research article a framework for the estimation and validation of energy consumption ... · 2019....

Research ArticleA Framework for the Estimation and Validation ofEnergy Consumption in Wireless Sensor Networks

Alexandros Karagiannis1 and Demosthenes Vouyioukas2

1School of Electrical and Computer Engineering National Technical University of Athens 15780 Athens Greece2Department of Information and Communication Systems Engineering University of the Aegean 83200 Samos Greece

Correspondence should be addressed to Alexandros Karagiannis akaragmobilentuagr

Received 6 April 2015 Revised 26 May 2015 Accepted 1 June 2015

Academic Editor Eduard Llobet

Copyright copy 2015 A Karagiannis and D Vouyioukas This is an open access article distributed under the Creative CommonsAttribution License which permits unrestricted use distribution and reproduction in any medium provided the original work isproperly cited

Body sensor networks and implantable and ingestible medical devices energy efficiency is a substantial key factor in networklifetime and functionalityThis work confronts the nodesrsquo energy problem by establishing a unified energy consumption frameworkcomprised of theoretical model energy simulator model and electronic metering modules that can be attached to the nodes Atheoretical analysis a simulation procedure and the design and development of three prototype electronic metering modules arepresented in this paperWe discuss the accuracy of the proposed techniques towards a unified framework for the a priori estimationof the energy consumption in commercial sensor nodes taking into account the application functionality and the energy propertiesof the incorporated electronics Moreover body network nodes are considered for the application and the measurements of theproposed framework

1 Introduction

Wireless sensor networks (WSNs) found their way in thepast years to both research fields and the commercial sectorThe application of WSNs in body physiological parametersmonitoring set a specific category termed as body sensornetworks (BSNs) [1]The latest development is in implantableand ingestible medical devices (IIMDs) that form in-bodyand on-body network

Dynamic and highly customized body sensor nodesaddress the clinical needs for individualized health moni-toring and delivery especially in several vulnerable groupsChronic ill patients and episodic incidents require pervasivehealthcare services utilizing BSN technology to accuratelyaccess medical related information and improve the qualityof services provided Specifically episodic events in patientsand the role of BSN in addressing them change the way ofhealthcare services delivered

Usual configuration of implantable and ingestiblemedicaldevices and BSN nodes follow the pattern of microsensors inclose proximity or contact for monitoring of biosignals that

feed a processing subsystem and a radio communication sub-system Implantable ingestible and BSN nodes are equippedwith energymanagement subsystems that incorporate batter-ies and optionally energy harvesting circuitry

The energy management subsystem is considered a crit-ical component of IIMDs and BSNs with a role in theoptimization of available energy reservoirs usage Severalproposals have been made concerning the powering ofimplantable devices and BSN nodes that include transcuta-neous external sources for battery recharge electromagneticenergy transmission piezoelectric power generation ther-moelectric devices ultrasonic powermotors radio frequencyrecharging optical recharging methods biofuel cells andlithium batteries [2]

The design of applications and protocols for BSNsand IIMDs requires that energy consumption optimizationshould be addressed in a multiparameters fashion thatinvolves the energy-efficient operation of nodesrsquo subsystems

BSNrsquos energy consumption issue has been highlightedin research literature as a key factor for performance andnetwork functionality There is an increasing number of

Hindawi Publishing CorporationJournal of SensorsVolume 2015 Article ID 124987 13 pageshttpdxdoiorg1011552015124987

2 Journal of Sensors

published works dealing with various aspects of energy con-sumption issue mainly in simulation level energy-efficientprotocol design and low power electronics developmentHowever a unified approach to combine theoreticalmodelingof WSN and BSN applications simulation estimation andelectronic metering module design and development for theenergy consumption is the main contribution of this work

In this paper a technical approach is discussed towardsthe design and development of a framework for the esti-mation and measurement of energy consumption in BSNand IIMD nodes For this purpose three electronic meteringmodules have been designed and developed for the validationof theoretical and software based estimations of energy con-sumption in BSN nodes Furthermore TinyOS applicationshave been developed for the implementation of sensor nodeshealth monitoring applications Energy-efficient design ofTinyOS applications is considered a primary target in energyconsumption minimization with an impact on the operationof individual subsystems the nodesrsquo and networkrsquos lifetime ofBSN and IIMDs battery-powered devices

The structure of this paper is as follows Section 2 intro-duces and analyses the related work in this area In Section 3basic assumptions are exhibited for the development of aframework for energy consumption study on BSN and IIMDnodes In Section 4 the methodology is presented regardingthe theoretical procedure for energy consumption estima-tion a simulator-based on PowerTOSSIM is developed andthree electronic metering modules are designed and devel-oped for energy consumption monitoring and measurementIn Section 5 results are provided for the energy consumptionof BSN nodes which monitor biosignals and finally theconclusions are discussed in Section 6

2 Related Work on Energy Consumption Issue

BSN and IIMD scientific progress is developing in two frontsthe research oriented and the commercial one Body wornsensing devices with computing and networking capabilitiesare presented in MIThril project [3] CodeBlue [4] has devel-oped medical sensor network using various transducers ofvital signs based on commercially available WSN nodes suchas TelosB and MicaZ [5] designs Healthy Aims [6] projecthas been focused on specific sensor applications for hearingvision glaucoma and intracranial pressure CardioMEMS [7]is one of the first attempts to develop an implantable pressuresensor into an aneurysm sac Various healthcare applicationsinvolving body sensor networks are presented in [8]

Typical BSN node consists of four main subsystems (i) asensing subsystem (ii) a microcontroller (MCU) micropro-cessor subsystem constituting the processing component ofthe node (iii) a radio communication subsystem providingthe wireless communication and (iv) an energymanagementsubsystem According to the specific requirements posed bythe applications additional componentsmay be incorporatedsuch as a secondary wireless communication module Thisis the case in IIMDs that operate two radio communicationsubsystems one for transcutaneous data communication andone for wake-up functionality [9]

In the case of a commercial implementation of medicalimplantable RF transceiver manufactured by Zarlink [9]the wake-up radio subsystem is an extremely low poweroperating in Industrial Scientific and Medical (ISM 24ndash25 GHz) band to provide wake-up receiver option whilst thedata communication radio subsystem operates in MedicalDevice Radio Communications Service (MedRadio) fre-quency bands 401ndash406MHz [10] The dual band operationin the MedRadiondashISM bands has an impact on the energyefficiency of the implantable medical device with the specifictransceiver

Improvement in electronics design and development interms of energy efficiency has resulted in ultralow powerimplementations in all four basic subsystems of BSN nodesand IIMDS However real life application experiences of BSNshow that incorporation of energy-efficient electronics onboard a BSN node partially guarantees low power perfor-mance unless energymanagement is taken into consideration[11]

A common assumption in BSN and IIMDs considers thatthe main consumer of energy is the radio communicationsubsystemThis stands true if certain conditions are met suchas the energy consumption of sensory subsystem being lowerdue to the incorporation of low power sensors or the imple-mentation incorporates an energy-efficient microprocessingunit Previousworks have revealed that one bit of informationtransmitted through the radio communication subsystem isequivalent to the execution of a few thousand instructions interms of energy consumption [12]

Radio communication subsystem has various states ofoperation regarding the energy consumption point of viewIn receiving mode regardless of the protocol or standardsupported the energy consumption is almost equivalent totransmitting mode However in idle mode the energy con-sumption of the radio communication subsystem is reducedand especially in themode of sleep it is an order ofmagnitudeless exhibiting energy-efficient power management of BSNnodersquos subsystems

There are several testbeds of WSNs running applicationsfor energy consumption and protocols performance evalua-tion Motelab [13] consists of a set of permanently deployednodes connected to a central server Each node is monitoredin terms of energy consumption by a network-connectedmultimeter Trio [14] is another large-scale testbed used foroutdoor purposes that is supplied by solar panels SenseNet[15] consists of low cost sensor nodes facilitating multipleusers to interact with the network

Research is mainly focused on energy-efficient protocolsimplemented on the network however there are efforts forapplicationsrsquo energy signature formation and optimization ofsubsystems operation with a direction towards minimizingenergy consumption There is an extensive variety of proto-cols that deal with the energy consumption issue by differentperspectives and suggest interesting and highly efficient waysto conserve energy A thorough presentation and a high leveltaxonomy of these protocols and techniques are presented ina recently published survey of Anastasi et al [12]

The incorporation of high-cost high accuracy infrastruc-ture in order to measure the energy consumption on single

Journal of Sensors 3

BSN nodes or IIMDs is prohibitive and inflexible [16] Onthe other hand on purpose designed hardware modulespartially deal with the energy monitoring on the node[17] Testbed deployment approaches oriented to softwarebased and hardware supported schemes to measure energyconsumption face this issue in a holistic way [18]

Energy consumption is measured by means of speciallydesigned hardware modules attached to BSN nodes [17] Theconcept for software based estimation methods is that totalenergy consumption is equal to the summation of individ-ual subsystem energy consumptions and the energy statetransitions PowerTOSSIM [19] is a representative softwarebased method which relied on TinyOS [20] A drawbackof this energy simulator is related to the intrinsic energymodel which is nodersquos technology dependent AEON [21]as an energy simulator is implemented on top of AVRORA[22] emulator Moreover there are some simulation toolsand platforms such as OPENET NS2 SHAWN SensorSimEmStar OMNet++ and GloMoSim which are widespread inresearch WSN community for their capacity to investigateenergy-efficient protocols performance and simulate theenergy-oriented operation of WSN nodes and networks

