ZigBee-based Positioning and Navigation System for Robot

Transcription

1 ZigBee-base Positioning an Navigation Sste for Robot Journal of Convergence Inforation Technolog, Volue 6, Nuber. Januar 0 ZigBee-base Positioning an Navigation Sste for Robot Meiqian Ye, Tianing Chen, 3 Changhong Yu *,,3 School of Inforation an Electronic Engineering, Zhejiang Gongshang Universit, eiqiane@6.co oi:0.456/jcit.vol6. issue.6 Abstract At present, the research on the technolog of obile robot navigation is uner the spotlight. This paper introuces ZigBee technolog into obile robot s research, an constructs an eperiental sste for robot navigation base on ZigBee wireless sensor network localization technolog. In this sste, robot ae a ZigBee oule takes the position as a blin noe in ZigBee network, which fors a wireless location network, together with other ZigBee reference noes. Cobine with several other algoriths like robot obstacle avoiance algorith an navigation algorith, it realizes the position an navigation of obile robot. The eperient result shows that the ZigBee-base robot position an navigation etho iscusse in this paper can eet the requireents of obile robot navigation applications.. Introuction Kewors: ZigBee, Mobile Robot, Positioning, Navigation Mobile robot is a cople sste which rouns up a set of functions, such as environental awareness, naic ecision-aking, behavioral control an ipleentation an so on. In the stu an application of the robot, its position an navigation is the ost basic an iportant issue, an it is one of the ke technologies to achieve real intelligent an copletel autonoous obile robot. Wireless Sensor Networks (WSN) is constitute b a large nuber of low-cost icro sensor noes which are arrange in the onitoring area, foring a ulti-hop self-organization network through wireless counication. Therefore, the noes can collaborativel perceive, collect an process the object s inforation in the coverage area of the network, then sen it to the observers. ZigBee-WSN technolog will epan the perception real of obile robot, an provie a new wa for robot s localization an navigation. This paper cobines the avantages of ZigBee-WSN an obile robot to buil a ZigBeebase position an navigation sste of obile robot. In the environent of WSN, this sste etens robot s perception b ZigBee technolog, to locate robot accuratel an then navigate it... ZigBee-WSN overview WSN is copose of a large nuber of sall sensor noes. It onitors the object s real-tie inforation through various icro-sensors, which will be processe b the ebee coputing resource an be transitte to reote users via wireless counication network. The fraework of WSN sste is as shown in Figure, which generall inclues sensor noes, sink noe an anageent noe. Sensor noes are place in the eaination area an constitute network in a for of self-organization. The ata of a sensor noe onitore will be transitte hop b hop along other sensor noes after preliinar treatent b itself. During the transission, the ata a be ulti-processe, spreas to sink noe through an hops an finall reaches the anageent noe through the Internet

2 ZigBee-base Positioning an Navigation Sste for Robot Journal of Convergence Inforation Technolog, Volue 6, Nuber. Januar 0 Figure. Fraework of WSN sste Sensor noe is usuall an ebee sste, its processing power an storage capacit are relativel weak, an its counication istance is quite liite. Sink noe usuall has a strong processing power, storage capacit an counications capabilities. It can be either an enhance sensor noe which has sufficient energ suppl, ore eor resources an coputing power, or a particular gatewa evice with wireless counication interface. Sink noe connects WSN with the eternal network, achieves the counication between anageent noe an WSN through the protocol conversion. Accoringl, the ata collecte can be forware to the eternal network, an siultaneousl the tasks subitte b anageent noe will be poste. ZigBee is a new technolog of WSN eploent. It is a wireless network technolog focusing on low power, low cost, low copleit, low-rate an short-range wireless counications. ZigBee protocol efines several laers. Each laer is responsible for its own uties an provies services to the upper. Coplete ZigBee protocol is constitute b high-level application specification, application laer, network laer, ata link laer an phsical laer. Its phsical an MAC laer aopt IEEE stanar, but iprove an epane on that basis. Its network laer, application laer an high-level application specification rafte b the ZigBee Alliance. Accoring to the protocol, its frequenc ban is ivie into three bans, which are 868MHZ, 95MHZ an.4ghz, the oulation oe an transission rate on each ban are ifferent..4ghz frequenc ban is ivie into 6 channels an its ata rate is 50kbps. The channel nubers of 868/95MHZ ban are corresponingl an 0, transission rates are 0 an 40kbps. The ain features of ZigBee technolog are following: a) Low ata rate. Onl 0kbps ~ 50kbps, focus on low-transission applications. b) Low power consuption. In the low power stanb oe, two coon 5 V batteries can sustain ore than 6 onths. c) Low cost. Both low ata rate an siple protocol reuce the cost greatl. ) Large network capacit. Each ZigBee network can support at ost 55 evices, which eans each ZigBee evice connects with other 54 evices. e) Low ela. Dela is usuall 5 ~ 30s. f) Sall effective range. Its effective coverage ranges fro 0 to 75, epening on the actual transission power an application pattern... Mobile robot overview Mobile robot is a kin of intelligent robot with highl self-planning, self-organizing an aaptive capacit, which is suitable for working in cople unstructure environents. As one of the iportant robot branches, obile robot s cobination an foration are of great fleibilit because of its autono an obilit. An this gives it greater fleibilit an aaptabilit of applications. In orer to achieve autonoousl ove, obile robot ust have accurate perception, reliable navigation an goo oveent sste. The ke technolog involves sensing technolog, navigation an positioning technolog, echanical esign, otion control, architecture, sart technolog an so on. Aong which, the navigation is a vital issue in the research of

