Electronic System-Level Optimization of Network Slicing and Resource Allocation for 6G URLLC Applications

Abstract

In this paper, we propose a complete mathematical framework for optimizing network slicing and resource allocation at the electronic system-level (ESL) of 6G ultra-reliable low-latency communication (URLLC). We present a multi-dimensional, non-convex mixed-integer stochastic optimization problem that jointly optimizes slice admission control, radio resource block (RRB) allocation and transmit power distribution, computation offloading decisions, reconfigurable intelligent surface (RIS) phase-shift beamforming configurations and fronthaul/backhaul capacity partitioning under heterogeneous traffic demands and stringent quality-of-service (QoS) constraints. The proposed model incorporates finite block length (FBL) information theory, Lyapunov drift-plus-penalty stochastic control, and deep reinforcement learning (DRL)-assisted dynamic orchestration. Numerical results confirm reliability and low end-to-end latency across heterogeneous slice types.