Motion Planning for Optimal Information Gathering in Opportunistic Navigation Systems
Motion planning for optimal information gathering in an opportunistic navigation (OpNav) environment is considered. An OpNav environment can be thought of as a radio frequency signal landscape within which a receiver locates itself in space and time by extracting information from ambient signals of opportunity (SOPs). The receiver is assumed to draw only pseudorange-type observations from the SOPs, and such observations are fused through an estimator to produce an estimate of the receiver’s own states. Since not all SOP states in the OpNav environment may be known a priori, the receiver must estimate the unknown SOP states of interest simultaneously with its own states. In this work, the following problem is studied. A receiver with no a priori knowledge about its own states is dropped in an unknown, yet observable, OpNav environment. Assuming that the receiver can prescribe its own trajectory, what motion planning strategy should the receiver adopt in order to build a high-fidelity map of the OpNav signal landscape, while simultaneously localizing itself within this map in space and time? To answer this question, first, the minimum conditions under which the OpNav environment is fully observable are established, and the need for receiver maneuvering to achieve full observability is highlighted. Then, motivated by the fact that not all trajectories a receiver may take in the environment are equally beneficial from an information gathering point of view, a strategy for planning the motion of the receiver is proposed. The strategy is formulated in a coupled estimation and optimal control framework of a gradually identified system, where optimality is defined through various information-theoretic measures. Simulation results are presented to illustrate the improvements gained from adopting the proposed strategy over random and pre-defined receiver trajectories.