集约型时间触发以太网拓扑优化方法

Intensive Time-Triggered Ethernet Topology Optimization

时间触发以太网(time-triggered ethernet,TTE)是一种时间业务与事件业务混合技术,其在兼容标准以太网基础上,保障时间触发(time-triggered,TT)业务的确定性和可靠性,在航天通信领域具有显著的优势.但由于航天网络的高成本问题,在很大程度上限制了TTE的发展,因此研究集约型网络拓扑可降低系统架构成本.本文首先引入复杂网络理论,在满足速率约束(rate-constrained,RC)消息延迟界门限的前提下,建立简化网络拓扑结构的问题模型.其次提出最小介数节点删除及混合策略恢复的集约型拓扑优化算法,通过对节点和边的交替迭代收敛到目标函数最小化的状态.最后,在优化后的集约型网络基础上设计双冗余结构,提高时间触发业务的可靠性.实例的优化前后结果表明,不同恢复策略在不同网络结构下性能不同,说明所提出混合策略的优势.通过优化目标函数对比表明所提出的方法能兼顾时延的前提下简化网络拓扑结构,降低架构成本.本研究给出了基于TTE的实时航天通信的设备部署优化算法及恢复策略,可用于旧网络的升级与改造.通过实例给出了在随机的BA无标度网络中,实现模型建立和拓扑优化的过程,其算法思路和设计方法可进一步应用于更大规模的网络拓扑优化,为航天通信网络优化提供了参考.

Time-triggered ethernet(TTE)is a new technology of ethernet network communication,combining timetriggered and event-triggered traffic.This technology is compatible with the standard ethernet protocol,guarantees the certainty and reliability of time-triggered traffic,and has significant advantages in aerospace communication applications.However,the high cost of space networks limited the development of TTE to a great extent.Therefore,the study of intensive network topology can reduce the cost of system architecture.First,this paper introduced the complex network theory and established the problem model of a simplified network topology through a rateconstrained message delay threshold.Then,a topology optimization algorithm based on the deletion of the minimum number of nodes and mixed strategy recovery was proposed,converging to a state where the objective function was minimized using alternating iterations of nodes and edges.Finally,a dual redundancy structure was designed based on the optimized intensive network,improving the reliability of the time-triggered business.The optimization of objective function comparison shows that the proposed method can simplify the network topology and reduce the architecture cost when considering the time delay.This research provides the optimization algorithm and recovery strategy of real-time aerospace communication equipment deployment based on TTE.The algorithm and the strategy can be used to upgrade and optimize the old network deployment.The modeling and topology optimization process in a random Barabási-Albert scale-free network is explained by examples.The algorithm idea and the design method can be further applied to network topology optimization on a larger scale,which provides references for space communication network optimization.