Taking into account processing sensing communicationlogging network formation and actuation events an energyestimation model is developed in [23] Lifetime estimationof WSNs is a critical parameter closely related to nodersquosand networkrsquos energy consumption which are investigated in[24] In [25] an OMNET++ based simulation estimated theenergy consumption based on the widely used RF transceiverCC2420 [26] found in numerous commercial WSN nodes

Specifically in the field of IIMDs energy consumptionestimation and accurate measurement are critical in definingthe medical device lifetime Merli et al [27] suggest thatenergy signature of IIMDs subsystems plays key role indefining the most energy hungry subsystems In their imple-mentation an implantable medical device that monitorstemperature equipped with Zarlink RF transceiver [9] andDigital Signal Processing unit the energy consumed in radiocommunication subsystem is the 92 of the total energyconsumed on the IIMD

In [28] the study of WSN node subsystemsrsquo energyconsumption in different energy states and state transitionsthe authors present the energy models of the node corecomponents including processors RF modules and sensorsApproximate energy models are developed in [29 30] whichare proved to be sound with battery-powered devices [31]The advantages of approximate energy models include smallerror compared to high accuracy energy consumption mea-surements on WSN and BSN nodes and low complexity insimulation level

BSN and IIMDs subsystem optimization of operation is akey stage in the minimization of energy consumption Thereis a rich field of innovations towards this direction concerningthe antenna component of the radio communication sub-system The design takes into account several performanceparameters for the development of novel miniature antennasfor integration in IIMDs operating in the MedRadio (401ndash406MHz) and ISM (4331ndash4348 8680ndash8686 and 9028ndash9280MHz) bands [32] One special category of antennas

for IIMDs is patch antennas which are receiving significantscientific interest for integration into the implantablemedicaldevices and radio frequency- (RF-) enabled biotelemetry[33]

In a previous authorsrsquo work a theoretical tool for energyconsumption prediction is presented A metric of energyconsumed on the BSN node is proposed in [34 35] thatconsists of a hybrid method based on current consump-tion subsystem specifications and real or simulated timinginformation for each BSN subsystem The battery voltagelevels are sampled by ADC of MCU in the commercial WSNnode used Battery voltage current consumption and timinginformation provide the energy consumption in node andsubsystem level

The contribution of the current work with respect toprevious authorsrsquo work [34 35] lies in the unification ofthe theoretical modeling the design and development of anenergy consumption simulator for specific TinyOS applica-tions and the presentation of a theoretical and simulationprocess for energy consumptionmeteringmodules which aredesigned and developed Specifically in the last part of thiswork we present the parameters involved in the design anddevelopment process a set of measurements conducted for avariety of TinyOS applications the optimization process forthe minimization of measurement error in electronic meter-ing modules and a rich set of new data concerning TinyOSapplications the simulation results measurement data andcomparison of each part of the frameworkwith respect to oneof the electronic metering modules The current work sup-ports the idea that the energy consumption of WSNs-BSNsshould be studied under the consideration of the whole sys-tem of ldquoWSN-BSN node and the running applicationrdquo Thisconcept reveals two fronts of energy consumption optimiza-tion the low power electronics for node implementation andthe applicationsrsquo energy-efficient operational optimization

3 Assumptions Adopted for the EnergyConsumption Framework

Body sensor network is deployed in standard laboratoryconditions and monitors temperature as a typical case ofa biosignal for on-body and in-body medical applicationsTemperature sensors are attached to the commercially avail-able sensor node Tmote Sky a variant of TelosB which is usedfor network deployment Multiple sampling frequencies areused to investigate the effect of sampling on the total energyconsumption of the node Processing of the acquired timeseries is performed in the microcontroller subsystem of BSNnode

A single hop body sensor network is taken into consider-ation with three nodes communicating with a central pointBSN node that collects data from the network Communica-tion is accomplished in a single hop fashion by broadcast RFmessages produced at the radio communication subsystem ofeach node that implements IEEE 802154 standard The RFenvironment is considered to include only path loss

Focus is given on the total energy consumption of eachnode that executes TinyOS applications specially developed


for BSN temperature monitoring and processing The energymodel of the node incorporated in the energy simulator takesinto account current consumption provided by subsystemsrsquomanufacturers and verified by averaged current consumptionmeasurement at each possible energy stateThe energymodelconsidered is an approximate model as described in theliterature review section that neglects energy transition statesdue to the small time window and the low energy dissipatedin each transition

Transitions between different energy states that occurredduring the operation of nodersquos subsystems are considered asan additive factor in total energy consumption The fractionof energy consumed during a transition over total energyconsumption due to subsystem operation is low Howevertheoretical tool does not take into consideration the energyconsumed during transitions This assumption is necessaryto avoid complexity in estimating energy consumption bythe theoretical approach although it is a reason for deviationof actual energy consumption from the estimated one Thisassumption is reasonable as it is suggested in numerous ref-erences and literature review for approximate energy models

The three prototype electronic metering modules aredesigned and developed in order to be used in a modularfashion over any commercial WSN and BSN node Basicrequirements to calculate energy are the a priori knowledgeof the time function of voltage induced across the BSN nodethe time function of current drawn and timing informationfor the operation of the node

In this work the voltage is provided by the Analog-to-Digital (ADC) component of the microcontroller subsys-tem According to specifications of MSP430 microcontroller(MCU) used in the implementation of Tmote Sky the ADCinternal part of the MCU may be used to read and monitorthe battery voltage over time Time samples of battery voltageconstitute a series of voltage readings over time that providea time function of the battery voltage induced across the BSNnode

The main role of the three electronic metering modulesis to provide the function of the current drawn by theBSN node and subsystems over time Timing information isprovided and used by various sources such as the prototypemodulesrsquo output oscilloscope measurements the energy sim-ulator virtual timing machine and the internal ADC of themicrocontroller subsystem

A metric for total and subsystemsrsquo energy consumptionis adopted formulated in EV which is expressed in mA lowastsec The nominator represents the energy consumption andthe denominator is the time invariant voltage which is thereference point in this case If the time window of theobservation concerning the battery voltage is reasonablysmall the voltage is considered constant over time andthus can be used as reference point in order to incorporateinformation from current drawn and timing However themethodology in the development of the proposed frameworkenables the incorporation of battery voltage time functioncurrent consumption and timing information for the extrac-tion of energy consumption estimation and measurementvalidation

4 Proposed Methodology

41 Developed TinyOS Applications TinyOS is an opensource operating system which supports nesC and is widelyused inWSN commercial and research oriented applicationsNesC is based on wiring components together that provideinterfaces for communication purposes with other compo-nents

Since the radio communication subsystem of BSN andIIMDs is considered the most energy expensive one anassumption which in most cases stands correct TinyOSprovides a feature called Low Power Listening (LPL) in orderto reduce energy consumption during the radio commu-nication operation According to LPL technique the radiocommunication subsystem is turned on specific time periodsand checks the presence of a carrier In case of a radiomessagethe subsystem remains active to receive the RF message

A family of TinyOS applications are developed in orderto be used as reference software applications in the theo-retical calculation of the energy consumption the softwarebased estimation and the validation of energy consumptionmeasurements via the three electronic metering modulesThe basic part of the TinyOS applications called Oscilloscoperesembles the functionality of the oscilloscope device whichis to sample the biosignal selected temperature and comparethe values of the acquired data against a predefined thresholdWhen a number of temperature readings (NREADINGSas a parameter in the application) are acquired the radiocommunication subsystem is enabled to listen for carriertransmit the acquired data and then turn off the radio in aLPL mode

Furthermore data fusion is implemented as a secondaryway apart from LPL to minimize radio transmissions andapply on board data processing by activating the microcon-troller subsystem (OscilloscopeFusion) The combination ofLPL and data fusion targets the further minimization ofenergy consumption Every NREADINGS of temperature anaveraging takes place in data fusion section and the acquiredvalue is stored until 10 averaged values are available whichcorrespond to 10lowastNREADINGS temperature samples Radiocommunication subsystem is turning on and transmits theRF message with the 10 averaged values in one RF message(OscilloscopeWMR With Management Radio)

The selection of the processing algorithm is related to thenature of the sampled biosignal the sampling frequency andthe complexity of the implementation in the microcontrollersubsystem Oscilloscope based family TinyOS applicationshave a strong flexibility in the implementation of on boardprocessing algorithms taking into account the processingpower as a resource oriented constraint

Another branch of TinyOS applications used in thiswork based on the Oscilloscope family is the CountToLeds-CountToRadio applications A major difference of CountTofamily applications with Oscilloscope family applicationsis the use of microcontroller subsystem counter to countbetween intervals instead of acquiring data from the sensorysubsystem The communication and processing stages arecommon in the two TinyOS application families


42 Theoretical Tool for Energy Consumption CalculationTheoretical calculation of energy consumption for TinyOSfamily applications executed in BSN nodes is based on ahybrid approach Theoretical model combines a mathemati-cal formula of EV for summing individual subsystem energyconsumption with timing information derived either fromsimulation stage or from measurement procedures Currentconsumption information for each subsystem and energystate is derived from manufacturerrsquos specification Howeverit is verified in measurement stage by the three prototypeelectronic modules which are used for current monitoringof each subsystem

Timing information is a critical element of the proposedtheoretical procedure that provides the time period for theoperation of each subsystem in specific energy states Thetime variable is dependent on the running application onboard the BSN node If timing information from energysimulator output is incorporated a deviation from actualtiming is expected due to the assumptions imposed in thedevelopment of simulator approximate energy model andthe indirect way that timing information is extracted formicrocontroller subsystem

