In large-scale and dense wireless sensor networks, sensor observations often are correlated and the correlation impacts overall network performance. Another performance limiting factor comes from the non-ideal nature of the wireless links between network nodes. In this paper, we study the detection performance for a distributed detection system with dependent observations under noisy communication channels. In particular, by adopting a novel unified hierarchical independence fusion framework, we derive asymptotic performance limits in terms of error exponents by taking into account the impact of dependent observations and non-ideal channels. The error exponents are investigated under both the Bayesian and the Neyman-Person criteria. Moreover, conditions under which the detection system will result in zero error exponents are also presented in terms of channel capacity requirements.
Performance limit for distributed detection under correlated observations in wireless sensor networks
2012-11-01
129617 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
An isolation principle based distributed anomaly detection method in wireless sensor networks
British Library Online Contents | 2015
|