低复杂度大规模接入协议设计

发布时间：2017-12-28 11:01

本文关键词：低复杂度大规模接入协议设计　出处：《浙江大学》2017年博士论文　论文类型：学位论文

【摘要】：物联网是第五代移动通信的重要应用场景。物联网中联网设备数目将大幅增长,要求未来的通信系统以较高的传输速率同时支持超大规模的连接;同时,随着网络的快速发展,网络状态、业务状态和信道状态都将变得更加动态,这要求未来的通信系统有较高的自适应性。这些要求将极大地影响多址接入协议的设计。在本学位论文中,我们在信息论理论和编码理论的指导下,重点研究了大规模接入场景中的多址接入协议设计。本学位论文的内容可概括如下:首先,我们提出了一个新型随机非正交多址接入的基本框架,称为无速率多址接入。该框架与传统方案的中心式资源分配不同,接入点仅给用户指定一个随机接入控制函数,然后用户以伪随机方式共享资源进行接入。所有用户的接入过程可共同等效为一种特殊形式的线性叠加无速率码,接入点利用复杂度较低的置信传播算法进行译码。为了进一步提升系统性能,我们将空间耦合思想引入该接入框架,对资源单元进行分类并且设计了不同用户到不同类型资源单元集合的映射关系,据此提出了一种增强的无速率多址接入方案。此外,我们利用外信息传播理论对所提方案进行理论分析,得到了系统可达速率与随机接入控制函数、空间耦合类型、用户预编码等参数的关系,并据此对系统关键参数进行联合优化设计。分析结果和仿真结果一致表明,所提方案可以渐进地达到信道容量,并且具有较低的信令开销和较高的自适应性,符合大规模接入场景的需求。第二,通过利用上述框架中接入模式暗含身份信息的特点,我们提出了一种“免授权”的无速率多址接入方案。该方案可以省去基本框架的“注册”过程,能够降低信令开销和系统时延。此外,我们进一步提出了一种联合用户检测译码算法,以较低的复杂度同时实现信道估计、用户检测和信息译码。基于密度进化理论,我们提出了一种近似分析算法,用于分析所提算法的性能。分析结果和仿真结果一致表明,与基于“授权”的无速率多址接入相比,所提方案只引入极小的误块率性能损失,但可以大幅降低系统的信令开销和时延。此外,我们将该思想推广于基于固定签名矩阵的稀疏码多址接入方案中,提出了一种签名序列跳变的稀疏码多址接入方案。第三,为了简化上行非正交多址接入的用户配对,我们提出了一种基于速率分割的非正交多址接入方案。在所提方案中,每个用户通过将待发送数据分割成多个数据流进而等效成多个虚拟用户。即使不同用户到达接入点的功率相同,接入点仍然可以利用连续干扰消除算法恢复出各个用户的数据,因此该方案可避免复杂的用户配对,并且降低调度复杂度。此外,我们分析了所提方案的性能,得到了系统中断概率和可达速率与系统关键参数之间的闭式关系表达式。根据理论分析结果,我们提出了一种系统关键参数(包括速率分割类型和功率分割类型)的设计方法。最后,分析结果和仿真结果一致表明所提方案在可达速率与中断概率等方面都有明显提升。
[Abstract]:The Internet of things is an important application scene of the fifth generation of mobile communication. The IOT will significantly increase the number of networked devices, requirements of future communication systems with high data rate and large scale support connection; at the same time, with the rapid development of network, network status, traffic state and channel state will become more dynamic, which requires communication system to have higher adaptivity. These requirements will greatly affect the design of multiple access protocols. In this dissertation, under the guidance of the theory of information theory and coding theory, we focus on the design of multiple access protocols in large-scale access scenarios. The content of this dissertation can be summarized as follows: first, we propose a basic framework of a new random non orthogonal multiple access (random non orthogonal multiple access) access, called rate free access. The framework is different from the traditional scheme's central resource allocation. The access point only assigns a random access control function to the user, and the user shares the resources in the pseudo random way to access. The access process of all users can be equivalent to a special form of linear superposition rate free code, and the access point is decode using the belief propagation algorithm with low complexity. In order to further improve the performance of the system, we introduce the idea of spatial coupling into the access framework, classify the resource units and design the mapping relationship between different users to different types of resource units, and propose an enhanced rate free multiple access scheme based on this. In addition, we use the information propagation theory to analyze the proposed scheme, and get the relationship between the reachable rate and the parameters such as the random access control function, the spatial coupling type and the user precoding. Based on these, we design the key parameters of the system jointly and optimally. The analysis results are consistent with the simulation results, which show that the proposed scheme can achieve channel capacity gradually, and has low signaling overhead and high adaptability, which is in line with the needs of large-scale access scenarios. Second, by using the identity information of the access mode in the above framework, we propose a "no authorization" rate free multiple access scheme. The scheme can save the "registration" process of the basic framework, and can reduce the signaling overhead and system delay. In addition, we have further proposed a joint user detection decoding algorithm to achieve channel estimation, user detection and information decoding at the same time with lower complexity. Based on the theory of density evolution, an approximate analysis algorithm is proposed to analyze the performance of the proposed algorithm. The analysis results are consistent with the simulation results, which shows that the proposed scheme only introduces a minimal bit error rate performance loss compared with the "authorized" rate free multiple access, but it can significantly reduce the signaling overhead and delay of the system. In addition, we extend this idea to sparse code multiple access scheme based on fixed signature matrix, and propose a sparse code multiple access scheme with signature sequence skipping. Third, in order to simplify the user pairing of uplink non orthogonal multiple access, we propose a non orthogonal multiple access scheme based on rate segmentation. In the proposed scheme, each user is equivalent to a number of virtual users by dividing the data to be sent into multiple data flows. Even if the power of different users reaches the access point, the access point can still restore the data of each user by successive interference cancellation algorithm, so the scheme can avoid complex user pairing and reduce the scheduling complexity. In addition, we analyze the performance of the proposed scheme, and obtain the closed relation expression between the system interrupt probability and the reachable rate and the key parameters of the system. According to the theoretical analysis, we propose a design method for the key parameters of the system, including the type of rate segmentation and the type of power segmentation. Finally, the analysis results and simulation results show that the proposed scheme has a significant improvement in both the achievable rate and the interruption probability.
【学位授予单位】：浙江大学
【学位级别】：博士
【学位授予年份】：2017
【分类号】：TN929.5

【相似文献】