Optimal Deployment Control for a Heterogeneous Mobile Sensor Network

This paper proposes a hierarchical architecture for a heterogeneous sensor network consisting of both mobile and fixed agents. The hierarchy contains three layers. The lowest level, Layer 0, is the sensing layer and consists of fixed sensors deployed in an environment of interest according to a spatial Poisson distribution of intensity measure A. Layer 1, the next level in the hierarchy, is the aggregation layer and involves mobile agents (e.g., autonomous vehicles) gathering the sensed data from the Layer 0 nodes based on a nearest neighbour rule. The highest level of the hierarchy is the query layer, which represents requests for the data gathered by the sensors to be transmitted, via the Layer 1 agents, to a mobile base station (e.g., a flying aircraft periodically visiting the area of interest). We construct for this network a utility function which basically captures the cost associated with communication, as seen from the point of view of the mobile agents of Layer 1. Based on a gradient descent strategy we propose a motion coordination scheme to maneuver the mobile nodes from an arbitrary initial arrangement to one that minimizes the utility function. The application of the proposed algorithm in an elementary settings is illustrated via numerical simulations