#54   Privacy-Preserving Greedy Link Scheduling for Wireless Networks

Link Scheduling is an important branch of wireless scheduling algorithms for improving wireless throughput, with implications in domains such as Industrial and Agricultural Internet of Things (IIoT and AIoT), smart grids, and Vehicle Ad-Hoc Networks (VANETs). Current approaches for efficient link scheduling primarily focus on computing schedules that maximize the highest-weighted links in a network among all possible schedules. These algorithms are not designed for privacy, as they rely on information about the entire network topology and device link weights. This paper proposes a novel link scheduling algorithm called PriLink with built-in privacy protections, link scheduling performances close to high-performing greedy algorithms, and real-time execution times. PriLink provides privacy benefits over existing algorithms as the entire network topology is not shared, and devices share only links required for computing the schedule hiding everything else. To our knowledge, PriLink is the first implementation of a privacy-preserving link scheduling algorithm. A comparison with high-performing greedy algorithms (Greedy Maximal Scheduling, Local Greedy Scheduling, and Distributed Greedy Scheduling) shows that the PriLink algorithm achieves faster execution times than all algorithms and good scheduling performance with link schedules about 3% lower than the best-performing algorithm for wireless networks comprising 50 devices and about 5% lower for 250 devices. Regarding privacy, we observe that PriLink can hide nearly 85% of network links for networks containing 50 devices from an honest-but-curious adversary. For large networks containing 250 devices, the algorithm can hide more than 95% of the links providing significant privacy benefits for wireless network devices.