Decentralized estimation over noisy channels in cluster‐based wireless sensor networks

Abstract

SUMMARY In this paper, the problem of decentralized parameter estimation over noisy channels in a cluster‐based sensor network is studied. Each cluster head generates a local estimate by adopting a sample mean estimator. The local estimates from all cluster heads are compressed by using a one‐bit quantizer and then the bits are transmitted to a fusion center over independent binary symmetric channels. Two maximum likelihood estimators are proposed for estimating the parameter based on the received bits in two different scenarios: (i) all clusters have the same number of cluster members and (ii) the number of cluster members in the clusters are different. The Cramér–Rao lower bounds for the proposed estimators in both scenarios are derived. Simulation results show that the estimation performance for clustering with the same number of cluster members can be comparable to that for cluster‐free sensor networks. However, the estimation performance for clustering with a different number of cluster members is not as good as that for cluster‐free sensor networks. The trade off between estimation performance and energy consumption exists in both scenarios. The energy consumed for executing the estimation task in cluster‐based sensor networks is less than that in cluster‐free sensor networks. Copyright © 2011 John Wiley & Sons, Ltd.

Investigation of wind and rain effects in a foliated tropical region for fixed wireless access

“…From visually inspecting Figure 5, it is clear that Rician distribution is found suitable to model the measured data. There are several estimation methods have been developed to determine parameters of the sample such as method of moments (Hao & Tsai, 2011), method of weighted least square (Liu, Ding, & Shi, 2012), maximum likelihood estimation (MLE) (Liu, Xu, & Chen, 2011) and estimation algorithms as presented in (Vizireanu, 2012;Vizireanu & Halunga, 2011). In this work, the theoretical parameters of this distribution are estimated using the MLE method, while the root mean square (RMS) error (E rms ) calculation is used to assess the degree of agreement between the measured data distributions and the theoretical calculation.…”

Developing the IoT through Wireless Communication Networks: Analysis of Topologies

“…These networks are based on packet switching, so that each block of information is fragmented into transmission small units called packets [8][9][10]. The nodes must be very simple in hardware and user interface components.…”

Multiobjective Optimization for Topology and Coverage Control in Wireless Sensor Networks

“…WSNs play an important role in the improvement of the quality of people's lives [2,3]. Energy is an extremely critical resource in this kind of networks, thus prolonging the lifetime of network and making energyefficient protocols are major design challenges in WSNs [4,5].…”