An optimized architecture and algorithm for resource allocation in D2D aided fog computing

In a Fog computing system, efficient resource allocation is crucial for achieving ultra-low latency and high offloading ratio. This paper presents a Shortest Path Resource Allocation (SPRA) algorithm for the allocation of resources in device-to-device (D2D) assisted fog computing framework. In the proposed model, an interconnected fog network with multiple channels is considered. Moreover, each channel can accommodate multiple users based on the fulfilment of certain constraints. The presented algorithm works by allocating resources to end-users one by one while keeping the cost minimum, to maximize the network management profit. Algorithm finds the best path to connect an end-user to fog nodes with minimum possible cost. An end-user is assigned to a particular fog node on a given channel if all its requirements are satisfied, along with the channel's SINR and power constraints. The algorithm ensures a high offloading ratio along with very low time complexity. Simulation results examine the effectiveness of the proposed algorithm in allocation of resources, and the comparative analysis demonstrates the superiority of the proposed scheme.