DEVS-based Evaluation of UAVs-based Target-search Strategies in Realistically-modeled Missions

Abstract

Searching for targets from a group of Unmanned Aerial Vehicles (UAVs) is a complex problem, whose applications range from the localization of military targets to search and rescue missions. Determining the best locations to search within the mission scenario requires to consider the dynamics of the UAVs and of its onboard sensors, and the uncertainty of the problem, usually related with the target initial location and dynamics, and with the sensor likelihood. Besides, what is best is not always the same (e.g. it can be maximizing the detection probability and/or minimizing the target detection time, while ensuring communications, smooth trajectories, energy saving, etc). These makes the evaluation of UAVs target-search strategies a complex system itself. In this paper, we tackle this problem using the Discrete Event System Specification (DEVS) to exploit its modular and hierarchical design, and to improve the reusability and scalability of our evaluation system. DEVS also provides simple and clear semantics to manage the complexities of the system, represents an explicit separation between the model specification and the corresponding simulation, and helps us to debug and verify our model, as the results of the paper show.