Noise reduction and speech enhancement for high-quality wireless handsets

Summary form only given. In wireless communications, the noise is a more serious problem than in landline communications. To end users, the voice from a wireless handset includes various kinds of noises involved from different sources and communication stages, such as the environmental background noise, electronic device noise, coding and computation noise, communication channel noise, etc. How to reduce the noises and enhance the speech is an important research topic to high-quality wireless communications, and noise reduction is becoming a new feature and selling point to wireless handsets. Supported by the U.S. Government research funding, we have been studying the problem for years and have developed new algorithms and chips for noise reduction and speech enhancement for commercial wireless environments. Our algorithm is based on the mathematic models of the human hearing system, and it can improve the SNR for more than 15 dB in the environments of commercial wireless applications. The chip is a low cost and low power consumption implementation. It can be inserted into wireless handsets or equipments directly. In this presentation, we will demonstrate our noise reduction algorithms with various background noises in comparison with other products on the market. Our chips with pre-loaded algorithm are ready for cell phone manufactures for new high-quality wireless handsets.