Task Offloading and Scheduling in Edge Computing Networks with Dynamic Spectrum Sharing

Recent advances in wireless network research have significantly improved communication performance. Using emerging edge computing brings IT services closer to end-users, resulting in minimizing response times. However, several challenges must be overcome to fully realize its benefits, including efficient resource allocation, task prioritization, and processing before deadlines. This paper aims to maximize the number of users who offload their tasks. Specifically, we address two issues: (i) dynamic spectrum sharing between primary and secondary communications using cognitive radio technology, and (ii) task scheduling at the edge server while respecting strict deadlines. We develop algorithms based on various design approaches to jointly solve these issues while resolving performance/complexity trade-offs. Simulation results compare the performance of the proposed algorithms to two benchmarks and demonstrate the near-optimality of the algorithm based on the genetic approach.