外文文獻翻譯-基于ZigBee的車內(nèi)無線傳感器網(wǎng)絡【中文6400字】 【PDF+中文WORD】

外文文獻翻譯-基于ZigBee的車內(nèi)無線傳感器網(wǎng)絡【中文6400字】 【PDF+中文WORD】,中文6400字,PDF+中文WORD,外文文獻翻譯-基于ZigBee的車內(nèi)無線傳感器網(wǎng)絡【中文6400字】,【PDF+中文WORD】,外文,文獻,翻譯,基于,ZigBee,無線,傳感器,網(wǎng)絡,中文,6400 Seediscussions,stats,andauthorprofilesforthispublicationat:https:/ Intra-car Wireless Sensor NetworkHsin-Mu Tsai1,Cem Saraydar2,Timothy Talty2,Michael Ames2,Andrew Macdonald2,and Ozan K.Tonguz11Carnegie Mellon University,ECE Department,Pittsburgh,PA 15213-3890,USA2General Motors Corporation,ECI Lab,Research and Development,Warren,MI 48092-2031,USAEmail:hsinmut,tonguzece.cmu.edu,cem.saraydar,timothy.talty,michael.ames,AbstractDue to an increasing number of sensors deployedin cars,recently there is a growing interest in implementing awireless sensor network within a car.In this paper,we report theresults of packet transmission experiments using ZigBee sensornodes within a car under various scenarios.The results of theexperiments suggest that both Received Signal Strength Indicator(RSSI)and Link Quality Indicator(LQI)can only be used as athreshold-based indicator to evaluate the link quality-indicatingpoor link quality when dropping below a certain threshold.Preliminary experimental results show that a detection algorithmdeveloped by the authors based on RSSI/LQI/error patterns andan adaptive strategy might increase the goodput performance ofthe link while improving power consumption of the radio.I.INTRODUCTIONWireless sensor networks have been implemented in variousmonitoring applications such as industrial,health,environmen-tal,security,etc.Recently,vehicular applications have enteredthe list of applications,mainly through tire pressure monitoringsystems.More widespread use of wireless sensors in a vehiclewill result from one or more of several different factorsincluding difficulty with wired sensing and cost reductionopportunity.Fueled by the emerging interest in the industryfor deploying a higher number of wireless sensors,there isa need to understand and characterize the wireless channelwithin a vehicle.To this end,we report a case study usingwireless sensor nodes that are compliant with ZigBee.ZigBeeis an industry alliance that promotes a set of rules which buildson top of the IEEE 802.15.4 standards 1.Channel behaviorunder various scenarios is observed for ZigBee nodes placedthroughout a midsize sedan.To the best of our knowledgethis paper presents the first attempt to characterize ZigBeeperformance within a vehicle environment.The rest of the paper is organized as follows.In section II,the details of the experimental setup are described.The resultsof the experiments and the discussions of these results arepresented in section III.In section IV,we propose a set ofdetection algorithms and an adaptive strategy that can adjustto channel conditions for improving the error performanceof the wireless channel and preliminary evaluation resultsare presented.Finally,the concluding remarks are given insection V.II.EXPERIMENTALMETHODA.Sensor node hardwareIn our experiments,we use Crossbow MPR2400 2 as oursensor node hardware platform.The specifications are shownin Table I.Fig.1.The block diagram of the experimental setupB.Experimental setup and sensor node firmwareFigure 1 shows the experimental setup.In the experiment,sensor nodes(SN)are placed in different locations in thevehicle.The base station(BS)is placed inside the instrumentpanel of the vehicle,next to the vent.The base station isconnected to a MIB510 programmer and a RS-232 to USBconverter is used to connect the laptop and MIB510.The sensor nodes periodically retrieve sensor informationfrom the attached sensors and send(broadcast)sensor packetsto the base station.The base station acts as a bridging devicebetween the sensor nodes and the laptop,relaying the sensorpackets from sensor nodes to the laptop and the commandpackets from the laptop to sensor nodes,as well as loggingvarious metrics such as Received Signal Strength Indicator(RSSI),Link Quality Indicator(LQI),CRC,etc.,and append-ing them to each received packet.The packet logger/parsersoftware in the laptop processes the packets sent by the basestation,and saves them to a log file for further analysis.Thecommand sender in the laptop can be used to issue commandsto adjust the parameters of sensor nodes such as transmittingpower,packet sending rate,etc.The firmware of sensor nodes and the base station is basedon TinyOS 1.1.15 3.TinyOS is an open source component-based operating system and platform targeting wireless sensornetworks.Our implementations use various API and librariesprovided by TinyOS.1-4244-0353-7/07/$25.00 2007 IEEE This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the ICC 2007 proceedings.TABLE ICROSSBOWMPR2400(MICAZ)SPECIFICATIONParameterValue/DescriptionProcessorATMega 128L ProcessorRadio ChipChipcon CC2420 RadioOperating Frequency2.4 GHzEffective Data Rate250 Kb/sModulation FormatOffset Quadrature Phase-Shift keying(OQPSK)TABLE