Search

 Abstract—In order to solve the decision fusion issue under a dynamic wireless sensor network (WSN) based on noncoherent multi-access channels, a decision fusion rule based on the quadrature components of the information received by fusion center is proposed. It is assumed that the distribution of fading channels submits a Rayleigh distribution and the number of sensors follows a Poisson distribution in the WSN. The analysis and simulation on detection performance of the decision fusion rule are made under different system parameters. Compared with the optimal likelihood ratio decision fusion rule, the proposed decision fusion rule can provide very close detection performance with smaller computation and less information. Recently, due to their high flexibility, robustness, mobility, and cost effectiveness, wireless sensor networks (WSNs) have attracted much attention, and the issues of decision fusion based on WSNs have become a very active research area. Usually, a typical WSN consists of a large number of low-cost and low-power sensors. When the decision fusion is processing, each sensor makes a local decision based on its observation, and transmits its decision to the fusion center via a wireless channel, finally, the fusion center gives a global decision based on the local decisions. Considering the influences of nonideal channels on the decision fusion, fusion rules with the statistical information of channels have been proposed in (1) and (2). Aiming to improve the performance of decision fusion, in (3) and (4), decision fusion issues in large-scale WSNs has been studied. The above-mentioned researches are based on the parallel access scheme. Under this scheme, an independent wireless channel is required between each sensor and fusion center for information transmission, which requires large communication bandwidth. In order to reduce the requirement of communication bandwidth, some researchers investigated decision fusion issues over multi-access channels (5), (6). The present researches are based on the static WSNs in which the number of sensors is constant. However, the number of sensors involved in decision making may change