Energy-Efficient Channel Access and Data Offloading Against Dynamic Jamming Attacks

Abstract

Multi-access edge computing (MEC) system is developing to take advantage of the adjacent computation devices to reach better performance for IoT networks, and has to fight against malicious jamming attacks and heavy co-channel interference. To address this issue, this paper investigates the energy-efficient MEC optimization problem in a jamming environment where the attack activities of the jammer are time-varying and a priori unknown, and tries to obtain the joint channel access, offloading ratio and power solutions for IoT device-MEC server pairs. To model the interactions between users and the malicious jammer, a joint channel access and data offloading anti-jamming game is formulated. Moreover, it is proved that the proposed game is an exact potential game (EPG), and there exist at least one pure Nash Equilibrium (NE). To obtain NEs in the dynamic jamming environment, a Multi-pattern Best Response based Channel access and Data offloading (MBRCD) algorithm is designed. Through the “game learning” structure, the global network energy consumption can be greatly reduced under the premise of completing computation tasks. Furthermore, simulations indicate that the proposed approach outperforms other comparative approaches, and is well-adapted to the dynamic jamming environment.