IISENSORNODELOCATIONS IN THECARNode No.LocationBEmbedded in the instrumentpanel,next to the vent6On the dashboard,next to the light sensor7On the right side of the trunk,next to the stability actuator1In the engine compartment,next to the fuse box0In front of the radiator,between the temperature sensorand the air quality sensorC.Sensor node locationsThe vehicle used in the experiment is a General Motors2005 Cadillac STS.Figure 2 and Table II show the locationsof the sensor nodes as well as the base station in the vehicle.D.Experimental scenariosWe performed different experiments under various scenariosshown in Table III.The details of these scenarios are discussedin the following.1)Locationa)Maintenance garage This is similar to the servicedepot of a regular car dealer.Technicians walk byfrequently and several other cars are parked nearby.There are a lot of service equipments in the garage.b)Corporate parking lot This is a regular corporateparking lot.The test vehicle was parked in one ofthe parking space and surrounded by other cars.Pedestrians passed by the test vehicle sometimes.Fig.2.Sensor node locations in the car.The numbers in the circles areshowing the number of the sensor node.See Table II for descriptions.TABLE IIIEXPERIMENTALSCENARIOSScenario No.LocationDriverEngine1Maintenance GaragePresentON2Maintenance GarageNot PresentOFF3Corporate Parking LotPresentON4Corporate Parking LotNot PresentOFF5On the roadPresentONNearby vehicles sometimes moved in or out of theparking lot.c)Road This is the driving scenario.The car wasdriven on the highway most of the time and some-times on large(multiple-lane)local roads.2)Driver In“driver present”scenarios,the driver wassitting in the car and has frequent movements,such asoperating A/C,radio,steering wheel,etc.In“driver notpresent”scenarios,the drivers seat was empty.3)Engine In“engine on”scenarios,the engine was startedand kept running throughout the whole measurement.The air conditioner and the radio in the vehicle werealso turned on.In“engine off”scenarios,the enginewas turned off(not in the accessory mode)and the keywas removed from the vehicle.E.Communication parametersTransmitting power:In the experiment,we set the trans-mitting power of the sensor nodes to be at 5 differentlevels:0,-5,-10,-15,and-25 dBm.The transmittingpower of the base station was fixed at 0 dBm(The basestation only transmits when sending commands to sensornodes).Packet sending rate:We configured the sensor node tosend a sensor packet every 100 ms.This sending rate issufficient for non-safety sensors in the vehicle.Channel selection:The physical layer standard of MI-CAz nodes follows IEEE 802.15.4 standard 4.Since802.11b/g devices are the most common devices in 2.4GHz ISM band and are likely to create a lot of interfer-ence,we select a channel that is away from the bandwidthoccupied by 802.11b/g standard.The bandwidth usedby 802.11b/g and 802.15.4 devices are 3 MHz and 22MHz,respectively.In our experiment,we configured thesensor nodes to use channel 26(2480 MHz)to avoidinterference from 802.11b/g devices.In this case,theclosest 802.11b/g channel is channel 11(2462 MHz)anddoes not overlap with our 802.15.4 channel.Packet format:Figure 3 shows the sensor packet formatused in the experiment.The total size of MAC ProtocolData Unit(MPDU)plus the frame length field is 31bytes.Note that most of the fields in the applicationlevel payload are used to record the information for theexperiments.For example,we used 12 bytes to recordthe sensor information,as well as 1 byte each to recordtransmitting power and version number of the firmware.The size of the sensor packet can be reduced by removingThis full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the ICC 2007 proceedings.Fig.3.Sensor packet format(the number above each field represent the sizeof the field in byte)5unnecessary fields and results in a lower packet errorrate.Depending on applications,the size of the sensorinformation field can also be reduced.Node transmissions:In the experiments conducted,onlyone of the sensor nodes transmitted at a time.We usedthis setting to avoid interference from other sensor nodesand focus on measuring the link quality.MAC related parameters:In the experiment,we disabledthe automatic ACK feature,as well as the retransmissions.The sensor nodes use a MAC protocol similar to the Car-rier Sense Multiple Access(CSMA)used in 802.11b/g,in which it will wait until the channel is clear(performclear channel assessment)and then start transmitting.Datacollection:Foreachscenario/transmittingpower/sensor node,we configured the sensor nodeto transmit 6000 sensor packets,which took 10 minutes.The total time to complete the data collecting processfor each scenario was around 200 minutes.F.Observable entitiesThe following describes various observable entities recordedby the base station.Link Quality Indicator(LQI)LQI is calculated by Chip-con CC2420 radio chip and is actually Chip CorrelationIndicator(CCI).It is