联合跳波束与干扰对齐的高通量卫星资源分配机制研究

High-throughput satellites have greatly increased capacity and reduced the cost and opens a new era of satellite communications. However, the dynamic changes of nodes and links in the system, the constraints of satellite resources and processing capacity, and the increasingly serious interference between multi-beam aggravate the contradiction between service transmission demand and transmission capacity. This project based on the idea of "beam hopping + interference alignment", designs the joint design of resource allocation, efficient utilization and non-orthogonal interference suppression, studies the service-driven beam hopping system and the time slicing mechanism of on-demand coverage, designs the on-demand resource allocation and beam-cluster configuration mechanism of satellite to avoids the fragmentation of resource utilization and the uneven busyness of beams; also explores the mechanism of efficient utilization of resources, which combines beam hopping and interference alignment to minimize the interference between beams and approach the link transmission limit from the perspective of multi-user information theory; and proposes an interference alignment precoding algorithm based on manifold optimization theory to make interference alignment strategy changes with the allocation of hopping beam resources in real time. Finally, it meets the requirements of flexible and mobile information services of high-throughput satellites, as well as high reliable and large-capacity transmission.

高通量卫星大幅提升了容量并降低了单位带宽成本，开启了卫星通信新纪元。然而卫星终端用户业务差异性大、类型和需求多样，时间和空间分布严重不均，波束之间的干扰日趋严重，加剧了业务传输需求与传输能力的矛盾。针对以上问题，本项目基于“跳波束+干扰对齐”的思想，将资源调配、高效利用和资源非正交的干扰抑制联合设计，研究业务驱动、随需覆盖的跳波束资源分配机制，提出面向资源全局调度的时间分片方法，设计分簇波束数目规划和分级组合的分簇构型流程，避免了资源利用的碎片化和波束的忙闲不均；研究跳波束和干扰对齐相得益彰的资源高效利用机制，提出最小化干扰子空间的干扰对齐机制，从多用户信息论的角度最大限度的降低波束间的干扰；研究基于流形优化理论的干扰对齐预编码算法，兼顾性能与低复杂度的要求，使得干扰对齐策略随跳波束资源分配实时变化。最终满足高通量卫星灵活机动的信息服务，以及高可靠、大容量传输的需求。

针对高通量GEO卫星系统，用户业务时空二维分布不均性、星地链路动态变化性、电磁环境的复杂性、波束间的同频干扰等问题，本项目的研究工作主要围绕跳波束技术的工作机制、资源分配与波束调度、干扰抑制与预编码设计、基于机器学习的智能跳波束方法等方面展开。主要包括：研究业务驱动的跳波束时间分片工作机制，建立跳波束时隙分配模型、并对波束跳变图案和跳波束工作流程进行设计；研究面向负载均衡的动态分簇与弹性波位设计，达到负载均衡和资源灵活利用的目标，并降低系统时延和波束管理的复杂度；研究跳波束系统的用户切换策略，从系统资源动态预留与调度的角度，解决跳波束系统中高速运动用户的无感切换问题；针对星地动态链路条件，研究面向时变雨衰的跳波束时隙分配方法，以获取更高的频谱效率和系统容量；针对密集波束之间的同频干扰，引入干扰对齐等预编码方法，并与跳波束进行联合设计，对干扰进行有效抑制；引入机器学习的人工智能方法，研究基于深度强化学习的跳波束方法，以实现复杂环境下的资源和波束的智能调度。通过对本项目的研究，切实提高跳波束系统的资源利用和系统性能，推动下一代高通量卫星的技术发展，促进我国卫星互联网工程的建设和实施。