Diffusion LMS Strategies for Distributed Estimation

• Published in: IEEE transactions on signal processing Vol. 58; no. 3; pp. 1035 - 1048
• Main Authors: Cattivelli, F.S., Sayed, A.H.
• Format: Journal Article
• Language: English
• Published: New York, NY IEEE 01.03.2010; Institute of Electrical and Electronics Engineers; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Adaptive networks; Algorithm design and analysis; Algorithms; Analytical models; Applied sciences; Convergence; Design engineering; Detection, estimation, filtering, equalization, prediction; Diffusion; diffusion LMS; diffusion networks; distributed estimation; energy conservation; Energy consumption; Exact sciences and technology; Information, signal and communications theory; Least squares approximation; Miscellaneous; Optimization; Parameter estimation; Performance analysis; Power consumption; Robustness; Scalability; Sensor phenomena and characterization; Signal and communications theory; Signal processing; Signal, noise; Strategy; Telecommunications and information theory; Wireless sensor networks; Performance evaluation; distributed estimation; Scalability; Wireless telecommunication; diffusion LMS; Algorithm; Failures; Implementation; Optimization; Remote sensing; Adaptive networks; Power consumption; Simulation; Energy conservation; Electric power consumption; Signal processing; diffusion networks; Wireless network; Robustness; Sensor array; Signal detection; Least mean squares methods
• ISSN: 1053-587X; 1941-0476; 1941-0476
• DOI: 10.1109/TSP.2009.2033729

We consider the problem of distributed estimation, where a set of nodes is required to collectively estimate some parameter of interest from noisy measurements. The problem is useful in several contexts including wireless and sensor networks, where scalability, robustness, and low power consumption are desirable features. Diffusion cooperation schemes have been shown to provide good performance, robustness to node and link failure, and are amenable to distributed implementations. In this work we focus on diffusion-based adaptive solutions of the LMS type. We motivate and propose new versions of the diffusion LMS algorithm that outperform previous solutions. We provide performance and convergence analysis of the proposed algorithms, together with simulation results comparing with existing techniques. We also discuss optimization schemes to design the diffusion LMS weights.