related to the chip error rate.LQIranges from 50 to 110 and is calculated over 8 bitsfollowing the start frame delimiter.Received Signal Strength Indicator(RSSI)RSSI is mea-sured by Chipcon CC2420 radio chip and represents theamount of energy received by the sensor node.Accordingto5,RSSI has a range from-100 dBm to 0 dBmand the maximum error(accuracy)is 6 dB.The RSSI iscalculated over 8 symbol periods.Sequence Number In the sensor data packet,there is anapplication-level sequence number field which will beincreased each time the sensor node sends out a packet.This can be used by the base station to detect a lostpacket.Cyclic Redundancy Check(CRC)field Chipcon CC2420radio chip has automatic CRC checking capability andTinyOS has a CRC field in its radio packet indicatingwhether the packet received pass the CRC checking.TheCRC scheme used in CC2420 is CRC-16(ITU-T).G.Definitions of metricsIn this sub-section,we define the metrics used later.First we define the following variables:G?The number of packets received by the base stationand passed the CRC check.LE?The number of packets received by the base stationand either the length of the packet or the type of thepacket(indicated by the type field)was not correct.CE?The number of packets received by the base stationand failed the CRC check.A?The total number of packets transmittedNote that our packet parser will first detect length/typeerrors.If the length/type of the packet is not correct,it will beput into LE category.The CRC field of these packets mightindicate that it is in error,but these will not be included inCE.Now we define the following error-related performancemetrics using the above variables:Packet Reception Rate(PRR):PRR=G+LE+CEA(1)Packet Error Rate(PER):PER=LE+CEG+LE+CE(2)Goodput:Goodput=GA(3)H.Experiment for understanding the impact of BluetoothTo study how the existence of an interference source can im-pact the performance of the ZigBee sensor nodes,we used theintegrated Bluetooth hands-free in the Cadillac and a MotorolaRAZR V3 cell phone to create interference.We performedthe experiment with and without the Bluetooth interferencein scenario no.3 in Table III(with limited Bluetooth dataset;each node transmitted using only one transmitting powersetting).In the experiment with Bluetooth interference,the cellphone was used to place a phone call and maintain a Bluetoothconnection with the hands-free during the whole experimentperiod.The Bluetooth protocol uses a Frequency Hopping SpreadSpectrum(FHSS)mechanism.It hops to one of the availablechannels every 0.625 ms according to a hopping sequencespecified by the master node.The Bluetooth standard used inU.S.has 79 1-MHz-wide channels spread from 2402 MHz to2480 MHz.Hence,the last two channels will overlap with the802.15.4 channel(2479 MHz)used in our experiment and willcreate interference to the sensor nodes.This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the ICC 2007 proceedings.3025201510505405060708090100Transmitting Power(dBm)Channel Loss(dB)Node 6Node 7Node 1Node 0Receive Sensitivity(94 dBm)Fig.4.Channel loss of the channels to all sensor nodes.The error bars showone standard deviation from the mean.III.EXPERIMENTALRESULTS ANDDISCUSSIONIn this section,we present the experimental results anddiscuss the implications of these results.A.Channel lossThe attenuation of signal strength experienced as the signalpropagates from the transmitter antenna to the receiver antennais referred to as channel loss and it is typically measured indecibels(dB):CL(dB)=Ptransmitted Preceived(4)Note that CL includes the antenna gains.The channel loss depends on a complex set of factors,including the distance between the transmitter-receiver pairand the type of medium along the path between the transmitterand the receiver.The location of each of our wireless sensorsidentifies a channel between the base station node and thecorresponding wireless sensor node.As seen in Figure 4,the best channel among the four that were measured in ourexperiments is the channel to node 6 on the dashboard(nextto the twilight sensor)whereas the worst channel has beenobserved as the channel to node 0 in front of the radiator(between the air quality sensor and the ambient temperaturesensor).Since we fix the locations of our wireless sensor nodes,ideally the channel loss curve should be a flat line(independentfrom the transmitting power)for each channel.However,itcan be observed from our results that for low transmit powervalues,especially for the channel to node 0,channel loss seemsto be decreasing!Such a result appears to be counter-intuitive,however a careful consideration of the circumstances revealsthat if the channel loss is large and transmitting power levelis sufficiently low,the received power level will be lower thanthe receive sensitivity of the receiver hardware most of thetime and will not be received successfully.Without these highchannel loss packets,the average channel loss is higher thanthe expected value.Observe