A Double-Ended Interference Assessing System for Short Wave Transmission

This paper presents an short wave transmission quality assessing system based on speech signal in which there are some jamming signals to interrupt the normal communication. This work aims to give proper objective assessment of the intensity of interference signal which is consistent with subjective one. Except interference signal, there are also delays at receiver side, which cannot be determined in advance. Firstly, cepstral mean normalization and weighted Mel frequency cepstral coefficient are applied to reduce channel noise. Furthermore, a sliding strategy is proposed to find the best matching point during the objective assessment to deal with the delay problem. In objective and subjective mapping model, two kinds of fitting approaches, named two-pass least square fitting and Akima fitting, are combined for different condition. Experiments show the effectiveness of proposed assessment approach.