Fog Computing Meets URLLC: Energy Minimization of Task Partial Offloading for URLLC Services

Abstract

Ultra-high reliability and ultra-low latency communication (URLLC) are critical challenges for upcoming 6G applications. Cloud computing and mobile edge computing (MEC) offer potential solutions but incur high deployment and maintenance costs due to reliance on central or edge servers. Moreover, the surge in users and data exacerbates latency concerns. Therefore, with more flexible servers deployment, fog computing is more capable of URLLC requirements. In this work, we propose a fog computing model utilizing mobile devices’ computing capabilities to mitigate latency delays. We characterise the problem as an optimisation problem in quadratic variables. And we reduce the problem to a mixed integer convex optimisation problem in two dimensions using decomposition subproblems. Based on this, we introduce a partial offloading algorithm based on the finite blocklength (FBL) mechanism, which improves the energy efficiency. Simulations demonstrate the efficiency of our algorithm in URLLC, with a 49% reduction in energy consumption compared to no retransmission and a 36% reduction in energy consumption compared to infinite blocklength (IBL) coding.

Keywords

• Fog computing
• dynamic voltage and frequency scaling (DVFS)
• partial offloading
• finite blocklength (FBL)
• ultra-high reliability and ultra-low latency communication (URLLC)
• 6G