3 ZigBee-base Positioning an Navigation Sste for Robot Journal of Convergence Inforation Technolog, Volue 6, Nuber. Januar 0 autonoous obile robots, an critical to achieve a real intelligent an full autonoous obile robot. Currentl, technologies use in location an navigation of autonoous obile robots can ainl be ivie into two categories: () Robot itself recors the oveent b sensors like coe iscs, electronic groscopes an acceleroeters, calculating the current location b integrating. () Deterine the relative position of the robot an the environent b raar, laser range finer, iage atching, etc. an then get the location. However, because of the strong obilit an oeling ifficult of obile robot, the error accuulate in the first etho has greater ipact on the positioning accurac. The secon approach is of high cost an nees for aitional epensive ancillar equipent..3. Integrate ZigBee with obile robot WSN can continuousl onitor a wie range of the environent, but its ata processing an eecution abilit is quite liite. Mobile robots are with strong coputing power an obilit, but their liitations of perception restrict their intellectual evelopent. In view of their respective avantages an characteristics of ZigBee-WSN an obile robot, people began to seek a cobination to for a new sste of istribute sensing an control network. Integrate ZigBee with obile robot can enhance the robot's perception abilit an the sensor network s control to environent. ZigBee can, not onl provie global real-tie awareness for the obile robot, but also be a eiu between calculation an counication. Meanwhile, obile robot can be a fleible eecutor to anage an aintain the network. Thus it is clear that the research on ZigBee-base obile robot navigation is of great significance.. Sste fraework overview ZigBee-base obile robot navigation sste is ainl constitute b three coponents: WSN use for navigation, navigate obile robot an ZigBee network coorinator. The fraework of sste is as shown in Figure. Figure. Fraework chart of sste ZigBee-WSN consists of an ZigBee wireless sensor noes with location abilit, which are bestrewe in the robot work area, as reference noes of the network with CC430 oule. Reference noe is a known static noe; its coorinate, is fie, an has no position calculation function. Mobile robot is a special feature noe in WSN, carring the wireless counications an positioning oule, as blin noe of the network with CC43 oule. Blin noe is a kin of obile noe, which can ake an oveent in the area encircle b reference noes, an blin noe ust use CC43 chip. The ifference between CC43 an CC430 is that CC43 has a wireless location engine, which can calculate the location of the blin noe in network to achieve Receive Signal Strength Inicator (RSSI)-base positioning. In other wors, blin noe autoaticall receives RSSI fro the reference noes, runs location algorith autoaticall with these RSSIs, an finall calculates its own position,

