【作者】 王伟；

【导师】 刘祖军；

【作者基本信息】 西安电子科技大学 ， 军事通信学， 2019， 硕士

【摘要】 全双工通信作为一种可以同时同频传输的通信技术,理论上可以使系统的通信容量翻倍。然而,全双工蜂窝网络中存在的小区间干扰以及小区内上下行用户间干扰严重损害了其全双工增益。干扰对齐作为一种有效的干扰管理方法,能够克服传统干扰管理方法的不足,并提高系统容量。本文将干扰对齐算法应用于多小区全双工蜂窝网络的干扰管理中,分析了多小区全双工蜂窝网络的自由度与全双工增益。并针对干扰对齐在多小区多用户场景下的不足,研究了基于干扰图的多小区全双工蜂窝网络干扰对齐算法,可以有效提升系统的自由度与全双工增益。论文的主要工作如下:1、建立了单小区全双工基站的蜂窝模型,分析了单小区全双工基站系统的理论自由度,设计算法对理论值进行逼近,并对算法进行了仿真。2、建立了多小区全双工蜂窝网络模型,推导和分析了多小区全双工蜂窝网络的理论自由度区域。通过系统等效多元方程组的可解性与干扰对齐可行性的关系,将多小区全双工蜂窝网络的自由度问题转换为最优化求解问题。通过对优化问题的求解,分别得到了多小区全双工蜂窝网络两用户和多用户的自由度区域和上界,且分别对两用户和多用户的自由度区域进行了分析,并求得相应的全双工增益,通过仿真和分析表明传统的干扰对齐方法在多小区全双工蜂窝的干扰管理中带来的全双工增益有限。3、针对传统干扰对齐方法在管理多小区全双工蜂窝网络的不足,提出了一种基于干扰图的多小区全双工蜂窝网络干扰对齐算法。该算法首先通过连续干扰抵消以消除干扰图中的定向干扰,将相邻小区间的双向干扰变为单向干扰,接着关闭部分小区的基站并对剩余小区进行分簇,最后通过簇内基站协作和干扰对齐的方法消除蜂窝网络中的小区间干扰和用户间干扰。仿真结果表明在一上行用户和一下行用户的条件下,每个工作的小区可实现的自由度为1.5,全双工增益达到1.5,而传统干扰对齐方法在相同情况下的全双工增益约为1.1,与之相比本文算法有效地提高了系统容量和全双工增益。更多还原

【Abstract】 Being a communication technology that can transmit data at the same time and in the same frequency,the full-duplex communication can theoretically double the communication capacity of the system.However,inter user interference both from the intra-cell and inter-cell in the full-duplex cellular network seriously degrades its full duplex gain.As an effective interference management method,interference alignment can overcome the shortcomings of traditional interference management methods and improve system capacity.In this thesis,interference alignment algorithm is applied to the interference management of multi-cell full-duplex cellular networks.The degrees of freedom(DoF)and full-duplex gain of multi-cell full duplex cellular networks are analyzed.To overcome the deficiency of the traditional interference alignment in multi-cell and multi-user scenarios,an interference graph based alignment algorithm for multi-cell full-duplex cellular networks is proposed,which not only can effectively improve the degree of freedom of the system,but also is not limited by network scale.The main work of the thesis is as follows: 1.A system model of single-cell full-duplex base station is established.The theoretical DoF of such system is analyzed,and an algorithm is designed and simulated to approximate the theoretical value.2.A multi-cell full-duplex cellular network model is established,and the theoretical DoF region of multi-cell full-duplex cellular network is derived and analyzed.By exploring the proper of the system and using the relationship between proper and feasibility,the degree of freedom of multi-cell full-duplex cellular networks can be treated as an optimization problem.By solving optimization,the DoF region and upper bound of two users(one uplink user and one downlink user)and multiple users(multiple uplink users and multiple downlink users)are obtained respectively.The DoF region of two users and multi-user is analyzed respectively,and the corresponding full-duplex gain is calculated and deduced.It is theoretically demonstrated that the traditional interference alignment methods give rise to limited full duplex gain in the interference management of multi-cell full-duplex cellular.3.To overcome the deficiency of interference alignment method in managing multi-cell full-duplex cellular networks,an alignment algorithm based on interference graph and incorporating multiple interference management methods is proposed.Firstly,successive interference cancellation is applied to eliminate the directional interference in the interference graph,turning the bidirectional interference between adjacent cells into unidirectional interference,then the base stations of some cells are closed and the remaining cells are clustered.The simulation results show that the DOF of each cell is 1.5 and the full-duplex gain is 1.5 when there are one uplink user and one downlink user,while the full-duplex gain of the traditional interference alignment method is about 1.1 under the same circumstance.Therefore,the proposed algorithm effectively improves the system capacity and full-duplex gain.更多还原