Battery voltage information is provided by theADCof themicrocontroller subsystem however in a short time periodvoltage level is expected to be constant and time invariant

The equation of the theoretical energy model for BSNnode energy consumption is expressed as

119864

119881

= 119868MCU119879MCU + 119868119871119909119879119871119909 + 119868119877119909119879119877119909 + 119868119879119909119879119879119909 +sum119894

119868119894119879119894 (1)

where 119868MCU is the current drawn by the microcontroller sub-system for time period 119879MCU 119868119871119909 is the current drawn by theradio communication subsystemat the state of listening to thechannel for time period 119879

119871119909 119868119877119909

is the current drawn by theradio communication subsystem at the state of receiving datafor time period 119879

119877119909 and 119868

119879119909is the current drawn by the radio

communication subsystem at the state of transmitting datafor time period 119879

119879119909 The sum term is a cumulative model of

all the other subsystemsrsquo energy consumptions whichmainlyrefer to the energy consumption of the sensory subsystemand the flash memory operation For short time periodobservation energy states transition of microcontroller andradio communication subsystems are considered negligiblein the approximate energy model

The terms of (1) are analyzed in detail as a sum ofenergy consumptions which correspond to various energystates of each subsystem The theoretical model can beadapted according to the specific operation of each TinyOSapplication implemented on BSN nodesThe first step for thetheoretical calculation of BSN nodesrsquo energy consumption isthe analysis of the operation in the subsystem level and theformulation of the appropriate mathematical expression

We provide hereinafter an example of this procedurefor a TinyOS application Initially the radio subsystem isin sleep energy state and the microcontroller is in energystate mode LPM3 (the microcontroller of the BSN nodebased on Tmote Sky has 5 different energy states forthe microcontroller operation denoted as LPM0ndashLPM4)

Current consumption during LPM3 is 10 uA according tomanufacturer specifications Microcontroller ADC starts thesampling of temperature sensor Every time a new samplingtakes place the content of data register is forwarded to radiocommunication subsystem for transmission via broadcastRF message In between the readings the microcontrollerand radio communication subsystem get into the sleep stateWhen new data reading is received the radio subsystemwakes up and the energy state changes

The operation of the Oscilloscope based TinyOS applica-tion leads to the formulation of the mathematical equationfor the theoretical estimation of energy consumption on BSNnode depicted in119864

119881

= 119879MCU-Idle119868MCU-Idle

+119879Radio-Activesum119894

119868Radio +119879MCU-Activesum119895

119868MCU

+119879Sensor119868Sensor +119879RF-Transition119868RF-Transition

+119879MCU-Transition119868MCU-Transition

(2)

where 119868MCU-Idle is the current consumption of the micro-controller at sleep state time period 119879MCU-Idle sum119894 119868Radio isthe total current consumption of the radio communicationsubsystem to listening transmitting or receiving mode forthe time period of 119879Radio-Active sum119895 119868MCU is the total currentconsumption of the microcontroller subsystem that transitsfrom sleep state to LPM3 for time period 119879MCU-Active 119868Sensoris the current consumption of the sensory subsystem for timeperiod119879Sensor 119868MCU-Transition is the current consumption of theenergy states transition for the time period 119879MCU-Transitionand 119868MCU-Transition is the current consumption of the MCUsubsystem for the energy states transition from sleep modeto LPM3

A formulation of the specific TinyOS theoretical anal-ysis may reasonably ignore the 119868sensor 119868RF-Transition and119868MCU-Trasnsition terms as long as these values are small enoughcompared to the other current consumption terms (an orderofmagnitudewould be an appropriate criterion) and does notcontribute significantly to the 119864119881 result

43 Simulator-Based Estimation of BSN Energy ConsumptionA variant of PowerTOSSIM called PowerTOSSIM-z is usedfor software-based estimation of energy consumption inTinyOS applications running on BSN nodes The adaptationof the energy simulator is critical for the energy consumptionestimation in order to fit the needs of the available BSN nodearchitecture which is based on Tmote sky

Recalibration of the simulator is taking place in twofronts First an approximate energy model is required thatincludes the necessary virtual subsystems which are the exactcorresponding of the actual BSN nodersquos ones In the nextstep a calibration of the new approximate energymodel takesplace The microcontroller subsystem imposes a challengefor the development of a sufficient energy model and asuitable virtual event machine This challenge is addressedby mapping the basic blocks of TinyOS application codeinto cycles of operation for each application In this way an


Table 1 Calibration of energy simulator model

Subsystem Current consumption (mA)LED 0 (red) 330LED 1 (yellow) 560LED 2 (blue) 353Radio receiving +microcontroller (MCU) 1995

Radio transmission +microcontroller (MCU) 1810

Radio listening +microcontroller (MCU) 1886

indirect estimation of microcontroller energy consumptionis extracted for the software-based total energy consumptioncaused by the execution of the application on BSN node

Energy simulator monitors the energy state for eachvirtual subsystem by outputting appropriate messages ofenergy state transitions as the code runs on the virtualmachine of the simulator At the end of the simulationa log file is produced with virtual subsystems messagesof the energy state transitions along with virtual timinginformation Energy consumption information is acquired byprocessing the log file for the various energy states of the BSNsubsystems as well as the virtual timing information

Recalibration of the approximate energy model is accom-plished by a measurement campaign using electronic meter-ing modules developed for energy consumption monitoringon BSN nodes The outcome of this process is depicted inTable 1 which includes current consumption measurementresults of BSN node for the parameterization of simulatorenergy model

Next step the adaptation of the architecture and thedevelopment of a modified energy model for the simulatorproceeds with the verification by uploading the TinyOSOscilloscope based applications The essential characteristicof these applications is related to the a priori identificationof specific operations and energy states that take placeduring the execution of each application An indicativeexample is depicted in Figure 1 where the application Oscil-loscopeWMR samples the temperature and transmits it tothe BSN At the time periods between successive samplingsradio subsystems energy state transits to sleep mode thus theenergy consumption is reduced

Oscilloscope family TinyOS applications developed in thecontext of current research work have also been presentedin a previous work [34 35] along with energy consumptionoptimization techniques Results depicted in Figure 1 for theverification of simulator energymodel are cross-checkedwiththeoretical model results The approximate energy modeltheoretical expression of OscilloscopeWMR operation fortime period 100m sec and number of samples transmitted (1sample) is

119864

119881

= 119879MCU-Idle119868MCU-Idle +119879Active (sum119868Radio +sum119868MCU)

= 0160mA sec(3)

25

20

15

10

5

001 02 03 04 05 06

Time (s)

Curr

ent c

onsu

mpt

ion

(mA

)

OscilloscopeWMR

Figure 1 Verification of energy simulator operation for speciallydesigned TinyOS application (OscilloscopeWMR) used for temper-ature monitoring in BSN node In 119910-axis the current consumptionis expressed in mA and in 119909-axis the time is in seconds

The energy consumption estimation ofOscilloscopeWMR from energy simulator is 0146mA secper time period There is an underestimation of 9 whichis due to the assumptions posed in energy simulatorapproximate energy model concerning the production ofvirtual machine energy states and timing of MCU subsystemand radio communication subsystem

44 Electronic Metering Modules for the Measurement of BSNEnergy Consumption Methodologies for hardware usage inenergy consumption measurement in BSN and IIMDs haveattracted small interest in research community Most of theimplementations propose high-cost solutions unsuitable forlarge-scale networks

In this work three prototype electronic metering mod-ules are proposed for energy consumption monitoring Prin-cipal goal of the design is the low cost solution with highaccuracy and modularity Suitable dynamic range in currentconsumption monitoring is considered as a key requirementsince current consumption in BSN nodes may vary upto three orders of magnitude according to the operationsimposed by the TinyOS applications (from a fewuA to severalthousands of uA)The electronic metering modules are smallsize with low power consumption

441 Shunt Resistor Based Electronic Metering Module Themain approach behind the first BSN energy consumptionelectronic module is based on a shunt resistor that intervenesbetween the BSN battery-powered node and the energyreservoir in this case the batteries It monitors voltage timefunction across the shunt resistor High accuracy shuntresistor is employed in order to acquire current consumptiontime function In this case shunt resistor should meet twocontradictory conditions The first one imposes that shuntresistor value should be high enough in order to detect thevoltage across it and the second one sets the shunt resistor at


RGain

RLoad

RShunt

VminusV+

V13V

0

1

1

2

2

334

5

45

6

6

78

7

+ minusAD620AN

IMOTE

U1

Figure 2 Electronicmeteringmodule of the energy consumption inBSNnodeswith shunt resistor and instrumentation amplifierModelIMOTE is used for the modeling of BSN node

0

5

10

15

20

25

100 200 300 400 500 600 700 800 900 1000

Erro

r (

)

IMOTE (120583A)

Figure 3 Deviation of theoretically expected output value againstthe simulated one in the range of hundreds of uA for the estimationof BSN energy consumption before the compensation of the instru-mentation amplifier offset voltage The error reaches 17 in 100 uArange In 119910-axis the error or deviation as it is calculated in (4) and in119909-axis the current is drawn by the BSN node as a parametric value

an ultralow value in order not to disturb the powering of theBSN node

At a constant battery voltage for short time periodsthe current information provided by the metering moduleenables the calculation of total energy consumption of theBSN node as a product of measured by the ADC battery-voltage current consumption and timing information

Simulation of the electronic module by means of a typicalelectronic circuit simulator provides information about thesimulated performance of the module The values for theelements used in the design of the electronic module in thesimulation stage are 119877shunt = 1Ω 119877Gain = 300Ω 119877Load =10KΩ Figure 2 depicts the actual design of the first electronicmetering module

