data in an internal memory andtransmit data via a UMTS connection to a clinical server.This paper describes two BAN approaches: 1: The WBANarchitecture is based on the ZigBee/IEEE 802.15.4 stack. 2:The BAN architecture uses an Atmel board as sensor platform.Both BAN designs are connected via Bluetooth to an Androidsmartphone, which features apps for analysis and visualizationof vital parameters. Example sensors foreseen in the firstprototype include sensors for O2 saturation of the blood, ECGand heart rate variability. In addition smartphone sensors (e.g.accelerometers) are used to gather additional information, e.g.fall detection. All information may be transferred to a medicaldatabase for further distribution and analysis.

II. DESIGN NO. 1: WBAN

A. ZigBee/ IEEE 802.15.4

Sensor nodes usually do not require high bandwidth fortypical applications (hundreds of bits/s). For a WBAN designsome requirements on the communication should be imposed:Resistance to interference, adequate throughput and security[25]. Therefore communication protocols such as Bluetooth,ZigBee, and IEEE 802.15.4 are serious candidates for the im-plementation of a WSN. ZigBee is a standard communicationprotocol for low-cost, low-power, wireless sensor and control

34

networks [26]. Just as Bluetooth, ZigBee operates on the 2.4GHz radio frequency but with a maximum data rate of 0.25Mbps [27].

Fig. 1. OSI vs. ZigBee/IEEE 802.15.4 Stack

In addition the ZigBee Alliance group has developed ZigBeeHealth Care to be used by medical and non-medical devicesin order to establish a universal communication standard in ahealth care environment. A major objective is to allow an indi-vidual to perform tasks that otherwise would be very difficultdue to a disability or medical condition. To accomplish thisZigBee Health Care fully supports standards like ISO/IEEE11073 [26].

B. WBAN Architecture

Figure 2 shows the WBAN architecture. In this architecturethe primary data processing is done by the sensor nodes,including the physiological signal processing in the micro-controller of the nodes. The secondary data processing is per-formed in the smartphone. This includes data representation,data filtering, graphical interface and data synchronization.Finally the last and most demanding data processing togetherwith the database management is performed in the medicalserver. The medical server allows local and remote access formedical personnel via the Internet.Our design encompasses communication protocols like theZigBee/IEEE 802.15.4 stack for intercommunication withinthe WSN, Bluetooth (Serial communication via RFCOMM) tolink the WSN with the smartphone, and WiFi or UMTS com-munication between the smartphone and the medical server.

Our mote design uses the Atmel ZigBitTM 2.4ATZB-24-A2/B0 [28], which features an ATmega1281V C,AT86RF230 Transceiver, 16 GPIOs, I2C Bus, 222kHz datarate [29], 1-wire interface 15.4kBits 125kBits [30], 4 ADCs10 Bit.

Wireless data transmission results in a higher patient com-pliance because there are no uncomfortable wires. On theother hand may the fear of electromagnetic radiation lower theacceptance. This served as a motivation for the IEEE 802.15.6

Fig. 2. WBAN Architecture

working group, with the goal of minimizing transmissionpower, range and SAR [13]. Future WBAN architectures willuse the new standard.

III. DESIGN NO. 2: BAN WITH ATMEL BOARD

In terms of the sensor nodes and the gateway, Atmel solu-tions promise to provide an ideal platform for Telecare devices,reliable communication together with power efficiency, in acompact design [31]. For this a high-performance, low power8-bit Atmel Microcontroller is available. It features a 8-bitAVR CPU, a maximum operating frequency of 32 MHz,128 KB In-System Self-Programmable Flash, 8 KB internalSRAM, 2048 bytes EEPROM, 2 Universal AsynchronousReceiver Transmitters (UART), 1 Serial Peripheral Interface(SPI) and 16 channels ADC with a resolution of 12 bits and2000 kSps speed, 4 channels DAC with a resolution of 12 bits.

