多层卫星网络QoS路由协议研究
发布时间：2018-08-28 07:52
【摘要】：卫星通信技术作为当今世上最先进的通信技术之一,其覆盖面积大、通信距离远等优点使其在各个领域发挥着举足轻重的地位。卫星技术的不断发展以及新兴应用的出现最终导致了多层卫星网络结构逐渐取代单层卫星网络结构,并相继带来一些网络中需要研究的新的问题。其中由于多层卫星网络结构不规则,连接复杂,星间链路有间歇性通断,网络拓扑动态变化频繁等特点,多层卫星网络路由协议的研究问题受到很多研究者的青睐。由于空间信息网络要求有一定的Qo S保证,卫星网络作为其核心网络,必不可少要研究它的Qo S路由协议。因此本论文主要针对多层卫星网络的Qo S路由协议进行研究。本文首先分析了课题背景以及研究的意义,着重分析在卫星移动通信中研究多层网络Qo S路由协议的重要性及卫星网络路由协议的国内外研究现状,发现当下多层网络的路由算法研究属于卫星通信的热点问题,尤其是很多业务分类的Qo S路由算法的研究相对较少。针对这种现象,本文首先研究移动用户位置管理策略确保移动用户位置可知,为后文路由算法研究奠定基础。其次针对当前GEO/LEO双层卫星网络中连接关系频繁变化造成链路频繁切换;以往网络的分层管理中,一颗GEO卫星与所有与其可视的LEO卫星之间都有链路,连接关系较为复杂;解决网络拥塞的重路由方法的开销较大等问题,本文提出一种基于族首管理的GEO/LEO双层卫星网络Qo S路由策略。该策略提出一种MSAP算法,充分利用算法的可离线计算的属性,使用该算法提前计算备选路径,实现负载均衡,缓解网络中的拥塞。然后针对MEO/LEO双层网络中拓扑更新频繁且规律差;解决网络拥塞、保证业务传输带宽的重路由方法的开销较大;不能充分利用多层网络的优势互补来提高不同业务的Qo S满意度等问题,本文提出一种基于蜂群位置优化的MEO/LEO双层卫星网络Qo S路由策略,该策略提出一种改进的BMDP算法,即M-BMDP算法,有效缓解网络拥塞问题,且更好的为不同类型业务提供Qo S服务,时延、丢包率特性都有很好的改善,且解决了重路由的延迟问题,降低了重路由开销。最后考虑地面接入业务对网络业务传输的影响,在MEO/LEO双层网络中研究一种基于流量预测的MEO/LEO双层卫星网络Qo S路由策略,通过神经网络预测地面接入业务流量模型,避免网络中由于地面流量接入可能引起的拥塞问题,减少了星间传输业务的丢包率。
[Abstract]:As one of the most advanced communication technologies in the world, satellite communication technology plays an important role in various fields because of its large coverage area and long communication distance. With the development of satellite technology and the emergence of new applications, the multi-layer satellite network structure is gradually replacing the single-layer satellite network structure, and some new problems need to be studied in the network. Due to the irregular structure of multi-layer satellite network, complex connection, intermittent intersatellite links, frequent dynamic changes of network topology, and so on, the research of multi-layer satellite network routing protocol is favored by many researchers. As the space information network requires certain Qo S guarantee, satellite network as its core network, it is necessary to study its Qo S routing protocol. Therefore, this thesis mainly focuses on the Qo S routing protocol of multi-layer satellite networks. In this paper, the background and significance of the research are analyzed firstly, and the importance of studying Qo S routing protocol in satellite mobile communication and the current research status of satellite network routing protocol at home and abroad are analyzed. It is found that the current research on routing algorithms in multilayer networks is a hot issue in satellite communication, especially in many traffic classification Qo S routing algorithms. In view of this phenomenon, this paper first studies the mobile user location management strategy to ensure the mobile user location knowable, which lays the foundation for the future research of routing algorithm. Secondly, link switching is caused by the frequent change of connection relationship in the current GEO/LEO double-layer satellite network. In the previous hierarchical management of the network, there is a link between a GEO satellite and all its visible LEO satellites, so the connection relationship is more complex. To solve the problem that the rerouting method of network congestion is expensive, this paper proposes a Qo S routing strategy for GEO/LEO double-layer satellite networks based on clan head management. This strategy proposes a MSAP algorithm, which makes full use of the off-line computable properties of the algorithm, calculates alternative paths ahead of time, realizes load balancing and alleviates congestion in the network. Then aiming at the frequent and irregular topology update in MEO/LEO double-layer network, the rerouting method, which solves the network congestion and guarantees the traffic bandwidth, has a large overhead. It is not possible to make full use of the complementary advantages of multi-layer networks to improve the Qo S satisfaction of different services. In this paper, a Qo S routing strategy based on hive location optimization for MEO/LEO double-layer satellite networks is proposed, and an improved BMDP algorithm is proposed. That is, M-BMDP algorithm can effectively alleviate the network congestion problem, and better provide Qo S services for different types of services. The characteristics of delay and packet loss rate are all improved, and the delay problem of rerouting is solved and the rerouting overhead is reduced. Finally, considering the influence of terrestrial access service on network traffic transmission, a MEO/LEO double-layer satellite network Qo S routing strategy based on traffic prediction is studied in MEO/LEO bilevel network, and the traffic model of ground access service is predicted by neural network. In order to avoid the congestion problem caused by ground traffic access in the network, the packet loss rate of inter-satellite transmission services is reduced.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TN927.2