100 200 300 400 500 600 700 800 900 1000

07

06

05

04

03

02

01

00

Erro

r (

)

IMOTE (120583A)

Figure 4 Deviation of theoretically expected output value againstthe simulated one in the range of hundreds of uA for the estimationof BSN energy consumption after the compensation of the instru-mentation amplifier offset voltageThe error reaches 06 in 100 uArange In 119910-axis the error or deviation as it is calculated in (4) and in119909-axis the current is drawn by the BSN node as a parametric value

Deviation of the simulated output from the theoreticallycalculated output is computed by

100(119881out minus Gain119881119877shunt)

Gain119881119877shunt

(4)

where 119881out is the theoretically expected value Gain is depen-dent on the selection of 119877Gain resistor and is related tothe specifications of instrumentation amplifier AD620 and119881119877shunt is the voltage output by the simulation processThemain source of deviation in small current range (uA)

is due to the offset voltage of the instrumentation amplifierFigures 3 and 4 depict the deviation in the range of hundredsof uA (typical current consumption range in BSNs) beforeand after the compensation of offset voltage in electronicmetering module based on shunt resistor Limiting offsetvoltage results in a deviation of less than 06 for smallcurrent range (uA) and less than 005 in the range of mA

In the development stage validation of the correct oper-ation according to the design specifications for accuracy isaccomplished by replacing the BSN node with a set of highprecision resistors in the range of 100Ωndash500KΩ Results ofthe validation process are depicted in Figure 5 Deviationaccording to (4) in the order of 20 at current range ofless than or equal to 10 uA has a negligible impact on totalenergy consumption since the excess energy consumption isminimal in that small current range For typical case currentconsumptions in BSN nodes in the range of hundreds of uAthe deviation is less that 5 reaching 04 at 100 uA

442 Capacitor-Based Electronic Metering Module Regard-ing the second electronic metering module a capacitorinitially charges with a current drawn submultiple of thecurrent drawn by the BSN node When the capacitor voltageis equal to an upper voltage threshold discharge process isinitiated until capacitor voltage drops to the lower voltagethreshold In every cycle of chargedischarge a pulse ofconstant energy is produced Pulse duration is proportionalto the current drawn by BSN node


25

20

15

10

5

0

minus5

minus10

minus15

minus20

Dev

iatio

n (

)

1 10 100 1000 10000 100000

Compensated current (120583A)

Figure 5 Deviation of simulated from theoretically expected outputas a function of BSN nodersquos compensated current consumption inlog scale

Expressing the operation of the electronic meteringmodule based on successive cycles of capacitor charge anddischarge in a mathematical formula yields (5) for theschematic of Figure 6

119868Node =119877scale119877sense

119862Δ119881

Δ119879

997904rArr

119881supply119868Node = 119881supply119877scale119877sense

119862Δ119881

Δ119879

997904rArr

119881supply119868NodeΔ119879 = 119881supply119877scale119877sense119862Δ119881 997904rArr

119881supply119868NodeΔ119879 = 119864ramp

(5)

where 119877scale and 119877sense are resistances of the current mirrorwhich also includes an operational amplifier and a MOSFETin order to achieve equal voltages across the 119877scale and 119877senseIn that case the current charging the capacitor is proportionalto the current drawn by the BSN node

The constant energy transfer 119864ramp per cycle ofchargedischarge is dependent on the capacitance theupper and lower voltage thresholds and the resistors 119877senseand 119877scale used in the implementation of the electronicmetering module Hence the total energy consumptionmeasurement is reduced to a number of pulses counting

Averaging in time domain provides an intuitive math-ematical equation (6) to calculate BSN nodersquos current con-sumption bymeasuring frequency of pulses produced at eachcycle of capacitor chargedischarge (Figure 7)

119868Node = 120572119891 (6)

where 119868NodeΔ119879 is the averaged BSN nodersquos current consump-tion 119891 is the mean frequency of pulse train and 120572 is aconstant coefficient equal to (119877scale119877sense)119862Δ119881

The values of the elements used in the simulation of theelectronic module are 119877

1= 50KΩ 119877

2= 35 KΩ 119877

3= 1KΩ

119877sense =1Ω 119877scale = 100Ω and 119862 = 100 nF

Thedeviation between the simulated and the theoreticallyexpected output after the procedure of offset voltage com-pensation is in the range of 10 for current consumptionvalues less than 20 uA In the range of current consumptionsof hundreds of uA to a few mA the deviation is less than 2

The development of the proposed electronic meteringmodule after the refinement in simulation stage proceedswith the verification of correct operation which is similar tothe procedure described in the first electronic module

443 iCount-Based Electronic Metering Module The iCounttechnique relies on the capacity of Pulse Frequency Modu-lated counting cycles of a node switching regulator [36] Theamount of energy transferred in each cycle of switching isconstant for constant input voltageThus a linear relationshipbetween BSN current and frequency is shown The range oflinearity spread from 5 uA to 35mA for current consumptionand constant input voltages in the range of 2-3V Theschematic of the implemented iCount technique is depictedin Figure 8

For inductance 119871 = 45 uH and capacitances 119862in =119862out = 10 uF the dynamic range of switching frequenciesis maximized for the aforementioned current range Lowervalues of inductance or capacitances result in the preservationof linearity for smaller range of switching frequencies

Simulation results of the electronic metering moduleimplementing the iCount technique are depicted in Figure 9for a wide range of input voltage from 16V to 309V

There is a disturbance of linearity for BSN currentconsumption in the range of 25mA for input voltage below19 V Particularly in BSNbattery-powerednodes the levels ofbattery voltage affect the linearity of the electronic meteringmodule even in the range of 20mA for input voltages of 16 V

Measurement of the switching frequency enables thedirect calculation of BSN current consumption according tomodeling of iCount technique as it is depicted in Figure 10

5 Results

TinyOS applications developed for the purpose of biosignalmonitoring in BSN nodes are employed for validation of theenergy consumption estimation and measurement resultsA family of TinyOS applications Oscilloscope is used fortemperature sensing

In Table 2 current consumption is presented for a seriesof Oscilloscope based TinyOS-2x applications measured viathe electronic metering module which is based on shuntresistor A common basis is considered concerning the timeperiod in order to provide comparative results

The performance of TinyOS-2x applications in termsof overall current consumption is dependent on the designdirections and the implementation of the applications Theincorporation of optimization techniques in order to min-imize energy consumption targets the reduction of energyexpensive subsystems usage such as the radio communica-tion subsystem The combination of this technique with lowpower listening modes in TinyOS applications has a directimpact on the energy efficiency of BSN nodes A typical case


U1AAD8552AR3

2 4

8

1

0Q2A

SI9933DY

3

0

U2B

TLC552C

GND7

2DIS13 2OUT92RST10

14

2THR12

2CON112TRI8

7

U2A

TLC552C

GND7

1DIS1 1OUT51RST4

14

1THR2

1CON31TRI6

0

0 6

0

5

0

1 9

4

0

2

10

0

8

IMOTE 1mA

100Ω

RScale

RLoad

RShunt

Vsupply1 Vsupply2

500kΩR1_TRIMMER

1Ω

33V

33V100nF

100nF

100nF100nF

CLoad

C3C2

C1

100 kΩ

10kΩ

10kΩ

R2_TRIMMER

R3_TRIMMER

+

minus

VCCVCC

Figure 6 Electronic metering module of the energy consumption in BSN nodes with a capacitor Model IMOTE is used for the modeling ofBSN node

10000

1000

100

10

1

01

1 10 100 1000 10000 100000

BSN current (120583A)

Pulse

freq

uenc

y (H

z)

Figure 7 Averaged current consumption of BSN in relation to thepulse frequency measured 119910-axis and 119909-axis are in log scale

is depicted in Figure 11 for two TinyOS applications basedon Oscilloscope which samples a biosignal (temperature)periodically and enables the radio communication subsystemonly when a specific number of temperature readings arecollected and processed

Implementation of low power techniques and time seriesprocessing methods verifies that energy-efficient approachesapplied onBSNnodes have positive impact in terms of energyconsumption Furthermore the effect of sampling frequencyis explored on total energy consumption As the samplingfrequency increases sensory subsystem energy consumptiontends to increase Moreover frequent sampling producesmore data on the node which raises energy burden of radiocommunication subsystem The processing methods appliedon collected data time series reduce the amount of data to betransmitted thus the total energy consumption

Figure 12 refers to the energy consumption comparativeresults between the theoretical model the energy simulatorand the electronic metering modules The biosignal tested is

Vin

Vin

Vout

Cin

Cout

GND

IMOTE

SW

LTC3525-33+minus

L

Switch_regulator

SHDN

Figure 8 Electronic metering module based on iCount techniqueThe model IMOTE is used for the modeling of BSN node

temperaturewith a small sampling rate Table 3 includes com-parative results of energy consumption on various TinyOSapplications executed on BSN nodes The measurementresults of the shunt resistor electronic metering circuit areused as a reference point for comparison purposes

The comparison between the theoretical model theenergy simulator output and the electronic metering mod-ulesmeasurements showsmatching results in terms of energyconsumption as depicted in Figure 12 The overall pictureshows similar results of themeasurement the simulation andthe theoretical methods however in software-based currentconsumption estimation it tends to be underestimated dueto the approximate energymodel and the timing informationwhich is produced from the virtual machine behind theenergy simulator

The relative deviation between the capacitor based mod-ule the shunt resistor module the iCount based module



Freq

uenc

y (H

z)

100000

10000

1000

100

10

11 10 100 1000 10000 100000

BSN current (mA)

