Coalitional Double Auction For Ridesharing With Desired Benefit And QoE Constraints.

This article focuses on addressing the ridesharing matching and pricing problem by modeling it as a coalitional double auction to maximize the total payoff of drivers while ensuring desired benefits and quality of experience (QoE) for both drivers and passengers. It proposes an efficient algorithm called CAMP, supported by two pricing strategies—secondary pricing (SEP) and sacrificed minimum bid (SMB)—that guarantee truthfulness, individual rationality, and budget balance, and converge to a Nash-stable coalition partition in finite steps. Simulations using a real-world taxi trajectory dataset from Beijing demonstrate that CAMP outperforms baseline algorithms in terms of drivers’ total payoff and matching ratios, with SEP-based strategy achieving higher matching ratios and SMB-based strategy yielding higher driver payoffs. The study confirms the NP-hardness of the formulated problem and suggests future exploration of reinforcement learning approaches for dynamic request dispatching.