Let's Share: A Game-Theoretic Framework for Resource Sharing in Mobile Edge Clouds

Faheem Zafari,Kin K. Leung,Don Towsley,Prithwish Basu,Ananthram Swami,Jian Li

DOI: https://doi.org/10.48550/arXiv.2001.00567

2020-01-03

Abstract:Mobile edge computing seeks to provide resources to different delay-sensitive applications. This is a challenging problem as an edge cloud-service provider may not have sufficient resources to satisfy all resource requests. Furthermore, allocating available resources optimally to different applications is also challenging. Resource sharing among different edge cloud-service providers can address the aforementioned limitation as certain service providers may have resources available that can be ``rented'' by other service providers. However, edge cloud service providers can have different objectives or \emph{utilities}. Therefore, there is a need for an efficient and effective mechanism to share resources among service providers, while considering the different objectives of various providers. We model resource sharing as a multi-objective optimization problem and present a solution framework based on \emph{Cooperative Game Theory} (CGT). We consider the strategy where each service provider allocates resources to its native applications first and shares the remaining resources with applications from other service providers. We prove that for a monotonic, non-decreasing utility function, the game is canonical and convex. Hence, the \emph{core} is not empty and the grand coalition is stable. We propose two algorithms \emph{Game-theoretic Pareto optimal allocation} (GPOA) and \emph{Polyandrous-Polygamous Matching based Pareto Optimal Allocation} (PPMPOA) that provide allocations from the core. Hence the obtained allocations are \emph{Pareto} optimal and the grand coalition of all the service providers is stable. Experimental results confirm that our proposed resource sharing framework improves utilities of edge cloud-service providers and application request satisfaction.

Networking and Internet Architecture,Distributed, Parallel, and Cluster Computing,Multiagent Systems

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to effectively share resources among different service providers in mobile edge cloud computing to meet the requirements of various latency - sensitive applications. Specifically, since edge cloud service providers may not have sufficient resources to meet all resource requests, and how to optimally allocate available resources is also a challenge, a mechanism is needed to achieve resource sharing among service providers. However, different edge cloud service providers may have different goals or utility functions, which makes resource sharing more complicated. For this reason, the paper proposes a multi - objective optimization framework based on cooperative game theory, aiming to achieve effective resource sharing by considering the goals of different service providers.

The key points of the paper include:
- **Problem Background**: Mobile edge computing aims to provide resources for different latency - sensitive applications, but due to limited resources, a single edge cloud service provider may not be able to meet all requirements.
- **Research Motivation**: In order to improve resource utilization and service quality, an effective resource - sharing mechanism is required, taking into account the different goals of different service providers.
- **Solution**: A resource - sharing framework based on cooperative game theory is proposed to solve the resource - sharing problem through a multi - objective optimization method.
- **Theoretical Contribution**: It is proved that for monotonically non - decreasing utility functions, the game is normal and convex, so the core is not empty and the grand coalition is stable.
- **Algorithm Design**: Two algorithms are proposed, namely Game - Theoretic Pareto Optimal Allocation (GPOA) and Polygamous - Polyandrous Matching Pareto Optimal Allocation (PPMPOA). Both algorithms can provide allocations from the core, thus ensuring Pareto optimality and the stability of the grand coalition.
- **Experimental Verification**: The experimental results show that the proposed resource - sharing framework can improve the utility of edge cloud service providers and the satisfaction of application requests.

In conclusion, this paper solves the resource - sharing problem in mobile edge cloud computing by introducing the method of cooperative game theory, and improves resource utilization efficiency and service quality.