Freq

uenc

y (k

Hz)

100

90

80

70

60

50

40

30

20

10

05 10 15 20 25 30 35

Vin = 309VVin = 3VVin = 27VVin = 24V

Vin = 21VVin = 18VVin = 15V



Figure 9 BSN current consumption against frequency switchingcurve for a wide range of input voltages

5

45

4

35

3

25

2

15

1

05

00 05 1 15 2 25 3 35 4 45 5

Log switching frequency

Log

BSN

curr

ent

y = 10008x minus 01101

R2 = 1

Figure 10 Modeling the linear relationship of BSN current drawnagainst the switching frequency for iCount technique The 119910-axisand 119909-axis are in log scale

25

20

15

10

5

0

minus515 2 25 3

Time (s)

Time (s)

Curr

ent c

onsu

mpt

ion

(mA

)

25

20

15

10

5

0

Curr

ent c

onsu

mpt

ion

(mA

)

OscilloscopeWMR

15 16 17 18 19 20 21 22 23 24 25

OscilloscopeFusionWM

Figure 11 Current consumption of TinyOS-2x applications Oscil-loscopeWMR and OscilloscopeFusionWMR based on the output ofthe electronic metering module for energy consumption measure-ment

045 05 055 06 065 07 075 08 085

5

10

15

20

25

Time (s)

Curr

ent c

onsu

mpt

ion

(mA

)

Oscilloscope based TinyOS application comparative results

MeasurementEnergy simulatorTheoretical model

Figure 12 BSN nodersquos comparative energy consumption results fortemperature monitoring using theoretical model energy simulatoroutput and electronic metering module measurement (shunt resis-tor)


Table 2 TinyOS-2x applications current consumption measure-ment results

TinyOS-2x applications Current consumption (mAlowastsec)per time-period 9767 sec

Oscilloscope(NREADINGS = 1) 186653

OscilloscopeWMR(NREADINGS = 1) 19304

Oscilloscope with LPL(NREADINGS = 10)(Wake-up period = 20msec)

74444


42223


OscilloscopeFusion with LPL(Wake-up period = 20msec) 68617


OscilloscopeFusionWMR 0635

the energy simulator and the theoretical model is in therange of less than 10 in reference to the shunt resistorelectronic metering module Only in one case the relativedeviation is in the range of 227 underestimating the totalenergy consumption from the energy simulator Indirectmicrocontroller energy consumption estimation and energystate transitions which are not taken into considerationexplain the difference of the energy simulator result from theshunt resistor electronic metering measurement

One interesting point from the comparison results ofTable 3 is the reduction of differences in all techniques whenthe application implements fewer energy consumption opti-mization procedures Focusing on the approximate energysimulator and the theoretical estimation tool fewer energyconsumption optimization procedures have a positive effecton more accurate timing information extracted from thevirtual machine of the simulator On the other hand whenenergy consumption optimization procedures are appliedsuch as low power listening with or without managementof radio communication subsystem the timing informationextracted seems to underestimate the energy state transitionswith an impact on the accuracy of the energy consumptionresult

The accuracy of the electronic metering modules isconsidered fairly acceptable Even in low energy consumptionTinyOS applications with optimization techniques despitethe fact that simulation results from the electronic meteringmodules design stage revealed increased deviation of themeasured output from the theoretically expected value of cur-rent consumption the three modules operate with credibilityin the region of their low sensitivity which corresponds tocurrent consumptions in the order of uA In higher currentconsumptions the accuracy in monitoring is increased dueto the high sensitivity in the region of current consumptionin the order of 10 uA to thousands of uA

The concept in the use of electronic metering modules inthe current work is the a priori validation of theoretical mod-eling and energy simulation results with energy consump-tion measurements The energy reservoir for the electronicmetering modules differs from the energy reservoir of theBSN-WSN node A coarse current consumption comparisonbetween the three components of the corresponding elec-tronicmeteringmodule reveals thatAD620 quiescent currentis 09mA in typical cases whilst for LTC3525 is only 7 uAand for the second electronic metering module is expectedto be higher than 1mA In terms of individual currentconsumption of the three electronic metering modules thesecond one which is based on the iCount technique seems tobe less energy consumable whilst maintaining sensitivity andmeasurement accuracy for a wide range of BSN-WSN inputvoltages

6 Discussion

Energy resourcesmanagement subsystem is a critical compo-nent in BSNnodes and IIMDs since reduced lifetime and lackof adequate power may compromise network functionalityand reliability Simulation and theoretical tools are essentialin designing energy-efficient applications for body sensornetworks Measurement results verify the available tool pro-viding a valuable insight into the actual energy consumptionon BSN nodes

The network topology is simple for the presented evalu-ation methodology as we are interested at the energy con-sumption framework of health monitoring body networks

Electronicmeteringmodules provide accurate time func-tion of current drawn by the BSN or IIMD node This infor-mation is combined with the sampling of the battery voltagelevels and timing information to produce an accurate energyconsumption signature of the BSN node system or the IIMDsystem The energy consumption issue in BSN and IIMDsintroduces a holistic approach in determining the energyprofiling which is composed of ultralow power electronicsenergy-efficient applications and means to determine theenergy consumed in BSN and IIMD nodes

A unified framework with components of energy con-sumption theoretical model simulation of approximateenergy models and electronic metering modules facilitatethe study of the energy consumption issue on body sensornetworks In this work an energy consumption estimationframework is presented which is composed of theoreticaltool calculation simulation of applications energy signatureand three prototype electronicmeteringmodules for accurateenergy consumption measurement

Conclusions drawn out of this framework allow thedesign of energy efficient hardware and software solutionsfor BSN nodes and the performance testing of the proposedsolution in terms of total energy consumption Furthermoreexperimental measurements of BSN nodersquos energy consump-tion by means of electronic metering modules attachedto the nodes verify in real time the efficiency of energyconservation techniques in order to extend node and networklifetime In the case of BSN and IIMDs applied to clinical


Table 3 Comparative results of BSN node energy consumption

ApplicationsTechniques

Current consumption per period (mAlowastsec)Shunt resistor Capacitor based iCount based Energy simulator Theoretical estimation

CountToLeds 4821 4811 4760 4863 4866mdash minus02 13 09 09

CountToRadio 1843 1872 1881 1842 1844mdash 16 21 minus01 01

CountToRadio with LPL 1371 1403 1399 1247 1321mdash 23 2 minus9 minus37

CountToRadioWMR 0154 0165 0160 0119 0147mdash 71 39 minus227 minus45

practice network and node lifetime extension is related topatient comfort and minimization of invasive procedures fordepleted energy reservoir replacement

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

References

[1] G Z Yang Ed Body Sensor Networks Springer 2006[2] X Wei and J Liu ldquoPower sources and electrical recharging

strategies for implantable medical devicesrdquo Frontiers of Energyand Power Engineering in China vol 2 no 1 pp 1ndash13 2008

[3] MIThril Project Home Page Massachussets Institue of Tech-nology httpwwwmediamiteduwearablesmithril

[4] D Malan T Fulford-Jones M Welsh and S Moulton ldquoCode-blue an ad hoc sensor network infrastructure for emergencymedical carerdquo in Proceedings of the 1st International Workshopon Wearable and Implantable Body Sensor Networks pp 55ndash58London UK 2004

[5] MEMSIC Solutions httpwwwmemsiccomproductswire-less-sensor-networkswireless-moduleshtml

[6] Healthy Aims EU Framework VI project httpwwwhealthy-aimsorg

[7] M G Allen ldquoImplantable micromachined wireless pressuresensors approach and clinical demonstrationrdquo in Proceedingsof the 2nd International Workshop onWearable and ImplantableBody Sensor Networks pp 40ndash43 2005

[8] D Vouyioukas and A Karagiannis ldquoPervasive homecare mon-itoring technologies and applicationsrdquo in Telemedicine Tech-niques and Applications G Graschew and S Rakowsky EdsInTech 2011

[9] Zarlink Semiconductors Integrated Circuit ZL70102 httpwwwzarlinkcom

[10] Medical Device Radiocommunications Service (MedRadio)ldquoFederal Communication Commission (FCC) Std CFR Part95601ndash673 Subpart E Part 951201ndash1221 Subpart I 2009formerly Medical Implanted Communication System (MICS)rdquohttpswwwfccgovencyclopediamedical-device-radiocom-munications-service-medradio

[11] C Kompis and P Sureka Eds Power Management Technologiesto Enable Remote and Wireless Sensing 2010 httpwwwlibeliumcomlibelium-downloadslibelium-ktn-power man-agementpdf

[12] G Anastasi M Conti M di Francesco and A PassarellaldquoEnergy conservation in wireless sensor networks a surveyrdquoAdHoc Networks vol 7 no 3 pp 537ndash568 2009

[13] G Werner-Allen P Swieskowski and M Welsh ldquoMoteLaba wireless sensor network testbedrdquo in Proceedings of the 4thInternational Symposium on Information Processing in SensorNetworks (IPSN rsquo05) pp 483ndash488 Los Angeles Calif USAApril 2005

[14] P Dutta J Hui J Jeong et al ldquoTrio enabling sustainableand scalable outdoor wireless sensor network deploymentsrdquo inProceedings of the 5th International Conference on InformationProcessing in SensorNetworks (IPSN rsquo06) pp 407ndash415NashvilleTenn USA April 2006

[15] T Dimitriou J Kolokouris and N Zarokostas ldquoSensenet awireless sensor network testbedrdquo in Proceedings of the 10th ACMSymposium on Modeling Analysis and Simulation of Wirelessand Mobile Systems (MSWiM rsquo07) pp 143ndash150 October 2007

