Consensus Sigma-Point Information Filter for Large-Scale Sensor Networks

We address an estimation problem of nonlinear dynamic system through a large-scale sensor network. Even though much research has been done in data fusion, the extension to nonlinear dynamic system is recently focused. The main difficulty in data fusion of nonlinear dynamic system comes from that effective nonlinear filters do not allow the information form. In this paper, two algorithms are considered to implement distributed Kalman filtering for a large-scale sensor network. Data fusion problem for a largescale sensor network is tackled by using Kalman-Consensus filter (KCF) whose scalability is suitable for a large-scale sensor network with random topology. Based on KCF fusion algorithm, Sigma-Point Information filter (SPIF) is proposed as a micro-filter of KCF to handle the nonlinear dynamic system. Because of its information fusion structure, it is simple and intuitive to be combined with the consensus algorithm. Newly proposed algorithm called Consensus Sigma-Point Information Filter (CSPIF) shows us the improved accuracy compared with local estimates.

[1]  R. Olfati-Saber,et al.  Consensus Filters for Sensor Networks and Distributed Sensor Fusion , 2005, Proceedings of the 44th IEEE Conference on Decision and Control.

[2]  Ian F. Akyildiz,et al.  Sensor Networks , 2002, Encyclopedia of GIS.

[3]  Fuwen Yang,et al.  Decentralized robust Kalman filtering for uncertain stochastic systems over heterogeneous sensor networks , 2008, Signal Process..

[4]  Robert J. Elliott,et al.  Discrete-Time Nonlinear Filtering Algorithms Using Gauss–Hermite Quadrature , 2007, Proceedings of the IEEE.

[5]  S. Mori,et al.  Distributed Tracking in Distributed Sensor Networks , 1986, 1986 American Control Conference.

[6]  Reza Olfati-Saber,et al.  Distributed Kalman filtering for sensor networks , 2007, 2007 46th IEEE Conference on Decision and Control.

[7]  Hugh F. Durrant-Whyte,et al.  A Fully Decentralized Multi-Sensor System For Tracking and Surveillance , 1993, Int. J. Robotics Res..

[8]  Mahbub Gani,et al.  Distributed estimation for nonlinear stochastic signals over sensor networks , 2008, 2008 IEEE International Conference on Acoustics, Speech and Signal Processing.

[9]  Xiaodong Wang,et al.  Decentralized sigma-point information filters for target tracking in collaborative sensor networks , 2005, IEEE Transactions on Signal Processing.