4 ZigBee-base Positioning an Navigation Sste for Robot Journal of Convergence Inforation Technolog, Volue 6, Nuber. Januar 0 When robot locates itself, it will enable its navigation algorith an behavior control progra to conuct path planning an behavior operation. ZigBee network coorinator serves as the gatewa, responsible for the services an coorination of the wireless network. It can connect with the onitoring terinal (PC) b wire or wireless eans. Mobile robot s coorinate is sent to the onitoring terinal b the gatewa, an the terinal controls the behavior of obile robot through the gatewa. In aition, the bining table, routing tables an equipent inforation can be store in PC, to reuce the buren on the coorinator an iprove network efficienc. 3. Ke algoriths 3.. Noe localization In applications of WSN, location inforation is ver iportant for onitoring network activities, eterining the location of the incient an collecting ata are two basic functions of WSN. Accoring to the nee for easuring the actual istance between noes when locating, the location algoriths can be ivie into istance-easure location an non-istance location. The forer uses actual istance to calculate the location of unknown noe, b easuring the istance or the location inforation between sensor noes; the latter has no nee to easure the actual istance between noes, it estiates the istance between noes an calculates the location onl b the network connectivit an other inforation. Non-istance location has nothing to o with the actual istance an orientation, which reuces the harware requireents of the noe. But its positioning accurac is relativel low. Distance-easure location can provie high positioning accurac. At present, the coonl use istance-easure technologies are RSSI, TOA (Tie of Arrival), TDOA (Tie Difference of Arrival) an AOA (Angle of Arrival). In which, RSSI has avantages of eas realization an low cost, its positioning accurac is up to 3. The basic principle of RSSI-base location is: knowing the signal strength of sening noe, the receiving noe can calculate the transission loss accoring to the RSSI. Then it will transfor transission loss into istance b use of theoretical or epirical oel, an use the eisting location algorith to calculate the location of the noe. Details see reference [4]. This paper uses the location engine ebee in CC43 to locate noes accoring to the istances. This engine uses Motorola wireless RSSI-base positioning etho. In this etho, 3~8 noes with known aress are as reference noes, which locate the location of the aress unknown noe. The scheatic iagra is as shown in Figure 3--. Figure 3--. RSSI location scheatic iagra First, all the noes in the network counicate with their neighbors, blin noe receives the RSSIs an coorinates of its neighboring reference noes. Then blin noe coputes its istances fro the reference noes accoring to the RSSIs receive. RSSI is a function of the TX power an the istance between the sener an receiver, the relation is as forulae (3-): RSSI 0nlog0 A (3-)

5 ZigBee-base Positioning an Navigation Sste for Robot Journal of Convergence Inforation Technolog, Volue 6, Nuber. Januar 0 A B X Here, is the istance between transitter an receiver; A is efine as the absolute value of the average power in B receive at a close-in reference istance of one eter fro the transitter; n is efine as the path loss eponent that escribes the rate at which the signal power ecas with increasing istance fro the transitter. The actual paraeter n value written to the Location Engine is an integer ine value nae N, their apping table coul be foun in reference []. Finall, after obtaining the istances fro the reference noes, the blin noe uses Maiu Likelihoo Estiation Metho (see reference [4]), to calculate the coorinates of itself. The Maiu Likelihoo Estiation Metho is as shown in Figure 3--. Figure 3--. Maiu Likelihoo Estiation Metho Set: the coorinates of reference noes are:,,,, 3 3,,, n,, their istances to the blin noe D are respectivel,, 3,,, the coorinates of blin noe D is,. Accoring to the istance forula between two points, the relationships between noes are as follows: (3-) Minus the forer n- forulas with the No.n respectivel, getting the following equation set: (3-3) The equation set can be epresse as: B AX Here (3-4) (3-5) (3-6)

6 ZigBee-base Positioning an Navigation Sste for Robot Journal of Convergence Inforation Technolog, Volue 6, Nuber. Januar 0 Use the least square etho to coputer the X estiates: X A A T A B estiate coorinates of the noe D,. 3.. Navigation algorith ˆ T, obtaining the In the research of obile robot navigation sste, there are two ajor probles: positioning an path planning. The RSSI-base positioning algorith entione above provies location inforation for the robot in unknown environent, which eans it has solve the positioning proble. In aition, the robot harware platfor-as-uiii intelligent obile robot aopte b this eperient, which coes with a collision sensor, can avoi obstacle when oving. This paper cobines the above two technologies, figuring out a siple navigation algorith. Robot can get its coorinates at an tie fro the wireless location oule uring locootion, an eterine its travel irection b coparing with the target coorinates. If the robot encounters obstacles in the locootion process, its collision sensor will be triggere, an it will autoaticall change the irection to avoi obstacles, an then re-obtain its current coorinates to eterine the irection of travel. Repeat this process until the robot reaches the target. In aition, the robot's current irection can be evaluate b the two consecutive coorinates easure. The principle of navigation is shown in Figure 3--, the WSN area covere b eight reference noes is a square, an the blin noe (blue ots) can ake an ove in it. The re ot represents the target, its coorinates M,N is set in the behavior control progra. Figure 3--. Navigation algorith scheatic iagra Suppose the two consecutive coorinates easure b blin noe were X, Y, X,Y, then the current irection can be consiere as angle with the X ais as arke in the iagra. Accoring to the latest coorinates X,Y an the target coorinates M,N, we can copute the net irection shoul be angle with the X ais as shown, an then we get the angle robot shoul turnθ α β. When selecting the net irection of travel, this etho uses the real-tie environental inforation of the WSN. So, it can be applie to robot navigation sste in naic environent. 4. Sste ipleentation 4.. Harware esign of sste 4... Harware esign of ZigBee network

