Competitive Task Offloading in Hierarchical Edge-Cloud Compute Continuum

The introduction of new technologies in user equipment (UE), radio access and the core network has facilitated service provision for task offloading in the recently developed cellular network architectures. One such advancement, has been the shift from classical cloud based designs to a more flexible compute continuum using mobile edge computing (MEC). To serve the requirements generated by novel applications, a redefinition of network interconnections to handle the new traffic considering the inherent challenges is essential. Different works have contributed to the analysis of MEC-based task offloading. The literature manages to capture specific variations of application and system modeling, but in turn, illustrates an absence of a holistic approach on the matter, working as a basis for all other studies. In this work,
we propose and evaluate a system for hierarchical task offloading, involving three computational tiers at the Internet of Things (IoT), MEC and the cloud. As the core objective of service deployment, the model includes monetary based metrics to devise offloading policies. Delay-cost trade-off will also be studied. Providing a game-theoretic assessment, the conflicting utility models of each service tier are highlighted, and competitive strategy settings are reported. The outcomes of this work can serve as a benchmark for following inter-tier task offloading mechanisms to be studied, while also stressing the limits of an unregulated system for service delivery, further motivating research directions for orchestrated and incentivized operation.