Energy-Efficient Computation Offloading for Indoor Localization Based on Game Theory

The topic of localization within indoor environments has recently received significant attention as localization has become an essential component of many Internet of Things applications such as object tracking and health care management. One of the promising approach to provide accurate localization while minimizing energy consumption is to use computational offloading under mobile edge computing system. Thus, the aim of this paper is to minimize the total energy consumption of multiple users by using computation offloading technique between users, mobile edge computing servers and cloud server. The offloading technique that is proposed in this paper should take in consideration users’ accuracy, latency requirements and the maximum capacity of each server. The paper presents the network model and the computation model of the proposed system. Then, the problem formulation is introduced to minimize the total energy consumption which is the sum of all energy consumed by the users in the local devices and the offloaded servers. In order to provide a distributed implementation that is more suitable for the users within localization environment, the paper formulates the proposed problem as a potential game and the existence of Nash Equilibrium is proved where all users have satisfied offloading decision. The paper obtains the optimal solution to act as a reference for the proposed potential game algorithm. Finally, the paper presents and analyzes the results of the potential game distributed computational offloading algorithm by comparing it to local computing, random offloading and the optimal solution techniques.