Reliability-Optimal Offloading in Multi-Server Edge Computing Networks with Transmissions Carried by Finite Blocklength Codes

Abstract

In this paper, we consider a multi-server mobile edge computing (MEC) network supporting low latency computation services, where the wireless data transmission/offloading are carried by finite blocklength (FBL) codes to satisfy the latency constraints. We characterize the FBL reliability of the transmission phase and investigate the extreme event of queue length violation in the computation phase by applying extreme value theory. Following the obtained characterizations, we provide an optimal framework design including time allocation and server selection, aiming at minimizing the overall error probability. Via simulations, we validate our analytical model and show the impact of the number available servers and total workloads on the system performance.