CDMR: Effective Computing-Dependent Multi-Path Routing Strategies in Satellite and Terrestrial Integrated Networks
Chen Wang, Zhiyuan Ren, Wenchi Cheng, Hailin Zhang
Abstract
With the development of science and technology, Satellite and Terrestrial Integrated Network (STIN) can effectively make up for the defects in the terrestrial network coverage and provide robust end-to-end transmission and computing supports for users in many communication scenarios such as emergency rescue, space exploration, and so on. However, many researchers only studied transmission or computing, ignoring that the task can be computed in task transmission. And limited by the satellite's battery resources and high dynamics, it is challenging to realize the fusion of communication and computing with the constraint of energy consumption in the high dynamic satellite network. Therefore, this paper proposes a Computing-Dependent Multi-path Routing (CDMR) paradigm to address such a challenge. In CDMR, the task is mapped to the multi-path route in STIN, and the energy consumption is taken as the optimization goal. Furthermore, the processing latency is the constraint to ensure that the optimal CDMR strategy is under latency constraint. Finally, the simulation results show that the CDMR proposed in this paper can effectively compute the task during the end-to-end transmission. Compared with other computing models, CDMR has more evident advantages in energy saving.