摘要
卫星网络有效地补充了地面网络,并在空间通信和组网中发挥着重要作用.分析卫星网络的容量可以为提高网络效率、增强业务保障能力以及设计高效通信协议等提供重要依据.然而,现有的地面无线网络容量分析方法和现有的在轨卫星与地面站之间上下行链路容量的分析方法都难以直接应用到位于空间的卫星网络容量分析.文中根据卫星网络的特点,提出了一种卫星网络容量智能分析方法,从网络数据传输能力的角度,分析了卫星网络容量.基于星间物理可见性、接收功率、载噪比、卫星可用资源,建立了星际链路有效性模型和网络容量模型;对分布式遗传算法进行了改进,并应用其求解节点分配带宽和功率到星际链路的优化方案,进而得出卫星网络容量上限.基于STK等进行了仿真实现和性能评价.研究结果表明,节点失效和邻频干扰会导致卫星网络容量下降,星座结构对卫星网络容量影响较大;文中提出的方法具有较高的搜索精度、较低的复杂度、较快的收敛速度和避免早熟的能力,可以有效地计算卫星网络容量上限,性能优于现有的方法.
Satellite network is an effective supplement of terrestrial network and plays an important role in space communication and networking. Its capacity analysis provides important basis for improving network efficiency, enhancing service guarantee ability and designing efficient communication protocol. However, existing capacity analysis methods for terrestrial wireless networks and up/down links among in-orbit satellites and ground stations are difficult to be directly applied to capacity analysis of satellite networks in space. With the characteristics of satellite networks considered, an intelligent analysis method of satellite network capacity is proposed in this paper. It analyzes satellite network capacity from the viewpoint of data transmission ability. Based on the visibility between satellites, satellite receiving power, carrier noise ratio and available resources, the availability model of inter-satellite links and the network capacity model are built. The distributed genetic algorithm is improved and applied to get the optimal satellite bandwidth and power allocation scheme and then the upper bound of the network capacityis obtained. It has been implemented by simulation over the STK. Experimental results have shown that node failure and adjacent channel interference can decrease network capacity while constellation structure has great influence on network capacity the proposed method can efficiently obtain the upper bound of network capacity with high searching precision, low complexity, fast convergence speed and avoiding premature solution, and its performance outperforms the existing methods.
出处
《计算机学报》
EI
CSCD
北大核心
2017年第7期1572-1582,共11页
Chinese Journal of Computers
基金
国家自然科学基金(61572123)
国家杰出青年科学基金(71325002
61225012)
辽宁省百千万人才工程项目(2013921068)
中国科学院科技创新交叉与合作团队计划-面向智能电网安全监控及信息互动的网络与系统技术资助~~
关键词
卫星网络
网络容量
容量分析
分布式遗传算法
satellite network
network capacity
capacity analysis
distributed genetic algorithm
