Research On Key Technologies Of Asynchronous CRA Random Access System In Satellite Networks

With the development of the communication technologies in recent years,the communication services have been gradually diversified,which has brought great benefits to people's production and life.Satellite,as a relay node with wide coverage and not affected by ground environment factors,overcomes the shortcomings of short communication distance,fixed communication range and high cost in terrestrial cellular network.Its simple structure and flexibility make satellite communication technology become the current focus of research.On the other hand,as an important application scenario in the fifth generation mobile communication,the Internet of Things is widely used in smart home,forest fire prevention,ship monitoring and so on.The performance of satellite access technology is directly related to the quality of service of the whole satellite communication system.Due to the limited resources of satellite upstream channel,a large number of users request random access at the same time,which results in data loss and low utilization of satellite channel resources.In this paper,the high throughput and low delay access schemes of M2M(Machine to Machine)communication in the satellite Internet of Things are studied from the point of Random Satellite access.The main contributions are summarized as follows:1)Based on Contention Resolution ALOHA,two improved access schemes,power-graded CRA and piecewise CRA,are presented respectively.Firstly,several improved asynchronous random access schemes based on ALOHA are introduced,and the CRA and ECRA access schemes are mainly expounded,and their performances are compared by computer simulation,and the existing problems are analyzed.Then,on the basis of CRA access scheme,the improved power-graded CRA based on power domain and the piecewise CRA access scheme based on fragmentation reorganization are proposed respectively,and the specific implementation schemes are also given.The fragmented CRA is suitable for the situation that the channel estimation of a single pilot is inaccurate due to long-packet length and bad channel environment.Finally,the performance of the presented access scheme is verified by computer simulation.2)Based on Q-learning algorithm,a QECRA access scheme is presented.Firstly,the problem of channel resource waste caused by random selection of data packet sending location in asynchronous random access schemes is analyzed.Then,combining with Q learning algorithm in reinforcement learning,a QECRA access scheme is presented.Aiming at the problem of large delay in satellite communication,users access the data frames in odd frames(even frames),divide the data frames into multiple uniform location intervals,and users learn the location intervals until users tend to choose the fixed position intervals to send data packets.Finally,the throughput and packet loss ratio of the presented scheme in three environments are simulated by computer,which proves that the presented scheme has good performance.