Work

A cooperative distributed model to evaluate and optimize task offloading in Mobile Edge Computing

Year: 2025

Type: article

Abstract: In the the context of Mobile Edge Computing, user devices (UDs) aim to direct their service requests to edge servers that provide the best Quality of user Experience (QoE). At the same time, edge domains must meet their own operational constraints, such as server consolidation, which –while reducing resource usage– may degrade QoE perceived by a UD. Given the highly dynamic and distributed nature of such systems, identifying a global optimal solution that balances client- and server-side requirements remains a complex challenge. To address this, we propose an algorithm that optimizes the match between user devices and edge servers by forming adaptive clusters based on two synthetic indicators: (i) the serverside utility of accepting a service request from a particular UD, and (ii) the QoE perceived and/or measured by the UD when sending a request to a specific edge domain. Our model also incorporates cooperative behavior among UDs, who share recommendations about QoE for specific services and edge domains. These recommendations are evaluated using a trust mechanism that assesses the reliability of each UD’s feedback. The proposed algorithm exploits this information to asymptotically maximize the collective utility of edge servers and UDs, which is computed on the basis of the server side utility and the QoE. We demonstrate through theoretical results that the computed solution constitutes a Nash equilibrium and that only those agents providing reliable feedback are rewarded –thus promoting trustworthy collaboration among UDs. Experimental validation confirms the effectiveness of the proposed method.

Source: Computer Communications

DOI: 10.1016/j.comcom.2025.108403

Authors: fabrizio messina, Domenico Rosaci, Athanasios Vasilakos

Institutions: University of Catania, University of Agder

Language: English

Topic: IoT and Edge/Fog Computing

Subfield: Computer Networks and Communications

Field: Computer Science

Domain: Physical Sciences

SDG: Industry, innovation and infrastructure

Open Access status: hybrid

Estimated APC paid: $2,620