Adaptive preprocessor system for noise suppression in cellular phones

Abstract

Digital voice systems have been in use for many applications including secure military voice communications, hands-free or cellular mobile telephone systems. In such applications, the speech signal is corrupted by environmental ambient noise. This acoustically added background noise degrades the performance of digital voice processors used for various functions like speech compressors, encoders etc. Further, due to dramatic variance in characteristics of noise in different environments, a single algorithm cannot prove equally beneficial in all circumstances. Finally, the complexity is a major barrier when dealing with systems involving runtime signal processing for mobile phones, as issues such as time complexity and MIPS as involved. Hence, some kind of noise reduction technique has to be employed in order to maintain the performance level of the system at par with that in case of a noise-free environment, regardless of the actual environment in which it is employed. To overcome these problems this paper propose to design a pre-processor system for noise suppression which should adapt to the environmental ambient noise and consequently apply a suitable algorithm.