7 ZigBee-base Positioning an Navigation Sste for Robot Journal of Convergence Inforation Technolog, Volue 6, Nuber. Januar 0 The ZigBee-WSN esigne in this paper is constitute b three parts: network coorinator, reference noes an blin noe. Appl ZigBee/ wireless SCM (Single Chip Micoco) evelopent sste of C5RF-CC43-ZDK Series as the harware platfor. Network coorinator inclues a backplane an a CC430 RF ZigBee oule. On the backplane las graphic an Chinese characters LCD, keboar, ZigBee oule interface of CC430/CC43, ajustable resistor, LED, power an RS-3 interfaces. The reference noes an blin noe respectivel use ZigBee wireless oules with CC430/CC43 chip. SCH (Scheatic iagra) of esigning the are shown as (a) (b) (c) in Figure 4--. Figure noes SCH of ZigBee network The CC430/CC43 chip integrate CC40 RF receiver, enhance 805 MCU an 8KB RAM. What s ore, CC430/CC43 has DMA (Direct Meor Access), prograing watchog tier, AES (Avance Encrption Stanar)-8 securit protocol processor, 8 to 4- bit ADC, USART, sleep tier, power on reset, brownout etect an prograable I/O. Aitionall, CC43 has a Motorola s location engine. So, in ZigBee-base positioning sste, CC430 oule can onl be use as a reference noe, while the CC43 oule can either be a reference noe or a blin noe Harware platfor of obile robot The object of this position an navigation sste is GuangMaoDa's AS-UIII intelligent obile robot. AS-UIII is equippe with collision sensors, infrare sensors an other sensors, an it can also epan vision sensors an range sensors, with a strong perception on the environent. In orer to achieve the ZigBee-base position an navigation sste propose b this paper, instea of using the AS-UIII vision an ranging sensors uring the eperient, we onl take the AS-UIII as a progra-controlle obile robot platfor. Three ain coponents of the obile robot are the ain controller, wireless counication oule an otion-control sensing oule. Fraework of the obile robot sste is as shown in Figure 4--. Figure 4--. AS-UIII structure sketch - 4 -

8 ZigBee-base Positioning an Navigation Sste for Robot Journal of Convergence Inforation Technolog, Volue 6, Nuber. Januar 0 Motion-control sensing oule is responsible for transitting the inforation collecte b each sensor to ain controller, receiving the coan fro ain controller, an controlling the rotation of the otors. The collision sensors contribute to obstacle avoiing b getting local obstacle inforation. AS-UIII uses NXP's LPC3 MPU as the core for control. LPC3 is base on a 3/6-bit ARM7TDMI-STM CPU which supports real-tie eulation an ebee tracing. With a high-spee 64 kb Flash eor ebee, it can provie a huge buffer space an strong processing power. The ain controller will process the inforation receive fro the otioncontrol sensing oule an the wireless counication oule, then return the corresponing results an orers. The wireless counication oule chooses Chipeon's wireless SoC-CC43, which is not onl responsible for the transceiving of ata, but also coputing the coorinates of blin noe b its location engine. The wireless counication oule is connecte with the ain controller through the ASBUS epansion car on AS-UIII control panel. 4.. Software esign of sste 4... Software esign of ZigBee network The software esign of the network noes is oularize into three laers, incluing harware laer, software laer an application laer. The harware laer is responsible for the collection of ata, eliver of coans an localization. Its core is CC430/CC43 which is with an 805 core an a RF transceiver in it. Software laer transplants the Z-stack to achieve the ata processing an transission. Application laer runs progras written b ourselves, incluing the positioning function, control to the robot, anageent of network ata an so on. The harware laer of ZigBee wireless oule on CC430/CC43 is copatible with /ZigBee. The MAC laer an phsical laer protocols are up to IEEE stanar; network laer an securit laer is forulate b the ZigBee Alliance, provie in the for of librar. Thus, the software esign of the entire sste is ainl about the ipleentation on AF (Application Fraework) laer an APS (Application Support sublaer) laer. Net, the etaile esign of the network noes will be presente. Reference noe: Reference noe is static an knows its own location, whose location in the region ust be configure correctl. Its role in the sste is to provie a packet with its own coorinates an RSSI to the blin noe as a reference. Usuall, location area is constitute b eight reference noes, at least 3~4, the ore reference noes the higher the positioning accurac. Reference noes in the network are with routing function, using chip CC430, Figure 4-- is its flowchart. Figure 4--. Reference noe flowchart - 4 -

