Beam hopping for massive MIMO LEO satellites : modeling and evaluation

Abstract

Low Earth orbit (LEO) satellite communication systems have become economically feasible in recent years due to the development of cost-e cient satellite launching technologies. This development renewed research interest in the satellite communications field. One technique that is in use in existing satellite systems is beam hopping. There, a multi-beam satellite activates and deactivates beams on a short time scale to serve the user demand by activating beams while avoiding unnecessary interference and resource consumption by deactivating beams. This thesis aims to evaluate the performance of beam hopping schemes in LEO satellite systems. For this, an existing scheme from a system operating in geosynchronous equatorial orbit is considered, as well as a newly developed scheme and three benchmark schemes. To evaluate the performance of beam hopping schemes, a reliable system model is needed. Therefore, the accuracy of existing fading models is evaluated through comparison with measurement data. Then, a LEO satellite system model for a multi-beam satellite operating in the Kaband is developed. The developed model includes a power consumption model of the satellite, which allows the evaluation of energy efficiency in a LEO satellite system. After evaluating the system behavior at different satellite elevation angles, the proposed beam hopping schemes are compared to three benchmark schemes. Evaluation of large-scale fading models shows that statistical models can closely reproduce the network behavior, while models proposed by standardization bodies strike a good trade-off between computational complexity and link modeling accuracy. These results motivate the use of models from standardization bodies for the subsequent satellite simulation. Evaluation of the LEO satellite system shows that the satellite elevation angle strongly influences the system behavior and thus needs to be considered in LEO satellite system modeling and design. Evaluation of the beam hopping patterns shows that the proposed simple greedy approach outperforms the complex optimal approach that uses inaccurate assumptions to simplify the optimization problem.