Uplink Rate-Splitting Multiple Access for Mobile Edge Computing with Short-Packet Communications

2025-02-07

AI Summary

This paper addresses the ultra-low-latency task offloading problem in uplink mobile edge computing (MEC) systems under finite-blocklength (FBL) communication constraints. To this end, it pioneers the integration of rate-splitting multiple access (RSMA) into FBL-MEC scenarios and formulates a joint optimization model for end-to-end successful computation probability (SCP). The proposed framework jointly optimizes task offloading ratios, transmit power, and edge computing resource allocation. An efficient hybrid algorithm—combining alternating optimization (AO) and successive convex approximation (SCA)—is designed to solve the non-convex problem. Compared with benchmark schemes such as NOMA, the proposed RSMA-based approach achieves significantly higher SCP and lower end-to-end latency. Results demonstrate RSMA’s unique capability to simultaneously enhance spectral efficiency and reliability under FBL constraints, establishing a novel paradigm for ultra-reliable low-latency intelligent edge computing.

Technology Category

• Machine Learning: Learning on the Edge & Model Compression
• Search and Optimization: Mixed Discrete/Continuous Search
• Multiagent Systems: Mechanism Design

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

In this paper, a Rate-Splitting Multiple Access (RSMA) scheme is proposed to assist a Mobile Edge Computing (MEC) system where local computation tasks from two users are offloaded to the MEC server, facilitated by uplink RSMA for processing. The efficiency of the MEC service is hence primarily influenced by the RSMA-aided task offloading phase and the subsequent task computation phase, where reliable and low-latency communication is required. For this practical consideration, short-packet communication in the Finite Blocklength (FBL) regime is introduced. In this context, we propose a novel uplink RSMA-aided MEC framework and derive the overall Successful Computation Probability (SCP) with FBL consideration. To maximize the SCP of our proposed RSMA-aided MEC, we strategically optimize: (1) the task offloading factor which determines the number of tasks to be offloaded and processed by the MEC server; (2) the transmit power allocation between different RSMA streams; and (3) the task-splitting factor which decides how many tasks are allocated to splitting streams, while adhering to FBL constraints. To address the strong coupling between these variables in the SCP expression, we apply the Alternative Optimization method, which formulates tractable subproblems to optimize each variable iteratively. The resultant non-convex subproblems are then tackled by Successive Convex Approximation. Numerical results demonstrate that applying uplink RSMA in the MEC system with FBL constraints can not only improve the SCP performance but also provide lower latency in comparison to conventional transmission scheme such as Non-orthogonal Multiple Access (NOMA).

Problem

• Enhances Mobile Edge Computing efficiency
• Optimizes uplink Rate-Splitting Multiple Access
• Ensures low-latency short-packet communications

Innovation

• Uplink RSMA for MEC
• Short-packet FBL regime
• Optimized task offloading