Learning- and optimization-based channel estimation for cognitive high-speed rail broadband wireless communications

In recent years, there is an ever-growing demand on high-quality broadband wireless communications (BWC) for offering high-quality multimedia information services (such as voice, Internet, and video) to passengers as well as improving the safety, security, and operational efficiency of high-speed rail (HSR) transportation. One of the most challenging issues is fast yet accurate channel estimation in HSR mobile environment, where various HSR scenarios and Doppler spread effects have to be considered. Considering the repetitive movement nature of high-speed train over a pre-determined course, a novel learning- and optimization-based channel estimation approach is proposed for cognitive HSR BWC systems in this paper. The key idea is to treat the channel estimation as a learning and optimization process, in other words, the HSR channel parameters are continuously fine-tuned while the train moves along the high-speed rail repetitively. In addition, the learning and optimization process is to be implemented offline, therefore, the computation time of the optimization algorithm (such as the genetic algorithm) is not a limiting issue for real-world implementation. The simulation results demonstrated the effectiveness and significant advantages of this cognitive approach to HSR channel estimation.