[Abstract]:As one of the most advanced communication technologies in the world, satellite communication technology plays an important role in various fields because of its large coverage area and long communication distance. With the development of satellite technology and the emergence of new applications, the multi-layer satellite network structure is gradually replacing the single-layer satellite network structure, and some new problems need to be studied in the network. Due to the irregular structure of multi-layer satellite network, complex connection, intermittent intersatellite links, frequent dynamic changes of network topology, and so on, the research of multi-layer satellite network routing protocol is favored by many researchers. As the space information network requires certain Qo S guarantee, satellite network as its core network, it is necessary to study its Qo S routing protocol. Therefore, this thesis mainly focuses on the Qo S routing protocol of multi-layer satellite networks. In this paper, the background and significance of the research are analyzed firstly, and the importance of studying Qo S routing protocol in satellite mobile communication and the current research status of satellite network routing protocol at home and abroad are analyzed. It is found that the current research on routing algorithms in multilayer networks is a hot issue in satellite communication, especially in many traffic classification Qo S routing algorithms. In view of this phenomenon, this paper first studies the mobile user location management strategy to ensure the mobile user location knowable, which lays the foundation for the future research of routing algorithm. Secondly, link switching is caused by the frequent change of connection relationship in the current GEO/LEO double-layer satellite network. In the previous hierarchical management of the network, there is a link between a GEO satellite and all its visible LEO satellites, so the connection relationship is more complex. To solve the problem that the rerouting method of network congestion is expensive, this paper proposes a Qo S routing strategy for GEO/LEO double-layer satellite networks based on clan head management. This strategy proposes a MSAP algorithm, which makes full use of the off-line computable properties of the algorithm, calculates alternative paths ahead of time, realizes load balancing and alleviates congestion in the network. Then aiming at the frequent and irregular topology update in MEO/LEO double-layer network, the rerouting method, which solves the network congestion and guarantees the traffic bandwidth, has a large overhead. It is not possible to make full use of the complementary advantages of multi-layer networks to improve the Qo S satisfaction of different services. In this paper, a Qo S routing strategy based on hive location optimization for MEO/LEO double-layer satellite networks is proposed, and an improved BMDP algorithm is proposed. That is, M-BMDP algorithm can effectively alleviate the network congestion problem, and better provide Qo S services for different types of services. The characteristics of delay and packet loss rate are all improved, and the delay problem of rerouting is solved and the rerouting overhead is reduced. Finally, considering the influence of terrestrial access service on network traffic transmission, a MEO/LEO double-layer satellite network Qo S routing strategy based on traffic prediction is studied in MEO/LEO bilevel network, and the traffic model of ground access service is predicted by neural network. In order to avoid the congestion problem caused by ground traffic access in the network, the packet loss rate of inter-satellite transmission services is reduced.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TN927.2
【相似文献】
相关期刊论文 前10条
1 毛腾跃;徐正全;朱容波;侯睿;;下一代卫星网络发展研究[J];电信科学;2012年06期
2 杜宗n，
本文编号：2208760
本文链接：https://www.wllwen.com/kejilunwen/xinxigongchenglunwen/2208760.html
