Abstract
Given the proximity of many wireless users and their diversity in consuming local resources (e.g., data-plans, computation and energy resources), device-to-device (D2D) resource sharing between directly connected devices is a promising approach towards realizing a sharing economy. However, most existing research on D2D resource sharing focuses on solutions in a centralized manner that requires global information and strict coordination. It does not allow for distributed computation to exploit users’ local D2D connectivity. This severely impacts scalability and suggests the need for a distributed solution for large-scale D2D resource sharing. First, we adopt an easy-to-implement greedy matching algorithm with distributed fashion and only sub-linear O(log n) parallel complexity (in user number n) for largescale D2D sharing. We provide a rigorous average-case analysis to show that the greedy matching’s average performance is far better than the worst-case approximation ratio 50% as compared to the optimum for multiple representative classes of graphs. For 1D linear networks, we prove that our greedy algorithm performs better than 86.5% of the optimum. For 2D grids, though dynamic programming cannot be directly applied, we still prove this average performance ratio to be above 76%. For the more challenging Erdos-Renyi random graphs, we equivalently reduce to the asymptotic analysis of random trees and successfully prove a ratio up to 79%. Further, we conduct experiments using real data to simulate realistic D2D networks, and show that our analytical performance measure approximates well practical cases. Second, we focus on a D2D trading market that can scale for large numbers of users. We propose a distributed double auction mechanism that exploits local D2D connectivity and uses distributed computation to achieve near-optimal allocative efficiency. The final prices for each matched pair of buyer and seller are adjusted in a way to induce incentive compatibility and depend on their own declarations in terms of quantity and valuation. We prove that the overall mechanism has significant social welfare gains compared to other widely-used distributed pricing mechanisms. It is also individually rational, ex-ante budget balanced using a subscription fee, and robust to perturbations of the model parameters. To render the system fully manipulation-proof, we further propose a distributed auditing scheme that prevents users from altering the decentralized computation to increase their profits. Finally, we model the repeated execution of the mechanism and determine the best trading frequency by taking into account the arrivals and departures of new participants.