Deep Learning Based Sensing Resource Allocation for Mobile Target Tracking

Edge computing has emerged as a prospective paradigm to meet ever-increasing service demands in wide ranges of Internet of Things (IoT) applications. However, massive amount of application data and real-time response are not orchestrated smoothly in this pattern, especially in battery-powered and mission-critical sensor networks. In order to cope with the challenges, this paper proposes a new cooperative resource allocation algorithm which couples reinforcement learning networks and prediction neural networks for accurate mobile targets tracking. Specifically, a hierarchical structure that performs collaborative computing is designed for alleviating computing pressure of front-end devices which are supported by edge servers. The optimization objective leading to the minimal system cost is formulated which is constrained by the system response latency and the tracking accuracy. Simulation results demonstrate that the proposed algorithm outperforms the greedy scheme 24.5% in terms of the energy cost. It also obtains significant 23.1% enhancement in tracking accuracy compared with the non-cooperative scheme.