Research And Implementation Of Key Technology For Uplink Channel In Broadband Cognitive Communication System

This paper mainly studies the implementation of the physical layer transmission technology of uplink channel in broadband cognitive trunked communication system based on Time-Division Long-Term Evolution（TD-LTE）technology on a multi-core digital signal processor（DSP）and application of frequency-domain energy detection algorithm in interference detection of uplink channel.Firstly,this paper introduces the development status of broadband trunked communication and the key role of interference cognitive technology in trunked communication.Then the paper analyzes the protocol architecture of the broadband cognitive communication system,introduces the physical layer resource definition,and studies the key processing modules of the uplink channel in detail.The paper establishes the floating-point data transmission link on MATLAB platform,and gets the simulation results of transmission performance under both gaussian and EVA channels.Simulation results show that using 16QAM,when the signal-to-noise ratio reaches 11.4dB in the EVA channel,the bit error rate can reach10^-4.Then,this paper uses Texas Instruments'TMS320C6670 multi-core DSP platform to establish a fixed-point data transmission link for the uplink channel.The inter-core synchronization mechanism is used to complete the multi-core task synchronization design of the transceiver considering the hardware feature.The paper studies in depth on the coordinated call of the DSP core and coprocessors,including BCP,FFTC,and TCP3d,to complete the signal processing modules of both the transmitting end and receiving end.Then the fixed-point transmission link is established on the DSP.The paper tests the accuracy error between the fixed-point results of DSP and the floating-point results of MATLAB.The time complexity of the fixed-point link is statistically analyzed,and the feasibility of multi-core pipeline processing at the receiving end is verified.Under both gaussian channel and EVA channel,the transmission performance of the DSP uplink fixed-point link is tested.The results show that the performance difference is only 0.3dB between DSP and MATLAB.The optimazations are made in terms of pipeline structure and memory allocation.Moreover,this paper studies the application of frequency-domain energy detection algorithm in uplink channel interference detection.The principle of the energy detection algorithm in the frequency domain is analyzed,and the advantages and disadvantages of CME and FCME algorithm for determining the detection threshold are compared.The statistical characteristics of the detection statistics under gaussian channel when there is data transmission on the uplink channel are studied,and frequency-domain energy detection algorithm is justified to be available with valid signal existing in uplink channel.The interference detection link is set up and tested with single-tone interference,multi-tone interference,narrow-band interference and SC-FDMA interference under the background of data signal and pilot signal respectively on MATLAB.Simulation results show that under the background of data signal,when the interference-to-noise ratio of narrow-band SC-FDMA interference reaches-8.4dB,the detection probability can reach 90%.Finally this paper researches,designs and implements the inter-layer high-speed communication interface among DSP,the host computer of the media access layer and the FPGA of front end at both transmitting end and receiving end of uplink channel.According to the protocol standards of the broadband cognitive communication system,the communication modes of the ethernet port and the high-speed serial port are used to design and implement the data channel and control channel between DSP board and each adjacent layer for the uplink transceiver.Under the scheduling control of each layer,the communication scheme are jointly tested.The timing order correctness,data processing correctness,and the feasibility of full-rate operation of the DSP baseband processing of the uplink transceiver are verified.Cooperated with the high-level protocol stack and lower layer,the broadband cognitive communication system is preliminarily realized.