Investigations On Key Techniques For Next Generation Broadband Wireless Network

With the popularity of the mobile smart terminals and the widespread application of the wireless multimedia data traffic,the data traffic requirement of the mobile broadband wireless network grows in an explosive way.Therefore,how to utilize the limited wireless frequency spectrum to enhance the system performance,e.g.,improve the efficiency of the frequency resources,promote the network capacity,reduce the interference,guarantee the QoS(quality of service)requirements of the users,lower the outage probability and realize the high-speed data transmission while meet the QoS of the users,becomes a crucial problem in the next generation broadband wireless communication.In order to relieve the network pressure and achieve the predicted 1000 times mobile data traffic growth in the next ten years,the communication industry conducts a wide range of researches on the key technologies of the next generation broadband wireless communication and proposes to enhance the service capability of the next generation broadband wireless communication network by introducing effective network architectures such as ultra dense heterogeneous networks and intelligent network with self-sensing and self-adaptioncapabilities.By densely deploying large scale low-power small cells within the coverage area of the traditional high-power macrocells,ultra dense heterogeneous networks can maximize the reuse of the spectrum,bring users and serving cells closer and improve the system throughput.However,the densely deployment of the low-power small cells will increase the co-channel interference in the network,cause the load imbalancing among cells and degrade the fairness among users,which finally limits the improvement of the network performance.Therefore,the study of the interference coordination and user association technologies with self-sensing and self-adaption abilities in the ultra dense heterogeneous networks is of great importance.Due to the coupled relationship between transmit power adjustment and user association,cell range expansion and user association,and user association and scheduling,the power adjustment based interference coordination problem and the cell range expansion bias adjustment based user association problem are non-deterministic polynomial hard problem which cannot be solved within polynomial time.This dissertation deeply investigates the power adjustment based interference coordination problem and cell range expansion bias adjustment based user association technologies under the scenario of the ultra dense heterogeneous network to maximize the system throughput and proportional-fair throughput,and minimize the outage probability.The main contributions of this dissertation are as follows:The interference coordination algorithm based on the modified particle swarm optimization in the ultra dense heterogeneous network is investigated.Since system throughput is a key evaluation criterion for interference coordination algorithms and power adjustment is an efficient interference coordination mothod,a power adjustment algorithm aiming at maximizing the system throughput is proposed.The downlink interference problem in the ultra dense heterogeneous network is investigated in depth.Since the serious multi-source interference from the densely deployed small cells in the ultra dense heterogeneous network will limit the improvement of the system throughput,an interference coordination algorithm which reduces the intercell interference among cells and improves the system throughput by adjusting the transmit power of the small cell is proposed.Considering the alteration of the serving cells happens during the transmit power adjustment of the small cells,it is difficult for the existing power adjustment algorithm to achieve the optimal transmit power of the small cells,therefore,a power adjustment algorithm based on the modified particle swarm optimization is proposed.The convergence and optimality conditions in the modified particle swarm algorithm are studied and the random local search and multi-restart processes are introduced to achieve the global optimal transmit power of the small cells and the maximum system throughput.Simulation results show that,compared with the existing algorithms which do not alter the serving cells,the proposed algorithm can further improve the system throughput and obtain the optimal transmit power of the small cells with polynomial complexity.The power adjustment algorithm aiming at maximizing the system throughput in the ultra dense heterogeneous network with the constraint of the QoS requirement of users is investigated.Since the power adjustment based interference coordination algorithm without QoS requirement may lead to the QoS degradation of the macrocell edge users,the QoS requirement of users is introduced in modeling the interference coordination problem.Both the QoS of users and the alteration of the serving cell during the power adjustment of the cells are considered,and the transmit power of the small cells is optimized through the modified particle swarm algorithm.As extra computational complexity,iterations and running time are needed in the process of abandoning non-feasible solutions and searching for feasible solutions,Lagrange duality is introduced into the initialization of the particle swarm algorithm to improve the quality of the initial particles.The combination of Lagrange duality and modified particle swarm algorithm can save the running time of the algorithm and reduce the computing complexity required by finding the optimal solution.Simulation results indicate that the proposed algorithm can achieve the optimal system throughput while guarantee the QoS requirement of the users.The cell range expansion bias optimization problem in the ultra dense heterogeneous networks is investigated,and aiming at maximizing the rate related utility,a cell range expansion bias optimization algorithm based on Gibbs sampling is proposed.Considering the conventional user association approach will restrict the coverage area of the small cells,lead to load imbalancing of the network and limit the improvement of the system throughput,a user association algorithm is proposed through optimizing the cell range expansion bias of the small cells,which can improve the system throughput,lower the number of the low-rate users and enhance the proportional fair throughput of the system.Considering the coupled relationship among the cell range expansion bias,user association and scheduling,the optimal solution of the cell range expansion bias is difficult to obtain in the optimization of the rate-related utility.Under such circumstances,a cell range expansion bias optimization algorithm based on Gibbs sampling is proposed.Since the centralized cell range expansion bias adjustment requires the knowledge of all the channel gains between each cell and its served users,the message exchange overhead and the computational complexity will be largely increased with the scale expansion of the ultra dense heterogeneous networks.Therefore,a decentralized cell range expansion bias adjusting algorithm which only requires the local message exchange is proposed by delicate deduction,and the optimality of the algorithm is proven.Simulations show that both the proposed centralized and decentralized cell range expansion bias optimization algorithms can achieve the optimal cell range expansion bias of the small cells,and the computational complexity and message exchange overhead of the decentralized algorithm is far less than those of the centralized algorithm.The low-complexity central aided distributed cell range expansion bias optimization algorithm in ultra dense heterogeneous networks is investigated.Considering the computational complexity and iteration time of the cell range expansion bias optimization algorithm in the ultra dense heterogeneous network rise sharply with the expansion of the network scale,the graph construction of the mutual bias influence relationship among small cells,graph coloring based cell grouping algorithm and candidate serving cell selection scheme are proposed.Based on Gibbs sampling,a central aided distributed cell range expansion bias optimization algorithm with less message exchange overhead,computational complexity and iteration time is proposed.The computational complexity and message exchange overhead of the algorithm is analyzed,and the optimality of the algorithm is proven.Simulation results show that compared with the centralized and decentralized algorithms,the proposed central aided distributed cell range expansion bias optimization algorithm can obtain the optimal solution of the cell range expansion bias with the least number of iterations,lowest computational complexity and least message exchange overhead.