Design And Implementation Of Lte Prach Sending And Receiving Based On Gpp Platform

With the expansion of broadband data services, as well as the increasing demand of a higher transmission rate, the3GPP (3rd Generation Partnership Project) launched a research project for Universal Mobile Telecommunications System technology evolution (Long Term Evolution, LTE), by introducing new technology such as OFDM (Orthogonal Frequency Division Multiplexing) and Multiple-Input Multiple-Output(MIMO) to it. The target is to achieve a high throughput rate, a maximum downlink speed of100Mbps and uplink speed of50Mbps in20MHz bandwidth. And the random access of the LTE plays a key role in the process of the mobile terminal to establish a radio link with the network. At the same time, the GPP (General Purpose Processor) technic is moreintroduced into the telecommunication fields driven by its rapid development. It also provides a new solution of real-time processing of the LTE baseband signal. Therefore, how to make LTE random access based on the general-purpose processor more efficiently becomes the focus of the research work in this paper.Firstly, analyze the Intel general purpose processor structure and performance and get development and optimization ideas of the program. Secondly, following this train of thought, design the LTE PRACH (Physical Random Access Channel) sending and the receiving blocks in the physical layer algorithm:(1) Analyze the basic structure of the LTE PRACH and the corresponding high-level configuration, and then design the PRACH sending block algorithm. An improved Chirp-z algorithm based on GPP platform is proposed to do DFT (Discrete Fourier Transform operation) At the same time, a low complexity IDFT(Inverse Discrete Fourier Transform) algorithm is designed by taking characters of the preamble signal into account. After that, correctness of each algorithm is proved one by one.(2) Design the PRACH receiving block by dividing it into two major modules:the signal receiving and the energy detection module. After that, design each part of the specific module clearly and definitely, including the front anti-aliasing filters, fast algorithm of related sequences, power detection algorithm, etc. Then obtain the performance data on MATLAB platform by simulating the entire link under both AWGN(Additive White Gaussian Noise) and ETU(Extended Typical Urban) channel conditions. The results show that the false alarm probability meet requirement, and the miss detection probability performance of the PRACH receiving algorithm above is2-3dB better than that specified in3GPP36.104. In the implementation, a large number of C language and SIMD (Single Instruction Multiple Data) instructions optimization were applied based on the targeted design using the above algorithm.At the same time, analyze the optimization process commenced by the VTune analyzer briefly. At last, a series of comprehensive tests are carried out to check the correctness of the program after optimization. The result shows that the false alarm probability meet requirement, and the miss detection probability performance on GPP is also about2dB better than that specified in protocol. From a real-time standpoint, the whole detection process only needs tens of microseconds or even less. It can also be pointed out easily that the real-time signal processing can be satisfied very well.