Logo image
Up-Downlink AoI-Driven Multi-Source Data Collection in UAV-Assisted Wireless Sensor Networks
Journal article   Peer reviewed

Up-Downlink AoI-Driven Multi-Source Data Collection in UAV-Assisted Wireless Sensor Networks

Mingxiong Zhao, Yiming Xiao, Jianping Yao, Tongda Wang, Jemin Lee and Tony Q. S. Quek
IEEE transactions on wireless communications, Vol.24(2), pp.1178-1192
01/02/2025

Abstract

age of information Autonomous aerial vehicles correlated information Correlation Data collection Downlink Optimization scheduling Simulation Time measurement Trajectory trajectory design Unmanned aerial vehicle Uplink Wireless sensor networks
This paper explores an unmanned aerial vehicle (UAV)-assisted wireless sensor network (WSN), in which one UAV-enabled mobile data collector periodically collects data from a set of ground sensor nodes (SNs) to the data center (DC) and then DC transmits the processed data back to a group of ground users to fulfill their diverse needs. To accurately evaluate information freshness, we introduce the Age of Multi-Sensor Association Information (AomaI) metric by incorporating the multi-source and up-downlink aspects. Under this framework, we formulate the optimization problem aiming to minimize the average AomaI for all users. To tackle this non-convex problem, we decompose it into two sub-problems: the SN-side optimization problem and the UAV-side optimization problem. For the first subproblem, we propose parallel optimization and primal-dual methods to obtain the optimal solution. For the second subproblem, we first determine the optimal UAV transmission power, then develop the data processing and results distribution scheduling strategies for the DC, and lastly propose the task-associated genetic algorithm (TAGA) and the improved Nawas-Enscore-Ham (INEH) algorithm to design the UAV's visiting order. Simulation results demonstrate that uplink and downlink AoI influence each other, and the consideration of up-downlink AoI can effectively enhance the freshness of AomaI.

Metrics

1 Record Views

Details

Logo image