IV. WIRELESS PLETHYSMOGRAPH/ PULSE OXIMETERMODULE

A. Pulse Oximetry Overview

Due to restricted space only the pulse oximetry modulewill be described. The ECG hardware was inherited from aprevious project and can be found in [32]. Pulse oximetry isan optical method to measure oxygen saturation, heart rateand heart rate variability. It is based on light absorption.Oxygenated Hemoglobin (Hb) has different light absorptionspectra than deoxygenated Hb. Thus the wavelength of redlight (660nm) is compared with the wavelength of infraredlight (940nm). The red/infrared light is received by a photodetector, which converts light into current. The output currentof the photo detector is proportional to the light intensity ofeach light source. The light sources are switched on alternately.Each source is switched on for a certain period and the currentis measured and converted into a voltage. The output signal

35

Fig. 3. BAN Architecture

lies between 1 Hz and 2 Hz [33] and consists of a small ACcomponent ( 1V) and a large DC component ( 10mV peak-to-peak) [34]. The AC component is caused by the arterialpulse. The DC component is caused by scattered light, residualarterial blood, venous blood and bloodless tissues. The SpO2level is calculated by taking the Root Mean Square (RMS)of the red/infrared AC values, computing the ratio betweenthe red and infrared RMS values and applying the ratio valueinto a 3rd order polynomial fit of the calibration curve foroximeters [35].

B. Design Outline

In order to tailor the system to specific application needsthe puloximeter application report SLAA274A from TexasInstruments (TI) [34] was used as basis and adjusted. Insteadof using the TI MSP430FG437 microcontroller, an AtmelATXmega128A1 is used with the following advantages:

Direct Memory Access Controller Fast 12 bit programmable (amplifying) ADC, which can

read 8 ADC channels in parallel (up to 2 MSps sampleand conversion rate)

Event system Interrupt handling with configurable priorities 8-bit RISC architecture with max. frequency of 32Mhz 128 kBytes Flash, 8 kBytes SRAM, 2 kBytes EEPROM

In addition to the microcontroller the hardware includes ananalog circuit, consisting of a LED drive circuit, photo diodeoutput processing and a ChipOx MiniMed fingerclip sensor.The firmware of the device is developed in C (GNU C com-piler WinAVR 20100110) and AVR Studio 4.18. The firmwareof the module consists of four major parts: LED control,signal amplification, LED data handling and processing anddata transfer. Data is transferred via serial interface. The DMAcontroller of the ATXmega128A1 is used to allow high speed

data transfer with minimal CPU intervention. In this way, thewhole data transfer can be processed in 200s CPU load.

V. ANDROID SMARTPHONE

A. Android Smartphone

An Android based smartphone has been chosen becauseof its powerful and Java-based development kit, AndroidSDK, its excellent documentation and library including classeslike BluetoothHealth, and the possibility to develop on manyplatforms, like Linux, Mac Os and Windows [36]. For devel-opment different smartphones are being used with Android2.3.5 and 4.03.

B. Android Apps

As mentioned in the System Architecture the smartphoneshould manage not only data acquisition from the W(BAN),but also synchronization and provide a Graphical UserInterface (GUI), among other tasks. In order to do soan Android application is necessary, this applicationshould feature several functions, among these are: Dataacquisition from the (W)BAN via Bluetooth; data analysis,i.e. comparison with medical norm values; GUI forconfiguration, data visualization, and communication;data transfer (synchronization) to a medical server via WiFior cellular network.Android applications are divided into Activity classes. AnActivity is both a unit of user interaction, and a unit ofexecution which provide reusable, interchangeable parts ofthe flow of UI components across Android Applications [37].

In essence the application is responsible to detect theBluetooth gateway and establish a full duplex communication,including device discovery, pairing, debugging andcommunication, and to be able to connect to the medical serverthrough the Internet enabling data synchronization betweenthe server and the W(BAN) in soft real-time. The applicationfeatures numerical analysis and graphical representationof the captured physiological data, an activity for thepatients profile, physical condition, disease history, etc., andactivities for the connection with the medical server. For thisproject some interesting packages available in the AndroidSDK are [36]: Android.Bluetooth, Android.database.sqlite,Android.net, Android.webkit, javax.net.ssl, Android SDK tool[17]. The use of third party libraries is optional keeping thevalidation effort for a Safety Critical System in mind.

The app is also responsible to present a GUI, whose designrepresents the captured data in an understandable way.Togetherwith the basic requirement of a state-of-the-art Android app,the GUI has therefore three principal modes: Configuration,display for patients, display for medical personnel with re-stricted access.

For the proposed application internal smartphone sensors,e.g. accelerometer, GPS, etc., provide additional opportunities,i.e. patient localization and possible detection of a fall. Basedon the evaluation of the acquired data the app starts

36

Fig. 4. Android Prototype Application

communication to predefined first responders.

Android based on Linux lacks a real-time kernel and cannotsupport hard real-time requirements. Based on the measure-ments of Mongia and Madisetti [38], tests of the complete,layered system are being conducted.

