Decentralized Parameter Estimation by Consensus Based Stochastic Approximation

By Srdjan S. Stankovic, Milos S. Stankovic, Dusan M. Stipanovic. 2007

201 citations recorded by the source corpus.

Subfields: Computer Networks and Communications, Statistical and Nonlinear Physics

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Networks and Communications

In Statistical and Nonlinear Physics

Citations in the corpus

Distributed Parameter Estimation in Sensor Networks: Nonlinear Observation Models and Imperfect Communication (2008, background): “In [23], the authors consider decaying weight sequences as we do in , thereby establishing also convergence to the desired parameter value.”
Adaptation, Learning, and Optimization over Networks (2014, unknown): “23), also with an iteration dependent sequence μ(i) that decays to zero, was employed in [34, 228] to ensure almost-sure convergence and agreement among agents.”
Diffusion LMS Strategies for Distributed Estimation (2010, methodology): “The work [15] is related to one of our diffusion algorithms from [8] and is discussed later in Section III-B.”
Performance of a Distributed Stochastic Approximation Algorithm (2012, background): “…process (Yn,i)n≥1 represents the observations made by agent i. [Gossip step] Node i is able to observe the values θ̃n,j of some other j’s and computes the weighted average: θn,i = N∑ j=1 wn(i, j) θ̃n,j , (2) where the wn(i, j)’s are scalar non-negative random coefficients such that ∑N j=1 wn(i,…”
Diffusion strategies for adaptation and learning over networks: an examination of distributed strategies and network behavior (2013, methodology): “” The ATC form (16), again with a time-dependent sequence ( ) i n that decays toward zero with time, was also employed by [42] to ensure agreement.”
Stochastic consensus over noisy networks with Markovian and arbitrary switches (2010, background): “Stankovic, Stankovic, and Stipanovic (2007) considered decentralized parameter estimation by combining stochastic approximation of individual nodes with a consensus rule.”
Adaptive Networks (2014, methodology): “The ATC form (49b), also with an iteration-dependent sequence (cid:8) ð i Þ that decays to zero, was further employed in [121] and [122] to ensure almost-sure convergence and agreement among agents.”
Convergence of a Multi-Agent Projected Stochastic Gradient Algorithm for Non-Convex Optimization (2011, background): “Let us also mention [38] which focuses on the case of quadratic objective functions.”
Diffusion Adaptation Strategies for Distributed Optimization and Learning Over Networks (2011, unknown): “The weighting matrix Σk,i−1 defined in (63) is a symmetric, positive semi-definite matrix, and satisfies: 0 ≤ Σk,i−1 ≤ γ2kIM (64) where γk , max {∣∣∣∣∣1−µk N∑ l=1 sl,kλl,max ∣∣∣∣∣, ∣∣∣∣∣1−µk N∑ l=1 sl,kλl,min ∣∣∣∣∣ } (65) May 15, 2012 DRAFT 19 Proof: By definition (63) and the fact that Hl,k,i−1 is…”
Diffusion Adaptation over Networks (2012, methodology): “A special case of the ATC strategy (134) corresponding to the choice C = I with decaying step-sizes was adopted in [34] to ensure convergence towards a consensus state.”
Distributed Network Optimization With Heuristic Rational Agents (2012, background): “These include the study of asynchronous updates [24], time-varying [15] or unreliable communication links [2], [3], [25], [26], and communication contaminated with random noise [2], [5], [8], [27].”