Cooperative Detection by Multi-Agent Networks in the Presence of Position Uncertainty

Abstract

Target detection in multi-agent networks encompasses a wide range of applications in both commercial and military contexts. Different from their conventional counterparts with static agents, mobile networks provide a natural alternative for performing detection tasks due to their advantages of flexibility and coverage. In this paper, we first establish a general framework of joint target detection and self-localization in mobile multi-agent networks. Next, we propose a cooperative target detection scheme and derive a generalized likelihood ratio test detector via direct localization of target and agents. The lower and upper detection performance bounds are derived as a function of effective signal-to-noise ratio (SNR). Moreover, we quantify the effects of position uncertainty, the number of agents and the SNR of received signals on the detection performance. A hybrid coordinate descent localization algorithm is developed for joint position estimation and cooperative target detection. Numerical results demonstrate the superior performance of the proposed direct detection scheme compared with conventional detectors, revealing its potential for target detection in mobile scenarios.