VI. SECURITY ISSUES

Security in Wireless Sensor Networks is of major concern(cf. [19], [22]). The security issues in a WBAN differ inseveral aspects from those in other applications. In a WBANfor medical applications there are only a few motes, as in ourdesign. Nevertheless a typical problem is to distinguish anattack from a network failure. Therefore detection mechanismsare required. Guided by the OSI model layering-based attacks[22] on the ZigBee/IEEE 802.15.4 stack (cf. fig. 1) may becountered by the combined ZigBee/IEEE 802.15.4 securitymodel (cf. [19]). Currently evaluations are under way toassess the permissible overhead introduced. In a BAN basedon sensors attached to an embedded control system securityissues are much simpler.

Both approaches communicate with an Android smartphonevia Bluetooth. Because of the widespread use of Bluetoothin mobile phone communication numerous threats exist

(cf. [39]). Threat mitigation involves properly designedcommunication between the (W)BAN and the smartphone,e.g. a defined software engineering process, use of theencryption and authentication mode, comprehensive testing.

Android is one of the most popular operating system forsmartphones. Despite this fact the Android security model isin need of further enhancements. Shabtai et al. [40] identifiedthe major threats to the operating system. The Android soft-ware stack is built on a Linux kernel. The Android securityframework is based on the Linux security system, the cellularnetwork provider and Android specific security mechanisms,i.e. application permissions, component encapsulation andsigning applications [40]. Shabtai et al. propose enhancementsto the Android system to increase protection.

VII. RESULTS

First reliability test runs with a WBAN prototype with ascalable number of ZigBee motes resulted in a 4-6.5 % packetloss (s. Table 1). In the first and third setup data were sentfrom a ZigBee endpoint via a router to the coordinator. In thesecond setup the endpoint sent data directly to the coordinator.

Run-Time Payload Packets sent Packets lost Security[Days] [Bytes]10 10 1.26 105 5 % none10 10 1.26 105 4 % standard10 10 1.26 105 6.5 % standard

TABLE IZIGBEE/IEEE 802.15.4 RELIABILITY TESTS

Trade-offs are necessary with respect to battery life,data rate and ZigBee/IEEE 802.15.4 security settings.Unfortunately increasing security does not only introducesa secure way to send and receive data, but also introducesan impact on the performance, and battery life of the sensornodes and the phone running the android application. Withstandard security only 54.54% of the usable payload isavailable. Battery life was reduced by 14.28% in a nodefunctioning as a coordinator. In order to mitigate the data lossre-transmitting information in slow sampling rate nodes canbe used, and interpolation techniques for nodes with fastersampling rates. Higher level encryption resulted in a drasticreduction of the data rate.

Range is not an issue for medical WSN with the used motes.Range tests revealed sufficient range for the application. A typ-ical setup for research in environmental physiology comprisesthree motes. ECG and Plethysmograph/Pulse Oximeter datahave been acquired, transferred and displayed in an AndroidApp. Two different smartphones, a HTC Desire HD withAndroid 2.3.5 and a Samsung Nexus S with Android 4.03,have been used to collect and visualize the WBAN data. Inaddition the Android App features the input of user/patientdata and the UMTS/WIFI data transfer to a medical serverdatabase.

37

VIII. CONCLUSION AND FUTURE WORK

The first design approach, a WBAN, fulfills the basicrequirements. Reliability and range are sufficient. Due to fearswith respect to transmission power of wireless systems theupcoming standard IEEE 802.15.6 will be considered for fu-ture designs. The combination of the WBAN with an Androidsmartphone offers a large functionality. Vital parameters canbe stored, analyzed and visualized with GUIs designed for theend-user. Security on all levels of the layered system must befurther investigated, especially to define trade-offs with respectto performance and comfortable use. Certification accordingto medical safety standards is currently impossible due to thedifferent components used, e.g. the Android operating system.The first version of the proposed system will therefore be usedin different research applications of environmental physiology,i.e. HRV measurement, heart rate, breathing rate, etc.

