User Demand Based Fairness Scheduling Research In Ultra Dense Networks

With the widespread popularization of various new mobile terminals such as smart phones and tablet computers,the amount of mobile data has exploded.A large number of hotspots have been formed in cities,posing a huge challenge to existing communication systems.Ultra Dense Networks(UDN)has become an inevitable trend for future mobile networks.In UDN,wired backhaul is limited by deployment costs and deployment environment.Compared with wired backhaul,wireless backhaul has lower cost and is more flexible and efficient.Due to the shortage of low-frequency wireless resources,millimeter wave(mmWave)with characteristics such as short transmission distance,narrow beam,and huge bandwidth has attracted communication workers' intrest.Therefore,the spectrum structure of UDN is coexistence of high and low frequencies.Low-frequency resources are responsible for seamless coverage,and high-frequency resources are responsible for traffic offload in hot areas.This article disassembles the UDN into high-frequency scenario and low-frequency scenario,and conducts fairness scheduling research based on the data demands of user equipments.The main controbutions are as follows:Due to the high propogation of mm Wave,the coverage area of small cell using high frequency resources(HSC)is small.In order to communicate with the gateway,HSCs which are far from the gateway need other HSCs' relay.We first propose a multi-path routing method to find the optimal path from each HSC to the gateway.Interference management is performed for high-frequency scenes and low-frequency scenes,respectively.In high-frequency scenario,HSC is densely deployed.There will be a lot of crossover between mm Wave beams,which will cause great interference to the system.By analyzing the location information of HSCs and the spatial trajectory information of millimeter-wave links,we control that the links with large mutual interference cannot be transmitted at the same time.It effectively reduces the interference between links.In low-frequency scenario,the macro cell and all small cells using low-frequency resources(LSCs)reuse the same radio resources.Therefore,there will be a lot of inter-layer interference and inter-layer interference in the system.In this thesis,a coloring method is proposed to apply the same color to LSCs with less interference.During the resource scheduling process,LSCs with the same color are controlled to reuse resources.For high-frequency and low-frequency scenarios,fairness scheduling schemes based on the data demands of users are preposed,respectively.In high-frequency scenario,the scheduling scheme control many links for simultaneous transmission.At the same time,when adjusting the scheduling priority of links,the data demands of users are included in the constraints to ensure fairness among users.In low-frequency scenario,we design a joint access and backhaul scheduling scheme.This scheme controls multiple LSC to transmit at the same time and adjusts the resource ratio between access links and backhaul links.This method improves the system throughput.In order to ensure the fairness among users,the priority of link being scheduled is adjusted by the demands of users.Simulation results show that our proposed schemes have better performance in improving system throughput and ensuring fairness than traditional scheduling schemes.