Find in Library

This paper considers the problem of efficient network design for data collection from various sensors in smart environments using flying base stations, which are realized using unmanned aerial vehicles (UAVs), or drones. The system efficiency is enhanced by maximizing the number of served sensors using the minimum number of UAVs while satisfying particular network constraints. Towards this end, a joint optimization problem is formulated for UAVs placement and sensors assignment in smart environments with massive sensor deployment. Due to the complexity of the optimal solution, a probabilistic learning approach is utilized to find a near-optimal solution. Further, a non-death penalty constraint handling approach is used to deal with difficult and conflicting constraints. Monte Carlo simulation is performed to evaluate the performance of the proposed algorithm in various scenarios, and compare it with the optimal solution to validate the efficiency of the proposed solution. The presented numerical results show that the solutions obtained using the proposed algorithm are generally close or equal to the optimal solution for several scenarios, but with significant complexity reduction, which confirms the efficiency of the proposed algorithm. Moreover, the proposed solution shows significant performance improvement when compared with an efficient greedy algorithm.