[16] C B Margi V Petkov K Obraczka and R Manduchi ldquoChar-acterizing energy consumption in a visual sensor networktestbedrdquo in Proceedings of the 2nd International Conference onTestbeds and Research Infrastructure for the Development ofNetworks and Communities pp 331ndash339 Barcelona Spain July2006

[17] X Jiang P Dutta D Culler and I Stoica ldquoMicro powermeter for energy monitoring of wireless sensor networks atscalerdquo in Proceedings of the 6th International Symposium onInformation Processing in Sensor Networks (IPSN rsquo07) pp 186ndash195 Cambridge Mass USA April 2007

[18] J-P Sheu C-C Chang andW-S Yang ldquoA distributed wirelesssensor network testbed with energy consumption estimationrdquoInternational Journal of Ad Hoc and Ubiquitous Computing vol6 no 2 pp 63ndash74 2010

[19] V Shnayder M Hempstead B Chen G W Allen and MWelsh ldquoSimulating the power consumption of large-scale sen-sor network applicationsrdquo in Proceedings of the 2nd Interna-tional Conference on Embedded Networked Sensor Systems (Sen-Sys rsquo04) pp 188ndash200 ACM Baltimore Md USA November2004

[20] TinyOS httpwwwtinyosnet


[21] O Landsiedel K Wehrle and S Gotz ldquoAccurate prediction ofpower consumption in sensor networksrdquo in Proceedings of the2nd IEEE Workshop on Embedded Networked Sensors pp 37ndash44 May 2005

[22] B L Titzer D K Lee and J Palsberg ldquoAvrora scalable sensornetwork simulation with precise timingrdquo in Proceedings ofthe 4th International Symposium on Information Processing inSensor Networks (IPSN rsquo05) pp 477ndash482 Los Angeles CalifUSA April 2005

[23] M N Halgamuge M Zukerman and K RamamohanaraoldquoAn estimation of sensor energy consumptionrdquo Progress inElectromagnetics Research B vol 12 pp 259ndash295 2009

[24] Y Liu and W Zhang ldquoStatic worst-case energy and lifetimeestimation of wireless sensor networksrdquo Journal of ComputingScience and Engineering vol 4 no 2 2010

[25] O Khader and A Willig ldquoAn energy consumption analysis ofthe Wireless HART TDMA protocolrdquo Computer Communica-tions vol 36 no 7 pp 804ndash816 2013

[26] CC2420 24 GHz IEEE 802154ZigBee-ready RF Transceiverhttpwwwticomlitdssymlinkcc2420pdf

[27] F Merli L Bolomey F Gorostidi Y Barrandon E Meurvilleand A K Skrivervik ldquoIn vitro and in vivo operation of awireless body sensor noderdquo inWireless Mobile Communicationand Healthcare vol 83 of Lecture Notes of the Institute forComputer Sciences Social Informatics and TelecommunicationsEngineering pp 103ndash110 Springer Berlin Germany 2012

[28] H Y Zhou D Y Luo Y Gao and D C Zuo ldquoModeling of nodeenergy consumption for wireless sensor networksrdquo WirelessSensor Network vol 3 no 1 p 18 2011

[29] F Kerasiotis A Prayati C Antonopoulos C Koulamas and GPapadopoulos ldquoBattery lifetime prediction model for a WSNplatformrdquo in Proceedings of the 4th International Conference onSensor Technologies and Applications (SENSORCOMM rsquo10) pp525ndash530 IEEE Venice Italy July 2010

[30] T Yang Y K Toh and L Xie ldquoRun-time monitoring of energyconsumption in wireless sensor networksrdquo in Proceedings ofthe IEEE International Conference on Control and Automation(ICCA rsquo07) pp 1360ndash1365 IEEE June 2007

[31] S Abbate M Avvenuti D Cesarini and A Vecchio ldquoEstima-tion of energy consumption for tinyos 2 x-based applicationsrdquoProcedia Computer Science vol 10 pp 1166ndash1171 2012

[32] A Kiourti andK SNikita ldquoMiniature scalp-implantable anten-nas for telemetry in the MICS and ISM bands design safetyconsiderations and link budget analysisrdquo IEEE Transactions onAntennas and Propagation vol 60 no 8 pp 3568ndash3575 2012

[33] A Kiourti and K S Nikita ldquoA review of implantable patchantennas for biomedical telemetry challenges and solutionsrdquoIEEE Antennas and Propagation Magazine vol 54 no 3 pp210ndash228 2012

[34] A Karagiannis D Vouyioukas and P Constantinou ldquoEnergyconsumption measurement and analysis in wireless sensornetworks for biomedical applicationsrdquo in Proceedings of the4th ACM International Conference on PErvasive TechnologiesRelated to Assistive Environments (PETRA rsquo11) Crete GreeceMay 2011

[35] A Karagiannis S Kokkorikos and P Constantinou ldquoEnergyconsumption analysis and optimization techniques for wirelesssensor networksrdquo in Proceedings of the IEEE Symposium onComputers and Communications (ISCC rsquo09) pp 8ndash14 July 2009

[36] P Dutta M Feldmeier J Paradiso and D Culler ldquoEnergymetering for free augmenting switching regulators for real-timemonitoringrdquo in Proceedings of the 7th International Conferenceon Information Processing in Sensor Networks (IPSN rsquo08) pp283ndash294 IEEE St Louis Mo USA April 2008

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Active and Passive Electronic Components

Control Scienceand Engineering

Journal of



RotatingMachinery


Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design



Shock and Vibration


Civil EngineeringAdvances in

Acoustics and VibrationAdvances in



Electrical and Computer Engineering

Journal of

Advances inOptoElectronics


Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of


Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014


Chemical EngineeringInternational Journal of Antennas and

Propagation




Navigation and Observation



DistributedSensor Networks



published works dealing with various aspects of energy con-sumption issue mainly in simulation level energy-efficientprotocol design and low power electronics developmentHowever a unified approach to combine theoreticalmodelingof WSN and BSN applications simulation estimation andelectronic metering module design and development for theenergy consumption is the main contribution of this work

In this paper a technical approach is discussed towardsthe design and development of a framework for the esti-mation and measurement of energy consumption in BSNand IIMD nodes For this purpose three electronic meteringmodules have been designed and developed for the validationof theoretical and software based estimations of energy con-sumption in BSN nodes Furthermore TinyOS applicationshave been developed for the implementation of sensor nodeshealth monitoring applications Energy-efficient design ofTinyOS applications is considered a primary target in energyconsumption minimization with an impact on the operationof individual subsystems the nodesrsquo and networkrsquos lifetime ofBSN and IIMDs battery-powered devices

The structure of this paper is as follows Section 2 intro-duces and analyses the related work in this area In Section 3basic assumptions are exhibited for the development of aframework for energy consumption study on BSN and IIMDnodes In Section 4 the methodology is presented regardingthe theoretical procedure for energy consumption estima-tion a simulator-based on PowerTOSSIM is developed andthree electronic metering modules are designed and devel-oped for energy consumption monitoring and measurementIn Section 5 results are provided for the energy consumptionof BSN nodes which monitor biosignals and finally theconclusions are discussed in Section 6

2 Related Work on Energy Consumption Issue

BSN and IIMD scientific progress is developing in two frontsthe research oriented and the commercial one Body wornsensing devices with computing and networking capabilitiesare presented in MIThril project [3] CodeBlue [4] has devel-oped medical sensor network using various transducers ofvital signs based on commercially available WSN nodes suchas TelosB and MicaZ [5] designs Healthy Aims [6] projecthas been focused on specific sensor applications for hearingvision glaucoma and intracranial pressure CardioMEMS [7]is one of the first attempts to develop an implantable pressuresensor into an aneurysm sac Various healthcare applicationsinvolving body sensor networks are presented in [8]

Typical BSN node consists of four main subsystems (i) asensing subsystem (ii) a microcontroller (MCU) micropro-cessor subsystem constituting the processing component ofthe node (iii) a radio communication subsystem providingthe wireless communication and (iv) an energymanagementsubsystem According to the specific requirements posed bythe applications additional componentsmay be incorporatedsuch as a secondary wireless communication module Thisis the case in IIMDs that operate two radio communicationsubsystems one for transcutaneous data communication andone for wake-up functionality [9]

In the case of a commercial implementation of medicalimplantable RF transceiver manufactured by Zarlink [9]the wake-up radio subsystem is an extremely low poweroperating in Industrial Scientific and Medical (ISM 24ndash25 GHz) band to provide wake-up receiver option whilst thedata communication radio subsystem operates in MedicalDevice Radio Communications Service (MedRadio) fre-quency bands 401ndash406MHz [10] The dual band operationin the MedRadiondashISM bands has an impact on the energyefficiency of the implantable medical device with the specifictransceiver

Improvement in electronics design and development interms of energy efficiency has resulted in ultralow powerimplementations in all four basic subsystems of BSN nodesand IIMDS However real life application experiences of BSNshow that incorporation of energy-efficient electronics onboard a BSN node partially guarantees low power perfor-mance unless energymanagement is taken into consideration[11]

A common assumption in BSN and IIMDs considers thatthe main consumer of energy is the radio communicationsubsystemThis stands true if certain conditions are met suchas the energy consumption of sensory subsystem being lowerdue to the incorporation of low power sensors or the imple-mentation incorporates an energy-efficient microprocessingunit Previousworks have revealed that one bit of informationtransmitted through the radio communication subsystem isequivalent to the execution of a few thousand instructions interms of energy consumption [12]

