A variable forgetting factor diffusion recursive least squares algorithm for distributed estimation

•A new variable forgetting factor diffusion RLS algorithm for distributed estimation.•Performance analysis of the diffusion RLS algorithm in time-varying systems.•Derivation of RLS solution to the distributed adaptive algorithm and study of the effect of the network topology.•Derivation of optimal f...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Signal processing Ročník 140; s. 219 - 225
Hlavní autoři: Chu, Y.J., Mak, C.M.
Médium: Journal Article
Jazyk:angličtina
Vydáno: Elsevier B.V 01.11.2017
Témata:
Adaptive networks
Diffusion RLS
MSD analysis
VFF
Adaptive networks
MSD analysis
Diffusion RLS
VFF
ISSN:0165-1684, 1872-7557
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:•A new variable forgetting factor diffusion RLS algorithm for distributed estimation.•Performance analysis of the diffusion RLS algorithm in time-varying systems.•Derivation of RLS solution to the distributed adaptive algorithm and study of the effect of the network topology.•Derivation of optimal forgetting factor selection formulae. Distributed recursive least squares (RLS) algorithms have superior convergence properties compared to the least mean squares (LMS) counterpart. However, with a fixed forgetting factor (FF), they are not suitable for tracking time-varying (TV) parameters. This paper proposes a novel diffusion variable FF RLS (Diff-VFF-RLS) algorithm based on a local polynomial modeling (LPM) of the unknown TV system. The diffusion RLS solution is derived analytically such that the estimation deviation from the true value is investigated. Based on the analysis and the LPM of the TV system, a new optimal VFF formula that tries to minimize the estimation deviation is obtained. Simulations are conducted to verify the theoretical analysis in terms of the steady-state mean square deviation (MSD) and the VFF formula. Results also show that the convergence and tracking performance of the proposed algorithm compares favorably with conventional ones.
ISSN:0165-1684
1872-7557
DOI:10.1016/j.sigpro.2017.05.010