Timeliness-Aware Computation Offloading Strategies for IIoT Networks

This paper investigates the peak age of information (PAoI) violation probability and mean PAoI of computation offloading strategies in multi-access edge computing-enabled (MEC-enabled) industrial Internet-of-Things (IIoT) networks. In particular, a comprehensive PAoI analysis framework for computation offloading strategies is proposed in this work. Through closed-form cumulative distribution function (CDF) expressions derived for received signal-to-interference-plus-noise ratios (SINRs) and PAoI arising from tandem M/M/1 queues, new closed-form expressions for PAoI violation probability and mean PAoI are obtained for the uplink timeliness-aware (UTA), joint uplink-and-computing timeliness-aware (JUCTA), and cloud-only (CL) computation offloading strategies. Extensive analysis demonstrate that the proposed UTA and JUCTA strategies outperform the CL strategy in MEC-enabled IIoT networks and are thus viable to support mission-critical IIoT applications. Crucially, it is also shown that the PAoI violation probability and mean PAoI of the considered computation offloading strategies hinges greatly on computation delay, communications radius, and task generation rates.