Energy-Efficient Distributed Cooperative Localization by Sensor and Communication Link Scheduling

Internet of Things (IoT) can improve efficiency and automation in industry through information sharing among different nodes. For the IoT applications such as unmanned air vehicle (UAV) swarms and other multi-agent systems (MASs), localization information is an essential input to enable autonomous operations. Distributed cooperative localization can improve localization performance and reduce hardware cost as compared to non-cooperative schemes, and it is more flexible than centralized schemes. In a large-scale MAS, inter-agent information exchange can lead to huge communication load. To reduce the communication cost of distributed cooperative localization, we consider a sensor and communication link scheduling problem in which a subset of the available relative measurements and communication links are chosen to be used at each time point. This scheduling problem is formulated as a combinatorial optimization model. As this model is proved to be NP-hard, we design a dynamic programming algorithm to solve it efficiently. In this algorithm, the first step is to determine a set of optimal solutions in an offline manner, each of which corresponds to a specific localization performance requirement. Based on these solutions, an optimal solution is then selected on-the-fly to satisfy the corresponding localization performance requirement and minimize the communication cost at each time point. Using UAV swarms as an example, simulations are carried out to validate the proposed algorithm, and the results suggest its effectiveness and high efficiency.