9 ZigBee-base Positioning an Navigation Sste for Robot Journal of Convergence Inforation Technolog, Volue 6, Nuber. Januar 0 Blin noe: Role of blin noe in this sste: Calculate its own coorinates b receiving all the RSSI of reference noes in the region an appling location algorith. As it involves coputation of coorinates, chip in this noe ust be CC43. The flowchart of blin noe is as shown in Figure 4--. Blin noe contacts its neighbor reference noes to collect their coorinates an the RSSIs, an calculates its own location accoring to these inforation an input paraeters (A,N), both of which have been entione in forulae (3-). Then it will sen appropriate inforation to the gatewa, siultaneousl to the robot control center through ASBUS epansion car to control the robot's oveents. Figure 4--. Blin noe flowchart Network coorinator: Coorinator (Gatewa) is the central controller, all the vali ata will be transitte through it to the PC, reference noes an blin noe. Coorinator plas a vital role in the whole sste, using chip CC430. First Coorinator receives the configuration ata of reference noes an blin noe provie b the onitoring software, an sens to the appropriate noes in ifferent was. Seconl, it transits the vali ata feeback fro noes to the onitoring software. The flowchart of coorinator is shown in Figure Software esign of obile robot controller Figure Coorinator flowchart

10 ZigBee-base Positioning an Navigation Sste for Robot Journal of Convergence Inforation Technolog, Volue 6, Nuber. Januar 0 The controlle ebers of the core controller involve a nuber of input parts an output parts. The iport parts are priaril wireless counication oules an sensors, an the output parts are a variet of servo otor. AS-UIII use hbri prograing environent VJC.0 which integrates flowcharts an C language to evelop High-level control proceures. VJC.0 prograing environent aopts graphical prograing language VJC, proviing a siple an powerful platfor for the evelopent of intelligent robot projects, proceures an algoriths. Developers can progra either b the intuitive flow chart or b C language. To eet the requireents of the navigation, instea of using the eisting software of AS-UIII, this paper evelops its own control software in VJC.0, realizing obstacle-avoiing for obile robot on the basis of positioning an navigation. The control progra flow chart of obile robot is shown in Figure The ipleentation of this navigation algorith is priaril base on the constant ccle operation of three oules as coorinate easureent, obstacle avoiance an angle rotation, an it will not stop until the robot reaches the target. 5. Eperient an analsis Figure Flowchart of robot control progra This paper takes C5RF-CC43-ZDK ZigBee wireless network/location sste as the eperient platfor, with integrate evelopent environent IAR Ebee Workbench IDE. The prograing environent of AS-UIII is VJC.0. Eperiental steps are as follow: 5.. Buil the harware platfor Insert a CC430 oule into the network epansion boar as the network coorinator, an connect it with PC through a RS3 serial port line, power on. This sste sets 4 reference noes as the reference for the location of blin noe. For higher precision, the nuber of reference noes can be increase to eight. Each reference noe with CC430 oule shoul connect an antenna an install batter. Insert CC43 oule with an antenna on the ASBUS epansion car of AS-UIII, turn on the power. 5.. Downloa progra In a location sste, each noe can onl connect with one coorinator (gatewa) to constitute a network. It will go wrong if equipent has joine the other networks. To avoi such case, we shoul first initialize the phsical aress of all noes. Open the project in IAR an