REFERENCES

[1] Cheung, S.S., Advanced Environmental Exercise Physiology, HumanKinetics, 2010, ISBN-10: 0-7360-7468-6

[2] Joyner, MJ, Exercise hyperemia: waiting for the reductionists?, Am. J.Physiol. Heart Circ. Physiol. 2006 Sep;291(3):H1032-3

[3] Kuch, B.; Koss, B.; Butazzo, G.; Dujic, Z.; Sieber, A., A novel wearableapnea dive computer for continuous plethysmographic monitoring ofoxygen saturation and heart rate, Diving and Hyperbaric MedicineVolume 40, No. 1, March 2010, p 34 -39

[4] Sieber, A.; LAbbate, A; Kuch, B.; Wagner, M.; Benassi, A.; Passera,M.; Bedini, R., Advanced instrumentation for research in diving andhyperbaric medicine. Undersea and Hyperbaric Medicine. 2010;37(5)

[5] Fabbri, G.; M. Gorini, M.; Maggioni, A.P.; Di Lenarda, A., Temporaltrends in pharmacological therapy in the IN-CHF registry from 1995 to2005, Area Scompenso ANMCO, Firenze, G Ital Cardiol (Rome), Feb.2007

[6] Alister, F.A.; Stewart, S.; Ferrua, S.; McMurray, J.J., Multidisciplinarystrategies for the management of heart failure patients at high risk foradmission: a systematic review of randomized trials. J. Am. Coll. Cardiol.2004 Aug 18;44(4):810-9.

[7] Jovanov, E.; Raskovic, D.; Price, J.; Krishnamurthy, A.; Chapman, J.;Moore, A.; Patient Monitoring Using Personal Area Networks of WirelessIntelligent Sensors, in Proc. of 38th Annual Rocky Mountain Bioengineer-ing Symposium, RMBS 2001, Copper Mountain Conference Center, COApr. 20-22, 2001.

[8] Young, A., Wireless Oximeter, School of Information Technology andElectrical Engineering, University of Queensland, 10/2003

[9] Sieber, A.; Yong, X.; LAbbate, A.; Woegerer, C.; Bedini, R., CardiacSentinel: a smart GSM based embedded solution for continuous remotemonitoring of cardiac patients, WISES 2008, Regensburg, Germany,2008, ISBN: 978-3-00-024989-1

[10] Bingchuan, Y. and Herbert, J., Web-based real-time remote monitor-ing for pervasive healthcare,Pervasive Computing and CommunicationsWorkshops (PERCOM Workshops), 2011 IEEE International Conferenceon, p. 625 -629

[11] Kohno, R.; Hamaguchi, K.; Huan-Bang Li; Takizawa, K. , R&D andstandardization of body area network (BAN) for medical healthcare,Ultra-Wideband, 2008. ICUWB 2008. IEEE International Conference on, vol.3, pp.5-8, 10-12 Sept. 2008

[12] Huan-Bang Li; Takizawa, K.; Kohno, R. , Trends and standardization ofbody area network (BAN) for medical healthcare, Wireless Technology,2008. EuWiT 2008. European Conference on , pp.1-4, 27-28 Oct. 2008

[13] http://www.ieee802.org/15/pub/TG6.html (last visited March 2nd, 2012)[14] Jinfeng Zhang; Canfeng Chen; Jian Ma; Nengqiang He; Yong Ren,

uSink: Smartphone-based mobile sink for wireless sensor networks,Consumer Communications and Networking Conference (CCNC), 2011IEEE, pp.90-95, 9-12 Jan. 2011

[15] Moreira, N.; Venda, M.; Silva, C.; Marcelino, L.; Pereira, A., @Sensor- Mobile application to monitor a WSN, Information Systems andTechnologies (CISTI), 2011 6th Iberian Conference on, pp.1-6, 15-18June 2011

[16] Oresko, J.J.; Zhanpeng Jin; Jun Cheng; Shimeng Huang; Yuwen Sun;Duschl, H.; Cheng, A.C., A Wearable Smartphone-Based Platform forReal-Time Cardiovascular Disease Detection Via Electrocardiogram Pro-cessing, Information Technology in Biomedicine, IEEE Transactions on, vol.14, no.3, pp.734-740, May 2010

[17] Postolache, O.; Girao, P.S.; Ribeiro, M.; Guerra, M.; Pincho, J.; Santi-ago, F.; Pena, A., Enabling telecare assessment with pervasive sensingand Android OS smartphone, Medical Measurements and ApplicationsProceedings (MeMeA), 2011 IEEE International Workshop on,pp.288-293, 30-31 May 2011

[18] Grimaldi, D.; Kurylyak, Y.; Lamonaca, F.; Nastro, A., Photoplethys-mography detection by smartphones videocamera, Intelligent Data Ac-quisition and Advanced Computing Systems (IDAACS), 2011 IEEE 6thInternational Conference on , vol.1, pp.488-491, 15-17 Sept. 2011

