Coordinated Access And Cache Resource Management Technology In Terrestrial-Satellite Integrated Network

Terrestrial-Satellite Integrated Network(TSIN)coordinates Low Earth Orbit(LEO)satellites with terrestrial communication networks to provide low-latency and highbandwidth access services for terrestrial users,which has also become one of the important development trends of the future communication.In this network architecture,the satellite network can not only provide high-capacity backhaul resources for small-cell base stations(SBSs),but also have the same on-board processing and access capabilities as SBSs.By integrating pre-caching mechanism into TSIN,the overall network performance can be further improved.Compared with the terrestrial communication networks,TSIN breaks the ideal or solidified backhaul capacity constraints.Multiple satellites are allowed to provide SBSs with high-capacity satellite backhaul resources,which makes the resource management needs to consider two levels of the terrestrial network and the satellite-toground network.Besides,the user association has also been converted from one-to-one association to one-to-many association.At the same time,the transmission characteristics of satellite networks and the requirement differences in transmission channel for different user services also need to be fully considered in the research on resource management.Therefore,how to efficiently design resource management techniques such as user association,resource allocation,access selection to improve the network efficiency of TSIN is a critical task.In view of the above reasons and challenges,this article focuses on two typical transmission modes in Terrestrial-Satellite integrated network: Backhaul mode and Regenerative mode.Aiming at maximizing the downlink sum rate of system and the number of accessed ground users,the resource management problems under these two transmission modes are studied to realize the efficient matching and full utilization of network resources.The main contributions and achievements of this paper are as follows:1.Backhaul Mode: We propose a resource management technique by jointly optimizing user association and resource allocation to maximize the downlink sum rate of system and the number of accessed ground users.Lagrangian Relaxation is applied to decompose the resource management problem in this mode into two subproblems: the resource management problem in terrestrial network and the resource management problem in satellite-to-ground network;Then,two low-complexity algorithms are proposed to solve the first subproblem with multi-objectives and the second subproblem with multi-connectivity.Simulation results validate the proposed algorithms can obtain the gain of the system sum rate from 28.5% to 120.7% compared to other algorithms in the typical settings,and simultaneously save the computation time.2.Regenerative Mode: We propose a resource management technique by jointly optimizing user association,resource allocation,user access selection to maximize the downlink sum rate of system while ensuring user service experience.Considering the non-convexity of the resource management optimization problem in this mode and the complex constraints,we propose a Genetic-Based resource management algorithm to obtain the global optimal solution.Simulation results validate the proposed algorithm can obtain the gain of the system sum rate from 25% to 381.6% compared to other algorithms in the typical settings.Furthermore,the proposed algorithm can reach the global optimal solution or the sub-optimal solution in short time.