from Figure 4 that insufficientamount of data and statistical effects result in some variationsof the channel loss curves,as opposed to perfectly flat lines.B.Error metrics and RSSI profilesThe receive sensitivity of the radio chip in the sensor nodesis-95 dBm(typical)and-90 dBm(minimum),as specifiedby 5.The sensitivity corresponds to the minimum receivedsignal strength beyond which the packet error rate exceeds1%,as defined in 4.To study the relation between various error metrics andRSSI,we computed the plots as follows.Each 6000-packetsequence of one setting was split to segments of 50 consecutivepackets.For each of the segments,error metrics were calcu-lated over these 50 packets,which is represented by y.Themean of the RSSIs of these 50 packets were also calculatedand represented by x.Then we plot the point with coordinate(x,y)on the figure to represent this segment,and repeat thisprocedure with all segments in this setting,and all the data ofother settings.Figure 5(a)-(e)shows the profiles of PRR versusRSSI for each of the scenarios.Figure 6 and Figure 7 showthe profiles of 1-PER and Goodput versus RSSI,respectively.In Figure 5,one can observe that,in agreement with thespecified receive sensitivity,the PRR drops from 1 to 0within the range-91 to-94 dBm.The outliers that violatethis general observation are due to external effects such asdriver movement within the cabin,interference from otherwireless devices,etc.For instance,802.11b/g access pointsare deployed in the maintenance garage in scenario no.1and 2,which is configured to operate on channel 11,whichis rather close to the frequency band the wireless nodesoperate at.As a result of such effects,the receive sensitivityboundary experiences slight shifts to the right of the figure,representing a less“friendly”propagation environment.Onecan also observe that the trunk data manifests a higher level offluctuations(more outliers),possibly due to the rich multipathenvironment caused by the presence and the motion of thepassenger along the direct path that lies between the sensornode and the base station node.In Figure 6,we observe that in scenarios where the engineis on or the driver is present,there is a higher noise levelwhich leads to poorer PER performance.We also observe thatthere are fewer outliers as compared to the results in Figure 5,which could be explained by a relatively lower impact of thedriver or the engine noise on the correlation between PERand RSSI.In Figure 7,one can also observe that the goodputperformance is good only when RSSI is much larger than thereceived sensitivity boundary.C.Error metrics and LQI profilesIn Figure 8,we calculate the mean(x)and the standarddeviation(x)of the LQIs of the 50 packets in a segment.Then we plot the points(x-x,y)and(x+x,y),andconnect these two points with a line to represent this segmentand repeat the procedure for all other segments.This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the ICC 2007 proceedings.10090807060504000.20.40.60.81RSSI(dBm)PRRScenario 1Node 6Node 7Node 1Node 0Receive Sensitivity(a)Scenario 110090807060504000.20.40.60.81RSSI(dBm)PRRScenario 2Node 6Node 7Node 1Node 0Receive Sensitivity(b)Scenario 210090807060504000.20.40.60.81RSSI(dBm)PRRScenario 3Node 6Node 7Node 1Node 0Receive Sensitivity(c)Scenario 310090807060504000.20.40.60.81RSSI(dBm)PRRScenario 4Node 6Node 7Node 1Node 0Receive Sensitivity(d)Scenario 410090807060504000.20.40.60.81RSSI(dBm)PRRScenario 5Node 6Node 7Node 1Node 0Receive Sensitivity(e)Scenario 5Fig.5.PRR versus RSSI profilesAs shown in Figure 8,one can observe that the variance ofeach segment(50 packets,which is received in a period of 5seconds)in these plots is too large,and hence we conclude thatit is difficult to estimate link quality based on the short-termaverages that our data points represent 6.It should be notedthat we observe a relatively high correlation between LQIand Goodput.Based on these observations,we can developa rule of thumb which says that the Goodput could be usedas an upper bound:If LQI is smaller than a certain value;theGoodput cannot be higher than a certain value.Closer lookat Figure 8 shows an example of such a curve(black dashedcurve)which could be used as this bounding function.D.Bluetooths impactFigure 9 compares the goodput performance with andwithout the Bluetooth interference.As expected,Bluetoothinterference has a big impact on the goodput performance to allnodes and the goodputs decrease by 3%-40%,depending onthe individual node and received power.One can also observethat the impact on goodput is larger on nodes which havepoorer channel quality than others.IV.DETECTIONALGORITHM ANDADAPTIVESTRATEGYExisting studies in the open literature of wireless sensornetwork usually concentrate on how to use various observable96949290888684828000.10.20.30.40.50.60.70.80.91RSSI(dBm)1PERScenario 1Scenario 1(fitted curve)Scenario 2Scenario 2(fitted curve)Scenario 3Scenario 3(fitted curve)Scenario 4Scenario 4(fitted curve)Scenari