Radio communication subsystem has various states ofoperation regarding the energy consumption point of viewIn receiving mode regardless of the protocol or standardsupported the energy consumption is almost equivalent totransmitting mode However in idle mode the energy con-sumption of the radio communication subsystem is reducedand especially in themode of sleep it is an order ofmagnitudeless exhibiting energy-efficient power management of BSNnodersquos subsystems

There are several testbeds of WSNs running applicationsfor energy consumption and protocols performance evalua-tion Motelab [13] consists of a set of permanently deployednodes connected to a central server Each node is monitoredin terms of energy consumption by a network-connectedmultimeter Trio [14] is another large-scale testbed used foroutdoor purposes that is supplied by solar panels SenseNet[15] consists of low cost sensor nodes facilitating multipleusers to interact with the network

Research is mainly focused on energy-efficient protocolsimplemented on the network however there are efforts forapplicationsrsquo energy signature formation and optimization ofsubsystems operation with a direction towards minimizingenergy consumption There is an extensive variety of proto-cols that deal with the energy consumption issue by differentperspectives and suggest interesting and highly efficient waysto conserve energy A thorough presentation and a high leveltaxonomy of these protocols and techniques are presented ina recently published survey of Anastasi et al [12]

The incorporation of high-cost high accuracy infrastruc-ture in order to measure the energy consumption on single


































119864

119881

= 119868MCU119879MCU + 119868119871119909119879119871119909 + 119868119877119909119879119877119909 + 119868119879119909119879119879119909 +sum119894

119868119894119879119894 (1)


119871119909 119868119877119909


119877119909 and 119868









119881




119868MCU



(2)















119864

119881


= 0160mA sec(3)

25

20

15

10

5

001 02 03 04 05 06

Time (s)

Curr

ent c

onsu

mpt

ion

(mA

)

OscilloscopeWMR







RGain

RLoad

RShunt

VminusV+

V13V

0

1

1

2

2

334

5

45

6

6

78

7

+ minusAD620AN

IMOTE

U1


0

5

10

15

20

25

100 200 300 400 500 600 700 800 900 1000

Erro

r (

)

IMOTE (120583A)





100 200 300 400 500 600 700 800 900 1000

07

06

05

04

03

02

01

00

Erro

r (

)

IMOTE (120583A)





(4)






25

20

15

10

5

0

minus5

minus10

minus15

minus20

Dev

iatio

n (

)

1 10 100 1000 10000 100000





119862Δ119881

Δ119879

997904rArr


119862Δ119881

Δ119879

997904rArr



(5)




119868Node = 120572119891 (6)



1= 50KΩ 119877

2= 35 KΩ 119877

3= 1KΩ









5 Results





U1AAD8552AR3

2 4

8

1

0Q2A

SI9933DY

3

0

U2B

TLC552C

GND7

2DIS13 2OUT92RST10

14

2THR12

2CON112TRI8

7

U2A

TLC552C

GND7

1DIS1 1OUT51RST4

14

1THR2

1CON31TRI6

0

0 6

0

5

0

1 9

4

0

2

10

0

8

IMOTE 1mA

100Ω

RScale

RLoad

RShunt

Vsupply1 Vsupply2

500kΩR1_TRIMMER

1Ω

33V

33V100nF

100nF

100nF100nF

CLoad

C3C2

C1

100 kΩ

10kΩ

10kΩ

R2_TRIMMER

R3_TRIMMER

+

minus

VCCVCC


10000

1000

100

10

1

01

1 10 100 1000 10000 100000


Pulse

freq

uenc

y (H

z)





Vin

Vin

Vout

Cin

Cout

GND

IMOTE

SW

LTC3525-33+minus

L

Switch_regulator

SHDN







Freq

uenc

y (H

z)

100000

10000

1000

100

10

11 10 100 1000 10000 100000

BSN current (mA)

Freq

uenc

y (k

Hz)

100

90

80

70

60

50

40

30

20

10

05 10 15 20 25 30 35






5

45

4

35

3

25

2

15

1

05

00 05 1 15 2 25 3 35 4 45 5


Log

BSN

curr

ent

y = 10008x minus 01101

R2 = 1


25

20

15

10

5

0

minus515 2 25 3

Time (s)

Time (s)

Curr

ent c

onsu

mpt

ion

(mA

)

25

20

15

10

5

0

Curr

ent c

onsu

mpt

ion

(mA

)

OscilloscopeWMR

15 16 17 18 19 20 21 22 23 24 25



045 05 055 06 065 07 075 08 085

5

10

15

20

25

Time (s)

Curr

ent c

onsu

mpt

ion

(mA

)










74444


42223









6 Discussion

















References








































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










































119864

119881

= 119868MCU119879MCU + 119868119871119909119879119871119909 + 119868119877119909119879119877119909 + 119868119879119909119879119879119909 +sum119894

119868119894119879119894 (1)


119871119909 119868119877119909


119877119909 and 119868









119881




119868MCU



(2)















119864

119881


= 0160mA sec(3)

25

20

15

10

5

001 02 03 04 05 06

Time (s)

Curr

ent c

onsu

mpt

ion

(mA

)

OscilloscopeWMR







RGain

RLoad

RShunt

VminusV+

V13V

0

1

1

2

2

334

5

45

6

6

78

7

+ minusAD620AN

IMOTE

U1


0

5

10

15

20

25

100 200 300 400 500 600 700 800 900 1000

Erro

r (

)

IMOTE (120583A)





100 200 300 400 500 600 700 800 900 1000

07

06

05

04

03

02

01

00

Erro

r (

)

IMOTE (120583A)





(4)






25

20

15

10

5

0

minus5

minus10

minus15

minus20

Dev

iatio

n (

)

1 10 100 1000 10000 100000





119862Δ119881

Δ119879

997904rArr


119862Δ119881

Δ119879

997904rArr



(5)




119868Node = 120572119891 (6)



1= 50KΩ 119877

2= 35 KΩ 119877

3= 1KΩ









5 Results





U1AAD8552AR3

2 4

8

1

0Q2A

SI9933DY

3

0

U2B

TLC552C

GND7

2DIS13 2OUT92RST10

14

2THR12

2CON112TRI8

7

U2A

TLC552C

GND7

1DIS1 1OUT51RST4

14

1THR2

1CON31TRI6

0

0 6

0

5

0

1 9

4

0

2

10

0

8

IMOTE 1mA

100Ω

RScale

RLoad

RShunt

Vsupply1 Vsupply2

500kΩR1_TRIMMER

1Ω

33V

33V100nF

100nF

100nF100nF

CLoad

C3C2

C1

100 kΩ

10kΩ

10kΩ

R2_TRIMMER

R3_TRIMMER

+

minus

VCCVCC


10000

1000

100

10

1

01

1 10 100 1000 10000 100000


Pulse

freq

uenc

y (H

z)





Vin

Vin

Vout

Cin

Cout

GND

IMOTE

SW

LTC3525-33+minus

L

Switch_regulator

SHDN







Freq

uenc

y (H

z)

100000

10000

1000

100

10

11 10 100 1000 10000 100000

BSN current (mA)

Freq

uenc

y (k

Hz)

100

90

80

70

60

50

40

30

20

10

05 10 15 20 25 30 35






5

45

4

35

3

25

2

15

1

05

00 05 1 15 2 25 3 35 4 45 5


Log

BSN

curr

ent

y = 10008x minus 01101

R2 = 1


25

20

15

10

5

0

minus515 2 25 3

Time (s)

Time (s)

Curr

ent c

onsu

mpt

ion

(mA

)

25

20

15

10

5

0

Curr

ent c

onsu

mpt

ion

(mA

)

OscilloscopeWMR

15 16 17 18 19 20 21 22 23 24 25



045 05 055 06 065 07 075 08 085

5

10

15

20

25

Time (s)

Curr

ent c

onsu

mpt

ion

(mA

)










74444


42223









6 Discussion

















References








































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation



















119864

119881


= 0160mA sec(3)

25

20

15

10

5

001 02 03 04 05 06

Time (s)

Curr

ent c

onsu

mpt

ion

(mA

)

OscilloscopeWMR







RGain

RLoad

RShunt

VminusV+

V13V

0

1

1

2

2

334

5

45

6

6

78

7

+ minusAD620AN

IMOTE

U1


0

5

10

15

20

25

100 200 300 400 500 600 700 800 900 1000

Erro

r (

)

IMOTE (120583A)





100 200 300 400 500 600 700 800 900 1000

07

06

05

04

03

02

01

00

Erro

r (

)

IMOTE (120583A)





(4)






25

20

15

10

5

0

minus5

minus10

minus15

minus20

Dev

iatio

n (

)

1 10 100 1000 10000 100000





119862Δ119881

Δ119879

997904rArr


119862Δ119881

Δ119879

997904rArr



(5)




119868Node = 120572119891 (6)



1= 50KΩ 119877

2= 35 KΩ 119877

3= 1KΩ









5 Results





U1AAD8552AR3

2 4

8

1

0Q2A

SI9933DY

3

0

U2B

TLC552C

GND7

2DIS13 2OUT92RST10

14

2THR12

2CON112TRI8

7

U2A

TLC552C

GND7

1DIS1 1OUT51RST4

14

1THR2

1CON31TRI6

0

0 6

0

5

0

1 9

4

0

2

10

0

8

IMOTE 1mA

100Ω

RScale

RLoad

RShunt

Vsupply1 Vsupply2

500kΩR1_TRIMMER

1Ω

33V

33V100nF

100nF

100nF100nF

CLoad

C3C2

C1

100 kΩ

10kΩ

10kΩ

R2_TRIMMER

R3_TRIMMER

+

minus

VCCVCC


10000

1000

100

10

1

01

1 10 100 1000 10000 100000


Pulse

freq

uenc

y (H

z)





