基于网络编码的光层组播最大吞吐量理论与方法研究

It is a key and difficult problem to maximize multicast switching and routing throughput all along in the all-optical network layer in the future Info-communication network. The contents of project research focus on the problem of maximizing multicast throughput theory and method for optical multicast communication network. The research contents include but not limited to: .① The way to construct network code and the scheme to resolve multicast contention are studied for optical multicast switching node. Moreover, the network coding system and information flow scheduling strategy are optimized by the enhanced graph theory and information flow disassembling method. Through this research, we want to expound the relation between network code and maximal throughput for optical multicast. Also, we want to extend the information delivery theory to optical multicast minimize code cost and maximize network throughput..② The maximum flow and minimum cut-set theorem and limit multicast information flow transmission are researched to expound the routing cost model considering network coding switching node and network coding link. We expect to find the laws between maximum information flow limit transmission and minimum optical multicast routing cost. Furthermore, the multicast routing strategy based-on minimal cost-constrained are probed by improved genetic algorithm..③ Considering the optical multicast traffic dynamic arriving and dynamic outgoing, the relations between network throughput performance and the method of network code node selection. The ways to select super network code node in the multi-source multi-homing optical multicast network are researched in the project. The methods of super network coding node selection are optimized to improve the optical multicasting network throughput.. Through the research of this project, we want to extend the theory of information delivery for optical multicast network based-on network coding, to expound the theorem of the maximum flow and minimum cut-set, and to resolve the difficulty problem of maximum throughput in network layer for optical multicast.

在光层实现组播信息交换和路由的最大吞吐量是未来信息通信网络面临的重点和难点问题。本项目紧紧围绕光网络通信领域有关组播最大吞吐量的理论与方法问题展开：①研究光层组播交换节点网络编码构造方法与冲突解决机制，基于增强图论法和流分解法优化网络编码系统与信息流调度策略，阐明光组播网络编码与最大吞吐量之间的关系，扩展光组播最小代价最大吞吐量编码信息分发理论。②研究最大流最小割定理及网络编码的组播信息流极限传输理论，探索光网络编码节点和编码链路的路由代价描述模型与最大流极限信息传输之间的规律，并基于改进遗传算法改进最小代价约束的光层组播路由。③研究光层组播业务动态到达和撤离复杂情况下的网络编码节点选择方法与光组播业务疏导性能之间的关系，研究多源多宿光组播网络中超级编码节点的选择方法，并优化节点编码与吞吐量。通过本项目研究，将扩展光网络编码组播信息流分发理论和最大流最小割定理，解决光层组播最大吞吐量难

组播吞吐量的最大化是通信网络一个基础而且困难的网络理论问题，光层组播吞吐量的最大化更是一个具有重要研究意义且提高未来光网络资源利用率和保证未来网络服务性能的关键支撑技术。本项目“基于网络编码的光层组播最大吞吐量理论与方法研究（NSFC 61275077）”紧紧围绕光网络通信领域有关组播最大吞吐量的理论与方法展开，主要研究内容包括：（1）研究光层组播交换节点架构和结合网络编码方法的冲突解决机制、交换节点的信息分发和信息流的调度策略，实现光交换节点的最大吞吐量，降低数据丢包率。发表相关的研究论文9篇（其中SCI检索4篇，EI检索8篇）；申请发明专利2项，授权1项。基于SPN-LMFRC光多播节点结构，提出的最大双向匹配最小权重匹配调度算法选择多波长转换器解决组播调度算法；设计了一种反馈共享有限范围多波长转换器和输出共享的基于逻辑运算的全光网络编码器结合的光组播节点结构，并提出一种基于改进的最大权重独立集算法的波长分配方法解决光组播调度策略中最小化波长转换度的组播调度方法及编码机制；设计了一种反馈共享有限范围多波长转换器和输出共享的基于逻辑运算的全光网络编码器结合光组播节点结构，设计了最小波长总转换度的多播冲突解决策略；设计一种输出共享的全光OFDM（正交频分复用）网络编码器结合反馈共享的光纤延迟线的弹性光交换节点结构，提出最大权重独立集信息流调度分发策略。.（2）基于最大流最小割定理及网络编码组播信息流极限传输理论，研究了最小光网络编码代价的光层组播路由和业务疏导传输方法，实现组播传输的最大吞吐量和降低带宽阻塞率。发表研究论文22篇（其中SCI检索5篇，EI检索２０篇）；申请发明专利13项，授权7项。研究光组播树中心编码节点优化选择问题，提出了组播路由中图压缩方法确定最小网络编码代价、节点最少数目的方法；为了降低光组播路由的光域网络编码代价和提高达到理论最大光组播容量的概率，提出一种基于共享链路和网络编码的优化光组播容量路由策略；提出一种改进的遗传模拟退火算法优化编码传输链路；提出了基于改进遗传算法的最小化光组播编码链路算法；提出一种基于网络编码的多核组播路由算法（MCMR-NC），减小目的节点数量对网络编码子图大小的影响；研究一种基于改进NSGA-Ⅱ的光组播编码-链路代价折中方法；提出了一种基于网络编码的最小代价组播树疏导算法。