Design Of Speech Signal Acquisition And Processing Based On Zynq-7000 Platform

As the carrier of information,electromagnetic wave monitoring is an effective means to obtain information.Wireless speech communication plays an important role in wireless comnunication.Therefore,the monitoring of analog speech communication is an important means of obtaining intelligence in military and civilian fields.The monitoring of analog speech signals requires advanced acquisition and processing equipment.The original device is based on multiple separate receivers,each receiving one speech.This type of equipment is bulky and suitable only for bases with good conditions.For mobile and small platforms,a new type of multi-channel simultaneous listening,suitable for a variety of communication modulation methods,modular portable receiving and collecting devices is needed.Since the speech signal is inevitably subject to external noise interference including natural noise and human noise,interference from network communication,interference between communication networks during wireless transmission,and noise introduced from problems of the communication device itself.The original clean speech signal is transmitted into the non-pure speech signal containing different kinds of noise.The monitors work in a strong noise environment for a long time,causing hearing and psychological damage.Therefore,it is necessary to denoise the noisy speech signal to improve the quality of the speech.Based on the above requirements,this paper studies and simulates the noise recognition and denoising algorithms based on the noise of the actual received signal,build the speech signal acquisition and processing environment based on the Zynq-7000-based embedded RF transceiver hardware platform,completes the algorithm's implementation based on Zynq-7000-based embedded RF transceiver hardware platform.The specific research content of this paper is as follows:Firstly,the paper studies the scheme of speech signal acquisition and processing system and various ninctional modules,and discusses the design of signal acquisition function module in detail.The Zynq-7000-based embedded RF transceiver hardware platform is highlighted.Secondly,aiming at the influence of noise on the analog speech signal,a data segmentation algorithm based on histogram analysis and fuzzy clustering and a new speech signal segment recognition algoritlun based on sub-band energy are proposed.The new algorithm uses the differences of noise and speech in the time-frequency domain distribution as the character.By using cluster analysis and time-frequency domain sub-band energy statistical distribution analysis method,the recognition and identification of speech signal segnents and strong noise signal segments are realized.Finally,the application running environment on the Zynq-7000 platform is built.The U-Boot,Linux kernel,device tree file configuration and compilation required to start the SD card,and the algorithm implementation of the speech signal processing application are implemented in detail.The algorithm was ported to the hardware platform.At the same time,the signal acquisition function and signal processing function in the speech signal acquisition and processing system are verified separately.