Deadline-aware heuristics for reliability optimization in ubiquitous mobile edge computing

With the advent of affordable and widely accessible broadband and mobile internet, there has been a significant surge in user demand. These demands, especially when considering latency and user preferences, exhibit a highly dynamic nature in the realm of Ubiquitous Mobile Edge Computing (UMEC). In (...

Full description

Saved in:
Bibliographic Details
Main Authors: Zaman, S.K.U., Maqsood, T., Ramzan, A., Rehman, F., Mustafa, S., Shuja, J.
Format: Article
Published: Springer Science and Business Media Deutschland GmbH 2023
Online Access:http://scholars.utp.edu.my/id/eprint/38006/
https://www.scopus.com/inward/record.uri?eid=2-s2.0-85176745485&doi=10.1007%2fs41060-023-00473-x&partnerID=40&md5=3f4227d45415634713df901c21abeefc
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:With the advent of affordable and widely accessible broadband and mobile internet, there has been a significant surge in user demand. These demands, especially when considering latency and user preferences, exhibit a highly dynamic nature in the realm of Ubiquitous Mobile Edge Computing (UMEC). In (UMEC), the end devices offload computationally intensive tasks to proximate edge servers. The subtasks of a large task are distributed to different edge nodes to improve reliability. Moreover, some tasks in UMEC are deadline constrained, neglecting which may lead to task failure. Optimizing reliability in Ubiquitous Mobile Edge Computing (UMEC) is crucial to ensure consistent and dependable performance of edge computing systems across a wide range of devices and environments. The existing work focuses on the tradeoff between latency and reliability in task offloading to UMEC. In this work, we define the deadlines of different tasks that require the offloading, considering their latency requirements and offloading failure probabilities. A critical issue in UMEC is to find a reliable server. We propose a novel deadline-aware heuristic for task offloading that divides the tasks into subtasks. The heuristic algorithm considers the latency and computing capacity of edge nodes to reduce task failure ratio and optimize the reliability. We consider the latency and offloading failure probability as performance evaluation parameters. The simulation results reveal that the proposed Deadline-aware Heuristic Algorithm (DHA) achieves a remarkable total latency of 12.67 ms, coupled with a mere 0.095 probability of offloading failure. In contrast, the state-of-the-art technique exhibits a latency of 19 ms, and a higher offloading failure probability of 0.38. © 2023, The Author(s), under exclusive licence to Springer Nature Switzerland AG.