Energy-Delay Aware User Association in mmWave Backhaul Networks using Matching Theory.

This paper investigates an energy efficiency (EE) and delay aware user association problem in millimeter wave (mmWave) downlink networks, in which the wire backhaul links are replaced by the wireless backhaul links to transmit traffic data. To characterize the mmWave network delay performance, the queue theory is adopted and an M/M/1 tandem queuing model is introduced to capture the data transmission latency over mmWave backhaul links and access links. Besides, in order to measure the network EE and delay performance simultaneously, a novel EE-delay aware utility function is constructed and the user association problem is formulated. It is proved that the formulated problem is NP-hard. To solve it, one many-to-one matching game is reformulated and a two-steps user association algorithm is proposed. Simulation results show that the proposed association scheme can obtain significant utility improvement, achieving up to 21% gains compared with the classic association schemes. Moreover, the proposed scheme also demonstrates its superiority on load balancing management in mmWave backhaul networks.