Research On Advanced Cooperative Communication Technology For 5G

The continued emergence of new service types,the great number of de-vices and the explosion of data traffic pose huge challenges to the fifth gener-ation mobile communications(5G)wireless networks.In order to increase the capacity of 5G network,ultra dense deployment becomes an important feature of 5G systems.In the ultra dense networks(UDN),the coverage of each cell becomes smaller,which leads to severe inter-cell interference and more fre-quent handovers.At the same time,the variation of signals caused by users'mobility increases the probability of handover failure and system signaling overhead.On the other hand,the scarcity of licensed spectrum has become a major bottleneck in improving system capacity.The use of unlicensed spectrum has the inability to ensure the quality of service(QoS)in addition to high proba-bility of collisions and spectrum inefficiency.The cooperative communication among multiple base stations can eliminate interference,reduce the probabil-ity of handover failure and the collision probability with the unlicensed spec-trum.Therefore,cooperative communication is considered to be one of the most promising technologies in 5G.To increase the system capacity,this dissertation carries out the research on advanced cooperative communication technology for 5G.By utilizing co-operative communication,the system capacity can be further increased with advanced physical layer technology,new network architecture and more spec-trum:(1)combining cooperative communication with adaptive modulation and low-density parity-check code(LDPC)to improve spectral efficiency;(2)em-ploying cooperative communication in UDN can effectively solve mobility man-agement problems;(3)using cooperative communication in new radio with un-licensed spectrum(NR-U)to ensure users' QoS to reduce the probability of collision,and improve spectrum efficiency.The main research contents are as follows(1)Cooperative communication adaptive transmission strategy and per-formance analysisIn order to improve the spectrum efficiency of 5G system,this disserta-tion proposes the power-rate adaptive modulation strategy in cooperative com-munication.Firstly,the probability density functions(PDFs)of the received signal-to-interference plus noise ratio(SINR)for three different dynamic point selection/dynamic point blanking(DPS/DPB)coordinated multipoint transmis-sion(CoMP)types are derived.Based on the PDFs,closed-form expressions of average spectral efficiency(ASE)for power-rate adaptive modulation strategy with constant power and rate are obtained.The method of Lagrangian multi-plier is used to analyze the ASE,SINR switching thresholds and power allo-cation schemes for discrete rate under the instantaneous bit error rate(BER)and average BER constraints.The simulation results show that the ASE is im-proved by 2.62bit/s/Hz after using DPS/DPB CoMP.The results also reveal that in DPS/DPB CoMP systems,the transmit power is the crucial parameter to enhance the ASE when the channel condition is unfavorable.For example,when the transmit power remains constant,the ASE can be reduced by 42.46%.Otherwise the transmit rate is the key parameter,when the transmit rate keeps fixed,the ASE can be reduced by 24.76%.When the channel condition is un-favorable,the ASE increases by 52.41%after employing LDPC code.(2)Mobility management scheme based on cooperative communication in ultra dense networkIn UDN,small cells are densely deployed and the distance among cells decreases.At the same time,the fast change of signal caused by users' mo-bility increases the probability of handover failure and system signaling over-head.In this dissertation,joint transmission(JT)CoMP is used to overcome the complicated mobility management problems in UDN.After introducing the reinforcement learning method into the selection of cooperative set,a mobility management scheme based on reinforcement learning for UDN is proposed.In the cooperative set selection algorithm,which is based on the Actor-Critic algo-rithm,the actor uses the strategy gradient to generate a parameterized strategy,and gradually optimizes the parameters.The critics evaluate the actor's strategy based on the value function.It can be seen from the simulation results that the cooperative set selection algorithm proposed in this dissertation can reduce the handover failure rate by 71.02%and increase users' throughput by 58.01%.(3)Design of cooperative transmission scheme based on LBT in unli-censed spectrumWhen using unlicensed spectrum in UDN,more and more NR-U base sta-tions attempt to access the channel,thus the probability of collisions increases which results in repeated backoff of NR-U stations and low spectrum efficiency.At the same time,because the unlicensed bands adopt a contention-based ran-dom access protocol,the QoS of the NR-U users cannot be guaranteed.Con-sidering the above problems,this dissertation designs a novel cooperative LBT protocol,in which cooperative communication among NR-U base stations is ap-plied to avoid collision and improve the spectrum efficiency.Based on the new protocol,a(N+3)-state semi-Markovian model is established to characterize the cooperative network for 5G NR in unlicensed spectrum.Lagrangian dual decomposition method and Karush-Kuhn-Tucker(KKT)condition are used to maximize the effectiveness of the system under the limited backhaul and max-imum transmit power constraints.Simulation results show that the usage of cooperative communication can reduce the collision probabilities from 0.4928 to 0.1632,and increase the effective capacities by 220.04%.