Research on end-side network computing power scheduling mechanism based on two-sided market game

This paper addresses the optimization problem of the allocation and pricing of device-to-device (D2D) computing power resources at the end-side network, and proposes a computational configuration model based on dynamic bilateral market game. By constructing utility functions involving three parties: sellers (SUs), buyers (RUs), and the platform, this paper analyzes the impact of commission rates on the equilibrium of the computing power market. Meanwhile, a genetic algorithm is employed to solve the optimal computing power allocation strategy. The simulation results show that: 1) There is a nonlinear inverted U-shaped relationship between the platform's benefits and the commission rate, and there exists a unique optimal commission rate (approximately 20%) that maximizes the platform's benefits. 2) The increase in the commission rate will significantly reduce the utility of both sellers and buyers. As for the platform's profit, due to the dynamic adjustment of supply and demand, it will show a trend of first increasing and then decreasing. This study provides a theoretical basis for the pricing design and platform regulation of the computing power market in the end-side network, and reveals the key mechanism of multi-party interest trade-off under the constraints of the externalities of the end-side network.