11 ZigBee-base Positioning an Navigation Sste for Robot Journal of Convergence Inforation Technolog, Volue 6, Nuber. Januar 0 progra three oules: coorinator, the reference noe an blin noe. Connect one sie of C5RF-3-ZDS eulator with coputer through the USB, the other sie with one noe oule, so that the appropriate progra can be ownloae to this oule. After ownloaing, set an unique non-efault IEEE aress for each noe, no repetition between an two noes. The high-level control progra of robot can be ownloae through VJC.0. Before ownloaing, connect the coputer with robot b USB Draw ap of the onitoring scope In this sste, the ap can be set an oifie through software, onitore object an its location can be visuall seen through PC. The ap is rawn in apping software. Loa the ap into the onitor software Z-Location Engine (see reference []), after it is finishe Position reference noes an blin noe Reference noes shoul be settle accoring to the actual istance; the ore accurate the istance calculate the higher positioning accurac. It woul be best to settle the blin noe within the range built b reference noes. Each noe ust be place with the cooperation of onitoring software accoring to its coorinates, which are set b the onitoring software. Ensure that the actual riing positions of the reference noes are the sae with that on the ap. An the blin noe (robot) will be etecte autoaticall after power on Sste ebugging The above steps have constructe the environent of the positioning an navigation sste, the net is ebugging this sste. Place the robot in ZigBee network, after the blin noe appears in onitoring software, set the two paraeters A an N of blin noe. Then start the robot, the position variation of blin noe (robot) can be seen in the onitoring software Results an analses The test site is a 0.5*5.5 square area, taking the northwest corner of it as the origin of coorinates. Put one reference noe at each corner of this fiel. The coorinates are (0, 0), (0, 5.5), (0.5, 0) an (0.5, 5.5). The navigation task is that the obile robot shoul ove fro coorinates (0, ) to (0.5, 5.5), avoiing obstacles in the environent. Robot s locootion path is as shown in Figure 5-. In the figure, the istance between each gri is 0; the black bo here stans for artificiall constructe environental barrier. Figure 5-. Robot s walking path with obstacle avoiing This eperient shows that the ZigBee-base location an navigation etho put forwar b this paper can fulfill the navigation task successfull, eeting the requireents for the navigation applications of obile robot. However, after an eperients, it turns out that the paths of robot are not totall the sae an soeties the location where it stops is not accurate

12 ZigBee-base Positioning an Navigation Sste for Robot Journal of Convergence Inforation Technolog, Volue 6, Nuber. Januar 0 This inicates that soe errors eist in the sste. In this regar, soe further research is in nee. 6. Conclusion This paper presents an iscusses a kin of location an navigation approach base on ZigBee. The blin noe is evelope on our own, which cobines obile robot with wireless location oule, to construct the ZigBee-base navigation eperiental sste for obile robot. Finall, we conuct an eperient an verif that this location an navigation sste can eet the requireents of obile robot navigation. 7. References [] Atul Vera, Hejit Sawant, Jinong Tan, Selection an navigation of obile sensor noes using a sensor network, Pervasive an Mobile Coputing, vol., pp.65-84, 006. [] CC430\CC43 PRELIMINARY Data Sheet (Rev..0) an Application Note (Rev..0) [3] Chu-Fu Wang, Shu-Chien Huang, Jau-Der Shih, Moveent-Assiste Deploent for Irregular Sensor Densit in Wireless Sensor Networks, JCIT: Journal of Convergence Inforation Technolog, Vol. 4, No., pp. 7 ~ 5, 009 [4] DING Ru, ZHENG Tong, ZHAO Li, Design of Oni-irectional Mobile Robot Bo Base on ZigBee Technolog, Machine Tool an Hraulics, vol. 36, no. 7, Jul 008. [5] LI Wenzhong, DUAN Chaou, ZigBee 006 Wireless Network an Wireless Location[D], Beijing Universit of Aeronautics an Astronautics, China, 008 [6] LIANG Hua-wei, Research on WSNs-base approach an sste for obile robot navigation[d], China Universit of Science an Technolog, China, 007 [7] LIU Zhen, DING Ming-Li, WANG Qi, Ipleentation of WSN Multi-Noe Decision Inforation Fusion in Autonoous Navigation of Robot, Acta Electronica Sinica, vol. 36, no., 008. [8] M.Asaque Hussain, Kung Sup Kwak, Positioning in Wireless Bo Area Network using GSM, JDCTA: International Journal of Digital Content Technolog an its Applications, Vol. 3, No. 3, pp. 67 ~ 7, 009 [9] XUE Han, LI Xun, MA Hong-u, Appling Wireless Sensor Network to Intelligent Navigation for Mobile Robots, Chinese Journal of Sensors an Actuators, vol., no. 5, Ma 008. [0] YANG Xiu-ping, LIU Song-an, Mobile Robot Locating an Tracking Sste Design Base on Wireless Sensor Network, Chinese Journal of Electron Devices, vol. 30, no. 6, Deceber 007. [] ZHU Feng-u, Research on location an navigation of robot base on ZigBee[D], Harbin Institute of Technolog, China,