[19] Healy, M.; Newe, T.; Lewis, E., Security for wireless sensor networks: Areview, Sensors Applications Symposium, 2009. SAS 2009. IEEE, pp.80-85, 17-19 Feb. 2009

[20] Yan-Xiao Li; Lian Qin; Qian Liang, Research on Wireless SensorNetwork Security, Computational Intelligence and Security (CIS), 2010International Conference on , pp.493-496, 11-14 Dec. 2010

[21] Burgner, D.E.; Wahsheh, L.A., Security of Wireless Sensor Networks,Information Technology: New Generations (ITNG), 2011 Eighth Interna-tional Conference on , pp.315-320, 11-13 April 2011

[22] Modares, H.; Salleh, R.; Moravejosharieh, A., Overview of SecurityIssues in Wireless Sensor Networks, Computational Intelligence, Mod-elling and Simulation (CIMSiM), 2011 Third International Conferenceon , pp.308-311, 20-22 Sept. 2011

[23] Tan, M.; Masagca, K.A., An Investigation of Bluetooth Security Threats,Information Science and Applications (ICISA), 2011 International Con-ference on , pp.1-7, 26-29 April 2011

[24] Bing, Han, Analysis and Research of System Security Based on Android,Intelligent Computation Technology and Automation (ICICTA), 2012Fifth International Conference on , pp.581-584, 12-14 Jan. 2012

[25] Chandrashekhar Dethe; Digambar Wakde; Chandrakant Jaybhaye, Blue-tooth based Sensor Networks Issues and Techniques Proceedings ofthe First Asia International Conference on Modelling and Simulation(AMS07), 2007 IEEE

[26] ZigBee Alliance, http://www.zigbee.org/About/AboutTechnology/ZigBeeTechnology.aspx (last visited March 2nd, 2012)

[27] http://sensor-networks.org/index.php?page=0823123150 (last visitedMarch 2nd, 2012)

[28] Atmel Products, http://www.atmel.com/dyn/resources/prod documents,Atmel Corporation, (last visited March 6th, 2012)

[29] I2C-bus specification and user manual, http://www.nxp.com/documents/user manual/UM10204.pdf, NXP Semiconductors, (last visited March6th, 2012)

[30] 1-Wire Tutorial, http://www.maxim-ic.com/products/1-wire/flash/overview/index.cfm, Maxim Integrated Products, Inc., (last visited March6th, 2012)

[31] http://www.atmel.com/applications/BuildingAutomation/telecare/default.aspx (last visited March 7th, 2012)

[32] Kuch, B.; Bedini, R.; LAbbate, A.; Wagner, M.; Buttazzo, G.; Sieber,A., Embedded Data Logging Platform for Research in Diving Physiology.Lecture Notes in Electrical Engineering, Vol. 81, Chapter 4, 2011, ISBN:978-94-007-0637-8

[33] Bailey, J., Wireless pulseoximeter: A Major Qualifying Project Report.Bachelor Thesis. Faculty of the Worcester Polytechnic Institute. 2008.

[34] Chan, V.; Underwood, S., SLAA274A: A Single-Chip PulsoximeterDesign Using the MSP430. Texas Instruments Application Report. 2005.

[35] Mendelson, Y., Pulse Oximetry: Theory and Applications for Noninva-sive Monitoring. Clinical Chemistry. 1992; 38(9): pp. 1601-1607.

[36] http://developer.android.com/index.html (last visited March 6th, 2012)[37] Zigrud Medniek, Laird Dornin, G. Blake Meike, Masumi Nakamura;

Programming Android: Java Programming for the New Generation ofMobile Devices & OReilly,

[38] Mongia, B. and Madisetti, V., Reliable Real-Time Applications onAndroid Operating System, http://users.ece.gatech.edu/vkm/, last visitedMarch 6th 2012

[39] Dunning, J.P.; Taming the Blue Beast: A Survey of Bluetooth BasedThreats, Security & Privacy, IEEE , vol.8, no.2, pp.20-27, March-April2010 b

[40] Shabtai, A.; Fledel, Y.; Kanonov, U.; Elovici, Y.; Dolev, S.; Glezer,C., Google Android: A Comprehensive Security Assessment, Security &Privacy, IEEE , vol.8, no.2, pp.35-44, March-April 2010

38