End-to-end QoS Architecture in Space Internet
- Paper number
IAC-05-B3.P.03
- Author
Mrs. Huang Wei, Beijing University of Posts and Telecommunications, China
- Year
2005
- Abstract
The increasing of space explorations requires space communication protocols to provide more capabilities, such as dynamic routing, adaptive data transformation and automatic resource allocation. Standardization is also need to increase cross-support between different space agencies and decrease development costs. Exploiting IP-based protocols as space communication protocols to construct space Internet is an attractive solution. The space Internet should not only rely on IP for packet transferring, but also exploit IP-based protocols to perform network-provided functions, such as quality of service (QoS). The advantages of applying Internet technologies in space mission have been analyzed in many papers. The practicability has been demonstrated. And modifications of transport layer and application layer protocols to make IP-based solutions fit better to space environment have also been identified. But papers that systematically study the end-to-end QoS architecture in space Internet have not been found. So this paper tries to discuss this issue. Quality-of-service(QoS) is an important concern of space network. In addition to enhance QoS by modifying transport layer or application layer in end systems, some kind of QoS guarantee is also need to be provided by space network. DiffServ and IntServ are two fundamental QoS Models of Internet. Because space missions prefer fine and absolute QoS control on per flow basis, and scalability problem is not crucial in space Internet domain, IntServ is selected as the base model. End-to-end QoS architecture must be based on the mobility management of space Internet. Mobility management is to solve the problem of accessing to spacecraft with unique IP address. Mobile IP protocol which is referenced as macromobility protocol in mobile network research area has been thought to be the solution. However, we would argue that micromobility protocol should be introduced to further solve the problem of mobility during cross-support. There have been some competing micromobility solutions in mobile network research area, such as HAWAII, Cellular IP, and hierarchical IP. In this paper, these micromobility solutions would be compared to select one that is most suitable to space network. Simulation comparison for hand-off performance of with and without micro-mobility solution during cross-support will also be presented in this paper to justify our conclusion. Referencing to the mobile resource reservation protocol in mobile network, and utilizing the fact that the trajectory of spacecraft could be known in advance, an end-to-end QoS architecture based on the mobility management mechanism mentioned above will be presented in this paper. The ability of supporting automatic resource allocation and dynamic data transfer is tried to achieved at least space-ground bandwidth expense. A performance analysis in terms of packet delay and packet loss will be made to justify the proposed architecture.
- Abstract document