An estimator model for distributed estimation in heterogenous wireless sensor networks

In this paper, we deal with distributed estimation using consensus algorithms for heterogenous wireless sensor networks (WSNs). To accommodate with the heterogeneity, we introduce a novel distributed estimator to track the weighted average of the input signals. Different from existing models, we consider a more practical scenario to take account of hierarchical processing abilities of different sensors: type-I sensors with high processing ability and type-II senors with low processing ability for distributed sensor fusion in WSNs. We investigate the properties of our model and illustrate the feasibility of the proposed estimator via a case study where we use the estimator to track the weighted average of a noisy time-varying signal based on the sensors´ noisy and distorted measurements. Convergence analysis in this scenario is given as well as the effect of network topology and estimator parameters are also studied. Simulation results are provided to demonstrate the performance and effectiveness of the proposed estimator.