The Research Of Coexistence Between LTE-U And WiFi And LTE Random Access Schemes

With a large number of multi-media services turns to the wireless network,the number of mobile communication services grows rapidly,thus take a new challenge to LTE network.As a public mobile communication system,the current LTE system,which deployed in the licensed band,faces the problem of the shortage of spectrum resources.In order to alleviate the problem between the large number of wireless services and the LTE network performance,two aspects could be considered.On the one hand,it is considered to extend the LTE system bandwidth to the unlicensed band,called LTE Unlicensed(LTE-U),which improves the system throughput.However,the unlicensed band is already deployed the WiFi devices,due to the difference between LTE-U and WiFi in the MAC access mechanism,when they are deployed in the same band,the LTE-U devices will excessively occupy spectrum resources,WiFi devices may fail to have opportunities to transmit.Therefore,it is necessary to study the coexistence between LTE-U and WiFi.On the other hand,we can consider optimizing the existing LTE random access mechanism to improve LTE network performance.Based on the above,the main work of this paper includes two aspects.On the one hand,in order to make a friendly coexistence between LTE-U and WiFi in the unlicensed bands,we propose a Q-learning based coexistence mechanism.First,the LTE-U system and the WiFi system in the coexistence scenario are mathematically modeled,in order to get an appropriate number of ABS subframes in a coexistence period.Secondly,the actions,states,reward values of Q-learning have been designed in the coexistence scenario.The LTE-U eNB uses the designed Q-learning module to determine the locations of the ABS subframes,try the best to match the silent period of LTE-U with the WiFi data transmission period.Finally,this mechanism is verified by simulation experiments,it is proved that this mechanism can implement a friendly coexistence between LTE-U and WiFi.On the other hand,when a UE starts an initial access or its uplink is non-synchronised,it will take a contention based random access procedure.In this kind of random access,UE selects the preamble by itself,thus UE may cause in a collision,due to multiple UEs simultaneously select the same preamble.Based on this problem,we propose a method that can quickly identify the occurrence of random access collision and allocates uplink resources to these UEs from a collision resource pool,so that they can successfully complete the random access with a lower system delay.First,we use a tagged preamble to allow eNB quickly identify the collisions,which has been provided in the existing paper.Secondly,we propose a configuration strategy for the collision resource pool.This strategy can ensure the collision UEs successfully obtain uplink resources without occupying uplink resources of non-collision UEs.In addition,a resource selection strategy for collision UEs is also proposed.This strategy can reduce the probability of collision UEs recurring collisions when select uplink resources from the collision resource pool.Finally,it is proved that the mechanism can effectively reduce the delay of contention-based random access,and improve the efficiency of the contention based random access.This paper is consisted of four chapters.Chapter 1 introduces the research background,leads to the significance of these researches;Chapter 2 introduces the basic theoretical knowledge involved in this paper;Chapter 3 proposes a Q-learning based coexistence mechanism between LTE-U and WiFi;Chapter 4 proposes a low delay LTE random access mechanism.