随着智能终端的普及，人们对多媒体业务的需求正急速增长，无线视频数据量在移动业务总数据量中所占的比重越来越高。为了缓解频谱资源短缺压力，信源分层技术和基于广播组播思想的下行多址接入技术成为了新一代通信系统提升频谱效率的重要手段。本文围绕“高谱效的无线资源分配与调度”这一主题，在网络信息论和最优化理论的指导下，结合对分层广播系统特征和非正交多址接入技术的深入分析，面向四种应用场景循序渐进地展开研究。 首先，退化广播信道的容量域上界可以通过发射端叠加编码和接收端串行干扰消除解码来实现。基于这一理想假设，本文研究了无反馈信道下分层高斯信源的各层发射速率、解码阈值与资源分配机制对系统效益的影响。利用凸优化工具，本文依次在发射速率受限与不受限的约束条件下分析了优化问题拉格朗日强弱对偶性质，并提出了最大化系统效益的资源分配算法。该算法高度抽象且通用性强，适用于多种信道分布、多种效用函数定义以及任意发射层数。其次，引入单位信源数据块的最大容忍时延，本文针对时延敏感的业务场景研究了分层广播组播系统的最佳资源分配机制。优化目标的定义是，在满足所有用户成功解码基本层的同时，为增强层选择合理的覆盖范围和发射参数，以最大化系统总收益。本文结合机会组播调度的思想，将FEC编码码率等价转换为均值意义下的用户调度比例，然后分别分析了同构网和异构网中的单层成功接收概率，通过迭代重复单层传输的优化过程提出了完整的多层优化算法。仿真结果表明，该算法在显著提升系统总效益的同时实现了多用户分集增益与组播增益的折中。 接着，本文考虑实际系统设计的复杂度限制，将分层传输与非正交多址接入技术相结合，在特定的星座映射和编码调制模式下研究面向差异化服务的资源分配方案。对于单小区传输，本文重点讨论了层分复用技术与比特分割复用技术的分层覆盖性能；对于多小区单频网传输，本文利用OFDM帧的保护间隔，通过联合优化多基站的发射参数进一步提升协同覆盖效果。 最后，针对多路分层业务的接入场景，本文提出了一种混合下行多址接入策略，不同业务以正交多址接入方式传输，业务内的不同分层以非正交多址接入方式传输，并简单讨论了典型场景下的用户分组原则，分析了组内和组间的资源分配算法。仿真结果验证了该策略在性能损失尽量小的前提下降低了接收复杂度。

With the popularity of smart terminals, the demand for multimedia services is growing rapidly, which results in an explosive amount of wireless video data. In order to alleviate the shortage of spectrum resources, the idea of layered transmission and downlink multiple access technology have attracted more and more attention in the new-generation communication systems. Under the guidance of network information theory and optimization theory, this paper analyzes the characteristics of layered broadcasting systems and non-orthogonal multiple access. With the topic of "high-spectrum-efficiency wireless resource allocation and scheduling", the research is then conducted step by step under four application scenarios.First, the upper bound of the capacity region of degraded broadcast channel can be achieved by combining superposition coding at the transmitter side and successive interference cancellation decoding at the receiver side. Based on this ideal assumption, this paper studies the influence of transmission rate, SNR threshold, and resource allocation of each layer on the overall system performance. The Lagrangian duality of the optimization problem is analyzed via convex optimization tools under the constraints of limited and unlimited rates, respectively. A resource allocation algorithm is accordingly proposed to maximize a predefined utility function. This algorithm is highly abstract and is suitable for all kinds of channel distributions, as well as arbitrary number of transmission layers.Secondly, considering the maximum tolerable delay of each data block, this paper studies the optimal resource allocation scheme for delay-sensitive layered broadcasting and multicasting scenarios. The objective function is defined as follows. On one hand, all users should successfully decode the base layer using the least amount of channel resources. On the other hand, reasonable transmission parameters should be selected for enhancement layers to maximize the block-level utilities. Inspired by the idea of opportunistic multicast scheduling, this paper transforms FEC code rate into the average user selection ratio. The successful reception probability under homogeneous and heterogeneous networks is then analyzed. By iteratively repeating the optimization process of single layer, this paper proposes a complete joint optimization algorithm for multiple layers. Simulation results verify that the algorithm achieves a trade-off between multi-user diversity gain and multicast gain at the same time improves the overall utilities significantly.Next, this paper combines layered transmission with actual system design, such as specific constellation mapping and modulation and coding scheme, and studies corresponding resource allocation strategies for differentiated quality of services. In single-cell transmission, this paper focuses on overcoming coverage mismatch under layer division multiplexing (LDM) and bit division multiplexing (BDM). In multi-cell transmission, this paper uses the guard interval of OFDM frames to jointly optimize transmission parameters of multiple base stations to further improve the collaborative coverage.Finally, for the scenario with two or more services, this paper proposes a hybrid downlink multiple access strategy, in which different services occupy orthogonal channel resources, and different layers in the same service occupy non-orthogonal channel resources. The user grouping principles are briefly discussed, as well as capacity analysis within and between groups. Simulation results show that the strategy reduces reception complexity with little loss of performance.