Research On Key Technologies Of Access Optimization Based On NB-IOT

With the development of modern communication technologies in the Internet of Things,the number of Internet of Things devices has increased dramatically.Among them,low-rate services have unique features of a large number of devices,wide distribution,and extremely low energy consumption.The third Generation Parnership Proj ect(3GPP)designed a dedicated narrowband Internet of Things(NB-IoT)network.However,when the IoT device access in large quantities at the same time,the base station cannot process in time and effectively,resulting in network congestion.In this paper,we study the NB-IoT communication access technology and optimize its access process and access control process to meet the service requirements of NB-IoT devices.It has significance reference for the development of IoT technology based on the cellular network.This paper firstly conducts an in-depth study on the NB-IoT random access protocol and analyzes the NB-IoT random access process in detail with the channel resources.For the NB-IoT protocol's proprietary access features,the user coverage level transition and connection into the process slotted for analysis and modeling.The theoretical analysis of system performance such as the probability of collision,the probability of access,and average access delay.The correctness verified by numerical simulation.Secondly,this paper optimizes the NB-IoT random access process from the application requirements.We propose an access optimization method based on the backoff mechanism by making access control requests for the backoff time and allocatable resources of different delay sensitive devices.Compared with the traditional fixed slot backoff method,this method can allocate random access resources reasonably and improve the overall system performance.The simulation results show that the proposed method can improve the probability of access.Finally,this paper studies the NB-IoT random access loading control algorithm for a large number of access congestion problems.We propose a load access control algorithm based on reinforcement learning,which enables the base station to dynamically learn and adjust the access level limit parameter value as the system congestion state changes.The simulation results show that the proposed algorithm can make the system converge quickly and increase the number of access devices.In the future work,we can study the problem of different retransmission time resources and preamble resources for NB-IoT devices with different coverage levels.The dynamic allocation of resources for different service scenarios enables the system to support more devices..We can consider the use of continuous reinforcement learning algorithm to improve the probability of access.