Optimization And Implementation Of Radio Link Control Protocol Of Narrowband Internet Of Things

With the development of information society,Narrowband Internet of things(NB-Io T)technology,as one of the low-power WAN technologies,has good potential of commercial application.The sliding window transmission mechanism of radio link control(RLC)protocol,a sub layer of NBIo T protocol stack,can not effectively adapt to the changes of network environment.When the network congestion occurs,the fixed parameters of RLC protocol may worsen the network condition,leading to degradation of system performance and resource waste.In order to promote the application of NB-Io T technology in low delay and high accuracy scenarios,the optimization of RLC protocol in NB-Io T is worth exploring.Machine learning algorithm and genetic algorithm are combined to optimize and implement RLC layer of NB-Io T protocol in this thesis.RLC layer of NB-Io T protocol is designed and modeled in this thesis.Under different network conditions,data sets of different parameter configurations and network performance of RLC layer are collected and used into the training of machine learning polynomial regression algorithm.By fitting the RLC data transmission process with regression model of quadratic polynomial with five variables,the network performance,network status and protocol parameters are expressed as quantitative mathematical expressions,and the complex relationship between the elements is clarified.By adding data transmission rate to RLC protocol to control the amount of network data,the efficiency of network data processing is improved.In addition,the RLC transmission model function is designed to work with genetic algorithm.The best RLC transmission parameters are calculated and configured when the network is congested,so as to improve the data processing efficiency of RLC layer.By designing the optimal parameter configuration package in the receiving side of RLC entity,the data transmission optimization process in terminal is simplified,and the resource consumption can be reduced.Finally,the standard RLC protocol function is implemented based on Free RTOS operating system,and the RLC congestion control optimization algorithm is implemented and tested under different network conditions.The test environment is built based on STM32 microprocessor and MATLAB software platform.The standard function and optimization algorithm of RLC protocol are realized in STM32 microprocessor,and the network environment is simulated by MATLAB.The test results show that:the function of the RLC layer implemented in this thesis conforms to the 3GPP protocol standard,and compared with the standard protocol,the average data processing delay of the optimized RLC protocol is reduced by about 32.16%,and the average throughput is increased by about 23.23%,which achieves the expected design index.