Single Bit and Reduced Dimension Diffusion Strategies Over Distributed Networks

Sayin, Muhammed O.; Kozat, Süleyman S.

doi:10.1109/lsp.2013.2273304

Cited by 36 publications

(36 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The data { d i ( n ), x i ( n ), v i ( n )} are zero mean jointly wide‐sense stationary random processes satisfying system model and the following assumptions: 1)The regression data { x i ( n )} are zero‐mean and spatially and temporally independent. 2)The noise process { v i ( n )} is zero mean, white, Gaussian, and spatially and temporally independent of x i ( n ) with variance

σ_{v_{i}}^{2}

at each node. 3)The regression and noise processes { x l ( m ), v i ( n )} are independent of each other for all i , l , m , n …”

Section: System Model and Assumptionsmentioning

confidence: 99%

See 1 more Smart Citation

A family of diffusion normalized subband adaptive filter algorithms over distributed networks

Abadi

Shafiee

2017

Int J Communication

View full text Add to dashboard Cite

Summary This paper solves the problem of distributed estimation in the diffusion networks based on the family of normalized subband adaptive filters (NSAFs). The diffusion NSAF (DNSAF), the diffusion selective partial update NSAF (DSPU‐NSAF), the diffusion fix selection NSAF (DFS‐NSAF), and the diffusion dynamic selection NSAF (DDS‐NSAF) are established based on the general formalism. In DSPU‐NSAF, the weight coefficients are partially updated rather than the entire weights at each node during the adaptation. The DFS‐NSAF selects a subset of subbands and uses them to update the weights. The dynamic selection of subbands is performed in DDS‐NSAF at each node during the weigh coefficients update. In comparison with DNSAF, the DSPU‐NSAF, the DFS‐NSAF, and the DDS‐NSAF have lower computational complexity while the convergence speed of them is close to the DNSAF. Also, by combination of SPU with FS and DS approaches, the DSPU‐FS‐NSAF and the DSPU‐DS‐NSAF are established, which are computationally efficient. In the following, based on the spatial‐temporal energy conservation relation, a unified framework for mean‐square performance analysis of the family of DNSAF algorithms in stationary and nonstationary environments is presented, and the theoretical expressions for learning curve and steady‐state error are derived for entire network. The validity of the theoretical results and the good performance of introduced algorithms are demonstrated by several computer simulations.

show abstract

σ_{v_{i}}^{2}

at each node. 3)The regression and noise processes { x l ( m ), v i ( n )} are independent of each other for all i , l , m , n …”

Section: System Model and Assumptionsmentioning

confidence: 99%

“…The data {d i (n), x i (n), v i (n)} are zero mean jointly wide-sense stationary random processes satisfying system model 1 and the following assumptions [13][14][15][16][17]30 :…”

Section: Assumptionsmentioning

confidence: 99%

A family of diffusion normalized subband adaptive filter algorithms over distributed networks

Abadi

Shafiee

2017

Int J Communication

View full text Add to dashboard Cite

show abstract

“…To aim this, various techniques have been proposed, such as choosing a subset of the nodes [16], [17], [18], [19], selecting V. Vahidpour, A. Rastegarnia, and A. Khalili a subset of the entries of the estimates [20], [21], and reducing the dimension of the estimates [22], [23], [24]. Among these methods, we focus on the second method in which a subset of the entries are selected in communications.…”

Section: Introductionmentioning

confidence: 99%

Analysis of partial diffusion recursive least squares adaptation over noisy links

Vahidpour

Rastegarnia

Khalili

et al. 2017

IET signal process.

View full text Add to dashboard Cite

Partial diffusion-based recursive least squares (PDRLS) is an effective method for reducing computational load and power consumption in adaptive network implementation. In this method, each node shares a part of its intermediate estimate vector with its neighbors at each iteration. PDRLS algorithm reduces the internode communications relative to the full-diffusion RLS algorithm. This selection of estimate entries becomes more appealing when the information fuse over noisy links. In this paper, we study the steady-state performance of PDRLS algorithm in presence of noisy links and investigate its convergence in both mean and mean-square senses. We also derive a theoretical expression for its steady-state meansquare deviation (MSD). The simulation results illustrate that the stability conditions for PDRLS under noisy links are not sufficient to guarantee its convergence. Strictly speaking, considering nonideal links condition adds a new complexity to the estimation problem for which the PDRLS algorithm becomes unstable and do not converge for any value of the forgetting factor.

show abstract

“…Although there is an extensive literature on this topic, e.g., [2,6,7,8,9,10,11] and references therein, we still have significant and yet unexplored problems for disclosure and utilization of information among agents. Prior work has focused on the computationally simple algorithms that aim to minimize certain cost functions through the exchange of local estimates, e.g., diffusion or consensus based estimation algorithms [2,9,3,12,13,14], due to processing power related practical concerns. However, there is a trade-off in terms of computational complexity and estimation performance.…”

Section: Introductionmentioning

confidence: 99%

Team-optimal distributed MMSE estimation in general and tree networks

Sayin

Kozat

Baar

2017

Digital Signal Processing

View full text Add to dashboard Cite

We construct team-optimal estimation algorithms over distributed networks for state estimation in the finite-horizon mean-square error (MSE) sense. Here, we have a distributed collection of agents with processing and cooperation capabilities. These agents observe noisy samples of a desired state through a linear model and seek to learn this state by interacting with each other. Although this problem has attracted significant attention and been studied extensively in fields including machine learning and signal processing, all the well-known strategies do not achieve team-optimal learning performance in the finite-horizon MSE sense. To this end, we formulate the finite-horizon distributed minimum MSE (MMSE) when there is no restriction on the size of the disclosed information, i.e., oracle performance, over an arbitrary network topology. Subsequently, we show that exchange of local estimates is sufficient to achieve the oracle performance only over certain network topologies. By inspecting these network structures, we propose recursive algorithms achieving the oracle performance through the disclosure of local estimates. For practical implementations we also provide approaches to reduce the complexity of the algorithms through the time-windowing of the observations. Finally, in the numerical examples, we demonstrate the superior performance of the introduced algorithms in the finitehorizon MSE sense due to optimal estimation.

show abstract

Single Bit and Reduced Dimension Diffusion Strategies Over Distributed Networks

Cited by 36 publications

References 9 publications

A family of diffusion normalized subband adaptive filter algorithms over distributed networks

A family of diffusion normalized subband adaptive filter algorithms over distributed networks

Analysis of partial diffusion recursive least squares adaptation over noisy links

Team-optimal distributed MMSE estimation in general and tree networks

Contact Info

Product

Resources

About