Handoff Strategy For Satellite-Ground Fusion Network

With the rapid development of mobile satellite communication technology in the21st century, the satellite is playing a more and more important role in our life. Mobile satellite has been utilized in emergency communication because of the advantage of high robust and wide covering range. On the other hand, satellite communication path depends on the line-of-sight transmission, which is suffered from the shadowing and covering of constructions. Therefore, the satellite-ground network has become the hot topic in recent years. The ancillary terrestrial component (ATC) has been widely used in the network, improved the communication quality within the urban area greatly and utilized the spectrum resources to a higher extent. This technology can combine the advantages of satellite’s wide covering range and ground network’s high quality channel and satisfy people’s starving for high speed communication.The mobile satellite system-ancillary terrestrial component, MSS-ATC is composed by two parts, including satellite network and ground network. The relay transmitter of satellite network is GEO satellite and the base station of ground network is the ATC base station. But MSS-ATC network has a long channel delay and high MU velocity, so the network handoff is confronted with a big challenge. Recent years, the network handoff problem has caught widely attention, but most existing research are focus on the ground heterogeneous network and LEO satellite system, so the existing handoff mechanism doesn’t match MSS-ATC system very well. This article did a thorough research about the MSS-ATC handoff problem, and put forward a new handoff decision algorithm and verified it in the simulation platform. The idea can be explained as follows: Firstly, to deal with the satellite’s unique feature of high channel delay and wide MU velocity range, the algorithm considered the user received signal strength and user velocity, and then put forward a handoff decision method based on the Kalman filter and Q-learning algorithm. We made a thorough comparison of channel attenuation rate, handoff times and network revenue between this method and traditional method. The results indicate that our algorithm has gained a high performance improvement and this can adjust to high speed scenarios very well.Secondly, we build a NS2simulation platform to implement the satellite-ground network’s mobility management module. This platform is composed with mobile satellite network and ATC based ground network. The platform implemented the different node structure, network linking and constructing, protocol stack design. In addition, we implement the MU position updating, authentication and encryption, handoff between spotbeam an ATC satellite. This platform is used in the simulation of this article.