TORE-NET: A Task Replication and Offloading Strategy in Reliable and Energy-Efficient Mobile Edge Computing Networks

Abstract

The rapid growth of Internet of Things (IoT) devices and the demands of ultrareliable low-latency communications (URLLCs) necessitate the deployment of mobile edge computing (MEC). Although MEC can reduce transmission delays and network congestion, its decentralization and processing load create challenges for energy efficiency, high quality of service (QoS), and reliable service delivery. Excessive energy consumption in MEC environments violates green IoT principles. Hence, to achieve green computing and sustainable development, it is vital to predict and monitor the energy consumed by edge servers (ESs). Prior strategies for offloading determination are deficient because they address ESs’ energy efficiency, QoS, and reliability as mutually exclusive objectives. These strategies fail to integrate reliability as a fundamental architectural constraint. This article proposes TORE-NET, a novel heuristic method that simultaneously maximizes QoS and the task success rate while meeting the reliability target and satisfying energy budget constraints across distributed ESs. The efficacy of this approach is ensured through the integration of dynamic energy, which accounts for heterogeneous, real-time workloads, and by applying the task replication technique to meet the reliability target. The experimental results demonstrate that our proposed method, TORE-NET, can achieve an average improvement in QoS of 27.6% (up to 33.9%) while reducing energy consumption by an average of 11.4% (up to 15.4%) compared to the state-of-the-art methods, while satisfying both reliability and energy budget constraints.