Research On Time-Frequency Resource Management For LTE-R Oriented Smart Base Stations

With the development of the high-speed railway industry,the high-speed railway radio communication technology that meets the demands of high communication performance services has become an increasingly important research topic.The fixed allocation of spectrum has been unable to meet the increasing communication demands.At the same time,the characteristics of high-speed movement make wireless communication in the field of high-speed rail face more challenges.Problems such as time-frequency resource management,vehicle body signal loss,Doppler frequency shift,and frequent handover are restricting the development of high-speed railway wireless communications.This article proposes a new solution for high-speed railway time-frequency resource management issues.Cognitive radio technology is a popular solution to time-frequency resource management.Dynamic management of resources is a reasonable solution to the lack of time-frequency resources.It is also the basic idea on which this paper relies.In addition,based on the idea of distributed management model,we will assign the base stations to the recognition,management,and calculation capabilities.It will cooperate with other base stations,trackside equipment,trains,consoles,and other infrastructure along the high-speed railway for information exchange.We define it as a Smart Base Station(SBS).SBS is given a variety of time-frequency resource management methods.The first is a time-frequency resource dynamic allocation method based on Bayesian theory.In the Naive Bayes algorithm,we establish a state-action set so that SBS uses experience to make decisions.We introduce a judgment mechanism to prevent local optimization.Under global perception,we not only enable SBS to gain more prior knowledge of the high-speed railway communication environment,but also consider the feedback of actions on decision making.The results show that Bayesian Learning has great performance in the field of high-speed railway communications,and can reduce the number of communication blocking times and switching times of unauthorized users effectively.The second is a time-frequency resource management method based on probabilistic graphical model.Based on the concept of directed acyclic graph in probabilistic graphical model,this paper defines key nodes for high-speed railway communication and seek the dependencies between nodes.In the simulation experiment,we take throughput,packet loss rate,delay and delay jitter as the standards to measure the performance of the algorithm.SBS makes sequential decisions through exploration,perception,and information exchange.The experimental results show that the improved time-frequency resource allocation scheme based on the probabilistic graphical model can better adapt to the special environment of high-speed railway and is superior to the traditional time-frequency resource management algorithm.The third is a time-frequency resource allocation model based on game theory.This paper studies the spectrum resource sharing theory of high-speed rail cognitive wireless networks under static and dynamic conditions,and proposes a auction model and a bidding model based on different subjects.In the auction model,this article emphasizes the protection of key services such as train control system.In the bidding model,more consideration is given to service priorities.The experimental results show that the game theory can effectively improve the high-speed rail spectrum utilization efficiency and provide a new idea based on the spectrum income balance for the high-speed rail cognitive wireless network.