Distributed Estimation of IIR System’s Parameters in Sensor Network by Multihop Diffusion LMS Algorithm
In literature, distributed LMS algorithm for finite impulse response (FIR) systems has been studied as it is stable inherently. But infinite impulse response (IIR) systems are used in wireless sensor network (WSN) based application. Hence, adaptive IIR filter is assumed at each sensor node to estimate the parameters in the network. Diffusion mode of cooperation among the sensor nodes is incorporated. By doing this, the probability of trapping in local minima, which is the major drawback in IIR system, is reduced if each node is well connected to more number of neighbors. But the connectivity is reduced in sparse sensor network. Therefore, IIR multihop diffusion LMS (DLMS) algorithm is incorporated to the sparse sensor network. It is seen from the results of simulation that 2-hop DLMS provides best results by providing least steady-state mean square error and mean square deviation with minimum number of iterations. In order to minimize the communication overhead, block LMS is proposed.
KeywordsIIR systems Diffusion LMS Distributed estimation Wireless sensor network Multihop diffusion
- 1.Li Y, Thai MTE (2008) Wireless sensor networks and applications. SpringerGoogle Scholar
- 3.Nayak M, Panigrahi T, Sharma R (2015) Distributed estimation using multi-hop adaptive diffusion in sparse wireless sensor networks. In: International Conference on Microwave, Optical and Communication Engineering (ICMOCE), Dec 2015, pp 318–321Google Scholar
- 4.Kong J-T, Lee J-W, Kim S-E, Shin S, Song W-J (2017) Diffusion LMS algorithms with multi combination for distributed estimation: formulation and performance analysis. Dig Signal Proc 71(Supplement C), 117–130Google Scholar
- 6.Panigrahi T, Pradhan PM, Panda G, Mulgrew B (2012) Block least-mean square algorithm over distributed wireless sensor network. J Comput Netw Commun., vol. 2012, Hindawi Publishing CorporationGoogle Scholar
- 16.Xiao L, Boyd S, Lall S (2005) A scheme for robust distributed sensor fusion based on average consensus. In: Proceedings of 4th international symposium on information processing in sensor networks, Loss Angles, CA, pp. 63–70, Apr 2005Google Scholar