Adaptive sign algorithm for graph signal processing

Yan, Yi; Kuruoğlu, Erçan E.; Altınkaya, Mustafa A.

doi:10.1016/j.sigpro.2022.108662

Cited by 14 publications

(3 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Bringing together equations ( 16), (17), and ( 18) takes one to (14). □ For our purposes of analytical description of the learning behavior of our novel method, a more adequate description of it is…”

Section: The Sm-sign-nlms Algorithmmentioning

confidence: 99%

“…This paper advances a framework that is able to furnish an algorithm that combines both Set-membership and signed approaches. The advanced algorithm eliminates the need for parameter selection by leveraging prior knowledge of the noise, a feature shared by both correntropy-based and the least mean p-th power algorithms [14], [15]. Despite its simple update equation and low computational burden, the learning behavior of the advanced algorithm is very sophisticated (a feature it shares with adaptive algorithms in general).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Stochastic Modeling of the Set-Membership- Sign-NLMS Algorithm

De Souza,

Henriques,

Siqueira

et al. 2024

IEEE Access

View full text Add to dashboard Cite

This paper proposes the Set-Membership sign-NLMS (SM-sign-NLMS) adaptive filter, which combines the ability for data censoring (offered by Set-Membership schemes) with robustness against impulsive noise (provided by signed schemes). The algorithm can present a much lower steady-state probability of update than the standard SM-NLMS algorithm when impulsive noise is present in the system. It is derived from a local deterministic optimization problem modulated by a minimum disturbance cost function combined with a bounded error criterion. Several stochastic models are proposed in order to extract insights and a time-variant step size extension of the algorithm. The first of them, based on energy conservation arguments, leads to a fixed-point analytic equation whose solution predicts the asymptotic performance of the algorithm. Further, a transient analysis based on a statistical decoupling of the radial and (discrete) angular distributions of the input vector is derived. Based on such an analysis, an efficient time-variant step-size version of the algorithm is proposed. Additionally, such an analysis is also utilized to obtain a fixed-point formula whose solution describes the asymptotic performance when the unknown plant that the filter intends to match varies according to a first-order Markovian model. Lastly, a novel stochastic model is advanced for the description of the algorithm learning behavior under a deficient-length scenario for a white input signal, which provides some insights about the asymptotic performance of the algorithm. The findings are confirmed by extensive simulations.

show abstract

Section: The Sm-sign-nlms Algorithmmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Stochastic Modeling of the Set-Membership- Sign-NLMS Algorithm

De Souza,

Henriques,

Siqueira

et al. 2024

IEEE Access

View full text Add to dashboard Cite

show abstract

“…If we assume the most common assumption where graph signals are under Normal noise, we can use the graph least mean squares (GLMS) algorithm to conduct an online estimation of node signals [8]. The GLMS algorithm is the foundation of many other adaptive GSP algorithms, which later improve the GLMS in aspects such as faster convergence speed and increased robustness under impulsive non-Gaussian noise [9], [10], [2], [11]. However, as pointed out earlier, GSP algorithms are defined solely on the nodes but we need algorithms that can operate on the graph edges.…”

Section: Introductionmentioning

confidence: 99%