Beamforming Based Mobile User Tracking

Beamforming Based Mobile User Tracking
of 2
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  Beamforming Based Mobile User Tracking P. T. Karttunen 1 , T. I. Laakso 1  and J. Lilleberg 2 1 Helsinki University of TechnologyLaboratory of Telecommunications TechnologyOtakaari 5A, FIN-02015 Espoo, FinlandE-mail:, 2 Nokia Mobile PhonesP.O. Box 50, FIN-90571 Oulu, FinlandE-mail: ABSTRACT - One way to increase system capacityin the future telecommunication systems is toemploy adaptive antennas at the base station. Theyenable reception and transmission utilizing narrowbeams which can dramatically reduce interferencefor other users. In this paper we propose a fastmulti-user tracking system based on the spatialdomain beamforming concept. The simulationresults show that the proposed tracking systemreacts fast and gives small tracking errors.1. INTRODUCTION The target tracking problem arises in numerousapplications, e.g., mobile communications where foreach moving user appropriate beamforming basedconnections have to be established and maintained.Target tracking methods in this context enablecontinuous locating of mobile terminals as they movearound in the cell. For this aim, efficient and robustmobile user tracking system is needed. The locationand tracking problem of multiple moving targets couldbe solved by utilizing numerous differentbeamforming methods like MUltiple SIgnalClassification (MUSIC) or Maximum Likelihood(ML) methods [1]. However, continual application of these kinds of algorithms is prohibited from thecomputational complexity point of view. Furthermore,they also introduce the data association problem, i.e.,the beamformer has no way to associate locationestimates to different mobile users. In this paper weemploy the method which can track multiple movingsources efficiently by using a conventionalbeamforming strategy without any greater performancelosses in the case of Direction-of-Arrival (DOA)pointing errors [ 2 ] . The tracking system is enhanced byintroducing an adaptive control strategy [ 3 ] . 2. TRACKING MODEL The tracking model for the antenna array used in thesimulations will be developed in this section. Theadaptive antenna configuration is illustrated in Figure1, where the antenna array receiver with  M   elementsat the base station and  N   surrounding mobile users areshown. The communication signals S n  ( n =1, …,  N  )are crude modeled as a zero-mean Gaussiandistributed processes [ 4 ] . The additive noise processis also drawn from a Gaussian distribution.Furthermore, it is assumed that samples from thesignal and noise process do not correlate with eachother. The beamforming concept is based on theantenna steering vector which represents the azimuth θ  n  response of the antenna array for each source. Thebeamforming operation forms the beams for eachsource and extracts the desired communication signals.The tracking problem becomes to that of continuouslocation estimation as the mobile users move aroundthe base station.Figure 2 shows the components of the tracking system.The communication signals impinged on the antennaarray are downconverted and digitized in the receiverfrontend. The baseband user signals are estimated inthe beamforming unit by using the conventional block beamforming strategy. In the tracking unit the steeringvectors are updated based on the received sampleblock which reflect new locations for mobile users.The updating rule can be derived by minimizing all theinterference signal and noise components orthogonalto desired user signal component [ 5 ] . In the modelfitting unit the new azimuth tracking angles for eachuser are determined by projecting the steering vector θ  n 41Referenceelement23  M  d  S 1 S  2 S  N  BroadsidedirectionEndsidedirection Figure 1 Adaptive antenna array receiver. BeamformeroutputBeamformingunitTrackingunitAdaptive stepsize control unitDynamicalmodel unitModelfitting unitz -1 Receiver Figure 2 Tracking system.  back to the array manifold. The dynamical model unittakes care of the quality of the location estimates. Asthe mobile users in the crossing stage are in theresolution range of the antenna array the locationestimates become worse and the beamformer may startto follow wrong mobile user. This problem can beavoided by switching to an appropriate locationestimation method, like linear regression on the pastestimated location values. Adaptive Step (AS) sizecontrol unit enhances the adaptation speed byadjusting the step size suitably. Large tracking errorsare introduced by using too large step size. On theother hand too small step size can not react in thenonstationary environment. 3. SIMULATION RESULTS The antenna array receiver consists of  M  =8 uniformlydistributed antenna elements with λ  /2 spacing. TwoGaussian distributed sources with equal variance withSNR=10 dB are at the initial azimuth locations of 10 ° and 40 ° . The pointing errors of 5 °  are introduced tothe location estimates of both sources so that theconvergence behavior can be inspected. Figure 3shows the simulation results for the dynamical signalscenario case. Figure 3a) illustrates the target trackingcurves. The moving sources are under the constantspeed of 0.25 Deg/sample. Figure 3b) shows thetracking error behavior for both Least Mean Square(LMS) and AS methods. As can be seen the ASmethod can easily outperform any fixed step sizemethod. Finally Figure 3c) confirms the steadybehavior for the step size. 4. CONCLUSIONS The proposed adaptive method gives good results interms of low misadjustment and fast convergence. Thisis advantageous especially for users followingcurvaceous trajectories. ACKNOWLEDGEMENTS This work is part of a research project of the Instituteof Radio Communication (IRC) funded by theTechnology Development Center (TEKES), NOKIAResearch Center, Finnish Telecom and the HelsinkiTelephone Company. REFERENCES [1]P. S. Unnikrishna,  Array Signal Processing . NewYork, NY: Springer-Verlag, 1988. [ 2 ] R. T. Compton,  Adaptive Antennas , Englewood Cliffs:Prentice Hall, 1988, pp. 448. 050100150−40−30−20−10010203040TIME [SAMPLES]    A   Z   I   M   U   T   H   A   N   G   L   E   [   D   E   G   ] 05010015000.511.522.533.544.55TIME [SAMPLES]    D   O   A   E   R   R   O   R   [   D   E   G   ]  LMS method AS method 05010015010 −2 10 −1 10 0 TIME [SAMPLES]    S   T   E   P   S   I   Z   E Figure 3  Dynamical signal scenario. a) Illustration of tracking of moving sources b) DOA error as a function of time by using LMS with the fixed step size 0.1 and ASmethod. c) Step size behavior for AS method. [ 3 ] P. T. Karttunen, T. I. Laakso and J. Lilleberg,“Tracking of Mobile Users in a MobileCommunications System Using AdaptiveConvergence Parameter”, PIMRC’97, Helsinki,Finland, Sep 1-4, 1997, pp. 989-993. [ 4 ] C. W. Therrien,  Discrete Random Signals and Statistical Signal Processing , Englewood Cliffs:Prentice Hall, 1992, pp. 727. [ 5 ] S. Affes, S. Gazor and Yves Grenier, “An algorithmfor multisource beamforming and multitargettracking”,  IEEE Trans. on Signal Processing , Vol. 44,No. 6, June, 1996, pp. 1512-1522.
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!