Algorithm Research And Implementation Of PESQ Speech Quality Assessment System

Due to the development of communication technology, the performance estimation of communication system and speech coding is becoming very important. So, it's a focus for researchers to develop the speech assessment system, which is flexible, convenient and reliable.PESQ algorithm is one of the speech quality objective assessments which are put forward by International Telecommunications Union. It takes the end-to-end delay into consideration, and adopts improved acoustical and cognitive technology. So it's robust to the communication delay and noise of environment. In the algorithms ITU declared, PESQ has the highest correlation with the result of subjective assessment. Considering the original and distorted speech signals required must be acquired synchronously, the paper proposes a PESQ system based on TMS320VC33, which uses the precise timing function of GPS to synchronize. Scheme of the system is presented in detail in the paper. The main work of this thesis includes the following parts:(1) The methods of speech quality assessment are introduced, including the performance comparison of various methods.(2) The earlier models of objective assessment and the disadvantages of them are introduced.(3) A popular applied speech objective assessment model-PESQ is analyzed in detail.(4) A scheme of end-to-end speech quality assessment system based on DSP is proposed, which apply GPS to synchronize. The function and performance demand of the system are analyzed.(5) The scheme of software and hardware design and the work flow of the system are introduced.(6) The floating-point DSP-TMS320VC33 is introduced. The format of GPS's output information is introduced. The software and hardware interface of the synchronous block in the system are designed and implemented.(7) The PESQ arithmetic on the platform of TMS320VC33 is implemented. The algorithm and codes are optimized and the methods of optimization are given in detail.