n n Development of algorithms Simulation of realistic transmitter and channel conditions Receiver with template correlation and path detection Support of for EU IP Pulsers WP3b Localization errror analysis and simulation Basis for integration of UWB and Inertial Measurement Unit - - - - localization and tracking UWB system design Jens Schroeder and Stefan Galler Analysis of UWB Localization Errors University of Hannover Institute of Communications Engineering Location Based Services and Systems Group Hannover, Germany n Simulation parameters - - - PN=12 chip BPSK, f =8.25GHz, f =700MHz, f =80GSp/s Pulse rate: , pulse train repetition rate: Varying additive noise: =N (thermal noise) + 10 or 20dB c LP s N t σ 2 repetition n n n n n Pure statistical model Space-variant model System design Simple NLOS detector Tracking (e.g. IEEE) LOS error: Gaussian above real value, additional noise adds uniformly distributed error NLOS error is statistic and does not represent reflections Unrealistic for tracking simulations (e.g. IMST) aspects Processing gain (code length, SNR) vs. desired range Range w/o “large” errors should be desired range works quite well in realistic scenario: Running variance ( ) > 10 cm² aspects Gaussian tracking filter (e.g. Kalman) not optimal for uniformly distributed ranging errors - - - - - - - - - - LOS error: either very accurate or uniformly distributed NLOS error shows influence of reflections Better suits reality for tracking simulations 1 second Motivation Future Work n n n n n n Increasing the number of channels for 3D-localization Evaluation of different channel models Incorporation of “hardware” Tx and Rx Analysis of different synchronisation schemes Development of localization and tracking filters Integration of Inertial Measurement Unit Localization Simulator n n Simulink model to simulate range estimates with varying channel models and a virtual trajectory Block diagram: n n n A random walk of a mobile object within a given area using the The object is initially placed at a fixed position, randomly chooses a new position, a velocity between 0.2- 0.7 m/sec and a pause time between 0.2-2 sec, walks to the new location and pauses, using the chosen parameters. The locations behind a virtual wall are perceived to be , else The resulting is input to the channel model. - - - random waypoint model NLOS LOS virtual trajectory [3]: . Virtual Trajectory Results Receiver Channel Transmitter PN f c f LP Virtual trajectory f c ±f LP x-Corre- lator Path detector Channel model AWGN σ 2 N n Implemented channel models [1]: statistical model with parameters for different envrionments. Here used: industrial LOS/NLOS [2]: space-variant channel model with spatial evolution of channel impulse responses. Here used: office LOS/NLOS - - IEEE 802.15.4a IMST whyless.com PUL ERS PUL ERS ∫ [1] A. F. Molisch, K. Balakrishnan, D. Cassioli, C.-C. Chong, S. Emami, A. Fort, J. Karedal, J. Kunisch, H. Schantz, U. Schuster, and K. Siwiak, "IEEE 802.15.4a channel model - final report," IEEE 2005. [2] J. Kunisch and J. Pamp, "An ultra-wideband space-variant multipath indoor radio channel model," IEEE Conference on Ultra Wideband Systems and Technologies (UWBST), Reston, USA, 2003. [3] D. B. Johnson and D. A. Maltz, "Dynamic Source Routing in Ad Hoc Wireless Networks," in , T. Imielinski and H. Korth, Eds.: Kluwer Academic Publishers, 1996. [4] J. Schroeder, S. Galler, K. Kyamakya, and K. Jobmann, "Analysis and practical comparison of Wireless LAN and Ultra-Wideband technologies for advanced localization," accepted at IEEE/ION Position, Location and Navigation Symposium (PLANS 2006), San Diego, USA, 2006. Mobile Computing Exemplary range in industrial environment [4] measurements 1 ns 1 100ms