Study of Self-organized Interference Management for Spectrum Sharing Minority Game-based Approach

摘要

A novel self-organized interference management framework for spectrum sharing among multiple wireless systems is proposed. When a low-priority system reuses the shared frequency band spatially, multiple simultaneous transmissions in the low-priority system may cause serious interference to a high-priority system. Therefore, transmitters of the low-priority system must learn to decide whether to transmit or not so that the interference plus noise level at the receiver of the high-priority system is closed to a predefined target value. This form of cognition cycle can be modeled by using a minority game (MG). Agents in the MG choose one of two actions so as to be in the minority side. It is known that the number of agents in the minority side converges to the half number of agents despite their selfish behavior without any information on the number of other agents. In the present paper, the MG-based interference management (MGIM) is proposed in order to manage interference at the high-priority system. In the MGIM, the winner side is determined by the relationship between the interference plus noise level at the receiver of the high-priority system and the predefined target value. Simulation results reveal that the MGIM can achieve high transmission capability of the low-priority system even without centralized control while maintaining the interference level at the receiver of the high-priority system.