Ground Station Network to Improve Operation Efficiency of Small Satellites and Its Operation Scheduling Method
- Paper number
IAC-06-C1.6.10
- Author
Dr. Yuya Nakamura, University of Tokyo, Japan
- Coauthor
Prof. Shinichi Nakasuka, University of Tokyo, Japan
- Year
2006
- Abstract
This paper will describe how we manage and solve an operation scheduling problem of multiple small satellites, using clustered ground stations connected one another with the Internet (Ground Station Network). Recently universities and venture companies all over the world have been engaged in developing small satellites. A "CubeSat," which is a nano-satellite with its mass of only 1kg, developed mainly by university students, is its good example. About 10 CubeSats were launched into orbit up to now, and some of them have been working perfectly over 2 years. What one small satellite can do is limited. However, if we have multiple small satellites in orbit, we may realize the same functional capabilities as a large satellite with them by having them work cooperatively. Small satellites have many advantages such as cost-effectiveness, easiness and promptness of development. On the other hand, it is also the fact that small satellites have disadvantages like low communication performance. One solution to overcome this disadvantage is to use multiple ground stations distributed throughout the world. Visible time at one station is surely short, but operations at multiple ground stations drastically improve communication performance of a small satellite as a whole. Ground Station Network (GSN) is based on this concept. An operator of a certain satellite (must be a member of GSN) can access a remote ground station participating in GSN through the Internet, and can make an operation even if the satellite is at the other side of the world. GSN will make operations of small satellites much more efficient. However, for all participants to enjoy the merit of GSN, we must deal with a complicated scheduling problem. For example, different two operators may want to use the same ground station at the same time. We have many things to be considered. A certain ground station may be under maintenance over a period of time. An operator may change his/her operation plan suddenly. In case of an awful natural disaster, a top-priority mission of image acquisition should be conducted by a remote-sensing satellite. The circumstances are changing every minute, and we have to take them into consideration when allocating an operation request given by a user (GSN participant) to a certain ground station. To solve this scheduling problem, we propose a new type of optimization method. As a multivariable optimization method, the Lagrangian decomposition method has recently gathered attention. We advance this method using concepts of potential, so that each task of a given mission (e.g. image acquisition, downlink) can, by itself, select the satellite to take a picture and the ground station to perform a downlink, considering imposed constraints and its priority. What is important here in this method is that selfish actions by each task agent eventually lead to a minimization of an evaluation function, that is, a total optimization. This means that if we can use a fast parallel computer (each task corresponds to a single computer node), we may obtain an optimized result in a very short period. This characteristic is advantageous to our GSN application, where the rescheduling is frequently required.
- Abstract document
- Manuscript document
IAC-06-C1.6.10.pdf (🔒 authorized access only).
To get the manuscript, please contact IAF Secretariat.