Vin

Vin

Vout

Cin

Cout

GND

IMOTE

SW

LTC3525-33+minus

L

Switch_regulator

SHDN







Freq

uenc

y (H

z)

100000

10000

1000

100

10

11 10 100 1000 10000 100000

BSN current (mA)

Freq

uenc

y (k

Hz)

100

90

80

70

60

50

40

30

20

10

05 10 15 20 25 30 35






5

45

4

35

3

25

2

15

1

05

00 05 1 15 2 25 3 35 4 45 5


Log

BSN

curr

ent

y = 10008x minus 01101

R2 = 1


25

20

15

10

5

0

minus515 2 25 3

Time (s)

Time (s)

Curr

ent c

onsu

mpt

ion

(mA

)

25

20

15

10

5

0

Curr

ent c

onsu

mpt

ion

(mA

)

OscilloscopeWMR

15 16 17 18 19 20 21 22 23 24 25



045 05 055 06 065 07 075 08 085

5

10

15

20

25

Time (s)

Curr

ent c

onsu

mpt

ion

(mA

)










74444


42223









6 Discussion

















References








































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










RGain

RLoad

RShunt

VminusV+

V13V

0

1

1

2

2

334

5

45

6

6

78

7

+ minusAD620AN

IMOTE

U1


0

5

10

15

20

25

100 200 300 400 500 600 700 800 900 1000

Erro

r (

)

IMOTE (120583A)





100 200 300 400 500 600 700 800 900 1000

07

06

05

04

03

02

01

00

Erro

r (

)

IMOTE (120583A)





(4)






25

20

15

10

5

0

minus5

minus10

minus15

minus20

Dev

iatio

n (

)

1 10 100 1000 10000 100000





119862Δ119881

Δ119879

997904rArr


119862Δ119881

Δ119879

997904rArr



(5)




119868Node = 120572119891 (6)



1= 50KΩ 119877

2= 35 KΩ 119877

3= 1KΩ









5 Results





U1AAD8552AR3

2 4

8

1

0Q2A

SI9933DY

3

0

U2B

TLC552C

GND7

2DIS13 2OUT92RST10

14

2THR12

2CON112TRI8

7

U2A

TLC552C

GND7

1DIS1 1OUT51RST4

14

1THR2

1CON31TRI6

0

0 6

0

5

0

1 9

4

0

2

10

0

8

IMOTE 1mA

100Ω

RScale

RLoad

RShunt

Vsupply1 Vsupply2

500kΩR1_TRIMMER

1Ω

33V

33V100nF

100nF

100nF100nF

CLoad

C3C2

C1

100 kΩ

10kΩ

10kΩ

R2_TRIMMER

R3_TRIMMER

+

minus

VCCVCC


10000

1000

100

10

1

01

1 10 100 1000 10000 100000


Pulse

freq

uenc

y (H

z)





Vin

Vin

Vout

Cin

Cout

GND

IMOTE

SW

LTC3525-33+minus

L

Switch_regulator

SHDN







Freq

uenc

y (H

z)

100000

10000

1000

100

10

11 10 100 1000 10000 100000

BSN current (mA)

Freq

uenc

y (k

Hz)

100

90

80

70

60

50

40

30

20

10

05 10 15 20 25 30 35






5

45

4

35

3

25

2

15

1

05

00 05 1 15 2 25 3 35 4 45 5


Log

BSN

curr

ent

y = 10008x minus 01101

R2 = 1


25

20

15

10

5

0

minus515 2 25 3

Time (s)

Time (s)

Curr

ent c

onsu

mpt

ion

(mA

)

25

20

15

10

5

0

Curr

ent c

onsu

mpt

ion

(mA

)

OscilloscopeWMR

15 16 17 18 19 20 21 22 23 24 25



045 05 055 06 065 07 075 08 085

5

10

15

20

25

Time (s)

Curr

ent c

onsu

mpt

ion

(mA

)










74444


42223









6 Discussion

















References








































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










25

20

15

10

5

0

minus5

minus10

minus15

minus20

Dev

iatio

n (

)

1 10 100 1000 10000 100000





119862Δ119881

Δ119879

997904rArr


119862Δ119881

Δ119879

997904rArr



(5)




119868Node = 120572119891 (6)



1= 50KΩ 119877

2= 35 KΩ 119877

3= 1KΩ









5 Results





U1AAD8552AR3

2 4

8

1

0Q2A

SI9933DY

3

0

U2B

TLC552C

GND7

2DIS13 2OUT92RST10

14

2THR12

2CON112TRI8

7

U2A

TLC552C

GND7

1DIS1 1OUT51RST4

14

1THR2

1CON31TRI6

0

0 6

0

5

0

1 9

4

0

2

10

0

8

IMOTE 1mA

100Ω

RScale

RLoad

RShunt

Vsupply1 Vsupply2

500kΩR1_TRIMMER

1Ω

33V

33V100nF

100nF

100nF100nF

CLoad

C3C2

C1

100 kΩ

10kΩ

10kΩ

R2_TRIMMER

R3_TRIMMER

+

minus

VCCVCC


10000

1000

100

10

1

01

1 10 100 1000 10000 100000


Pulse

freq

uenc

y (H

z)





Vin

Vin

Vout

Cin

Cout

GND

IMOTE

SW

LTC3525-33+minus

L

Switch_regulator

SHDN







Freq

uenc

y (H

z)

100000

10000

1000

100

10

11 10 100 1000 10000 100000

BSN current (mA)

Freq

uenc

y (k

Hz)

100

90

80

70

60

50

40

30

20

10

05 10 15 20 25 30 35






5

45

4

35

3

25

2

15

1

05

00 05 1 15 2 25 3 35 4 45 5


Log

BSN

curr

ent

y = 10008x minus 01101

R2 = 1


25

20

15

10

5

0

minus515 2 25 3

Time (s)

Time (s)

Curr

ent c

onsu

mpt

ion

(mA

)

25

20

15

10

5

0

Curr

ent c

onsu

mpt

ion

(mA

)

OscilloscopeWMR

15 16 17 18 19 20 21 22 23 24 25



045 05 055 06 065 07 075 08 085

5

10

15

20

25

Time (s)

Curr

ent c

onsu

mpt

ion

(mA

)










74444


42223









6 Discussion

















References








































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










U1AAD8552AR3

2 4

8

1

0Q2A

SI9933DY

3

0

U2B

TLC552C

GND7

2DIS13 2OUT92RST10

14

2THR12

2CON112TRI8

7

U2A

TLC552C

GND7

1DIS1 1OUT51RST4

14

1THR2

1CON31TRI6

0

0 6

0

5

0

1 9

4

0

2

10

0

8

IMOTE 1mA

100Ω

RScale

RLoad

RShunt

Vsupply1 Vsupply2

500kΩR1_TRIMMER

1Ω

33V

33V100nF

100nF

100nF100nF

CLoad

C3C2

C1

100 kΩ

10kΩ

10kΩ

R2_TRIMMER

R3_TRIMMER

+

minus

VCCVCC


10000

1000

100

10

1

01

1 10 100 1000 10000 100000


Pulse

freq

uenc

y (H

z)





Vin

Vin

Vout

Cin

Cout

GND

IMOTE

SW

LTC3525-33+minus

L

Switch_regulator

SHDN







Freq

uenc

y (H

z)

100000

10000

1000

100

10

11 10 100 1000 10000 100000

BSN current (mA)

Freq

uenc

y (k

Hz)

100

90

80

70

60

50

40

30

20

10

05 10 15 20 25 30 35






5

45

4

35

3

25

2

15

1

05

00 05 1 15 2 25 3 35 4 45 5


Log

BSN

curr

ent

y = 10008x minus 01101

R2 = 1


25

20

15

10

5

0

minus515 2 25 3

Time (s)

Time (s)

Curr

ent c

onsu

mpt

ion

(mA

)

25

20

15

10

5

0

Curr

ent c

onsu

mpt

ion

(mA

)

OscilloscopeWMR

15 16 17 18 19 20 21 22 23 24 25



045 05 055 06 065 07 075 08 085

5

10

15

20

25

Time (s)

Curr

ent c

onsu

mpt

ion

(mA

)










74444


42223









6 Discussion

















References








































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation











Freq

uenc

y (H

z)

100000

10000

1000

100

10

11 10 100 1000 10000 100000

BSN current (mA)

Freq

uenc

y (k

Hz)

100

90

80

70

60

50

40

30

20

10

05 10 15 20 25 30 35






5

45

4

35

3

25

2

15

1

05

00 05 1 15 2 25 3 35 4 45 5


Log

BSN

curr

ent

y = 10008x minus 01101

R2 = 1


25

20

15

10

5

0

minus515 2 25 3

Time (s)

Time (s)

Curr

ent c

onsu

mpt

ion

(mA

)

25

20

15

10

5

0

Curr

ent c

onsu

mpt

ion

(mA

)

OscilloscopeWMR

15 16 17 18 19 20 21 22 23 24 25



045 05 055 06 065 07 075 08 085

5

10

15

20

25

Time (s)

Curr

ent c

onsu

mpt

ion

(mA

)










74444


42223









6 Discussion

















References








































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation















74444


42223









6 Discussion

















References








































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation




















References








































RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation




























RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation









research article a framework for the estimation and validation of energy consumption ... · 2019....

Documents