No abstract

No abstract

No abstract

The aim of this paper is to show the problems of implementing the wireless adaptive networks with the free space optical (FSO) technology. Implementing adaptive networks with the wireless optical communication technology has several benefits and also some hindering problems. The thermal optical noise modeled with Gaussian distribution and link turbulence is two of the major problems of this implementation. In this paper, the theoretical analysis of the FSO link effects that are modeled with K-distribution and Negative exponential distributions are considered on the estimation performance of the adaptive incremental networks. These distributions arise when the FSO link is contaminated with strong optical turbulence. Experiments are designed to cover these conditions and the analysis is based on the steady state mean square deviation (MSD) and excess mean square error (EMSE) values for the incremental LMS (ILMS) algorithm and these are the metrics that show how well the adaptive network performs. Simulation results are presented for different parameters of K -distribution and negative exponential distribution and the results show perfect match with the theoretical outcomes. Based on these results, we show that implementing the incremental adaptive networks in the strong turbulence conditions is not feasible and we must think of some countermeasures for these cases. INDEX TERMSFree space optical communications, distributed processing, negative exponential distribution, adaptive networks, strong turbulence, estimation. JUAN WANG (S'19) is currently pursuing the M.Sc. degree in communication and information engineering with the Nanjing University of Posts and Telecommunications, Nanjing, China. Her research interest includes machine learning for wireless communications.

No abstract