Summary
International Technical Conference on Circuits/Systems, Computers and Communications
2016
Session Number:T3-5
Session:
Number:5238
On the Effect of Informed Nodes on Learning over Complex Adaptive Networks
Morteza Farhid, Mousa Shamsi, Mohammad Hossein Sedaaghi, Faramarz Alsharif, Bruno Senzio-Savino, Mohammad Reza Alsharif ,
pp.653-656
Publication Date:2016/7/10
Online ISSN:2188-5079
DOI:10.34385/proc.61.5238
PDF download (1.3MB)
Summary:
In this paper, we investigate the impacts of number of informed nodes on the performance of a distributed estimation algorithm, namely adaptive-then-combine diffusion LMS, based on the data with the temporal and spatial independence assumptions. The study covers different network models, including the regular, the small-world and the random networks. We have two scenarios for our simulation. We change the fraction of nodes according to their links densities. The simulation results indicate that the larger proportion of the uninformed nodes (90% in first and up to 50% in second scenarios) in a network causes lower convergence besides improvement in the mean-square-error performance and that acquiring more information is not necessarily better.