Transplant And Performance Optimization Of LTE Baseband System For Centralized Cellular

With the rapid increase in mobile data demand,facing wireless data needs,the traditional cellular network is facing unprecedented challenges in the transmission rate,the spectrum of resources and energy consumption,the bottleneck is more obvious.So the centralized cellular network architecture came into being,its large capacity,low power consumption,intelligent control and other characteristics can enhance the transmission rate,improve the spectrum utilization,conserve energy thereby to achieve green communications.Super base station,its overall program has been recognized in many ways.LTE physical layer,as a key part of re-definable multi-mode baseband pool of super base station architecture,its stable and efficient run has become a key issue.This paper first introduces LTE physical layer,channel processing flow and physical layer process,according to these designs,the physical layer control system is designed to complete the interface with the upper layer and control the uplink and downlink channel.On the basis of ICT-BBUH1000 and LTE-C model,at the system level,we design and implement the overall software architecture of the LTE physical layer,mix together integral structure and physical layer control system to get through links so that to complete the normal video data transmission.TD-LTE has high requirements of timing.Physical layer processing,as a time-consuming part,its performance directly affects the system.Therefore,performance optimization is of great importance.For performance optimization,first,we do the overall modularization and parallelization,analyze the part can be done in parallel,process unrelated module in parallel advantages of multicore DSP.Then we do the optimization in algorithm level and assembler instruction level of time-consuming modules.The paper introduces time-consuming modules such as de-scrambling and demodulation,and analyzes fine algorithms currently,does the code optimization in algorithm level and assembly instruction optimization of the repeated operation.The optimizations reduce the time-consuming and improve the system performance.