Improving speech understanding in communication headsets: Simulation of adaptive subband processing for speech in noise

摘要

Speech communication headsets are necessary for many high-noise environments to maintain interaction between individuals and facilitate safe working conditions. However, current hearing protection devices intended to protect hearing health can impede speech communication or expose persons to sound pressure levels (SPLs) that could lead to excessive noise exposure if a communication signal is presented improperly. This paper explores an adaptive subband communication algorithm, based on a delayless subband active noise reduction architecture, intended to adjust the communication channel gain to provide an appropriate speech signal power in relation to the instantaneous environmental noise power. The method monitors SPLs underneath the ear cup of a communication headset to provide a target speech signal-to-noise ratio without exceeding safe noise exposure thresholds. A series of computer simulations derived from a real-world communication headset model are used to compare the method developed with a traditional passive attenuation headset and a commercial active noise reduction design. The simulations demonstrate the ability of the adaptive subband communication algorithm to adjust automatically the speech signal gain for improved intelligibility while maintaining healthy noise exposure levels. Relevance to industry: The electro-acoustic performance of an active speech communication headset is explored by simulation. The concept integrates a subband active noise control algorithm with an adaptive gain control structure to improve speech intelligibility in a noisy environment. The concept automatically selects appropriate communication channel gain levels without exceeding hearing damage thresholds or requiring user input, and is directly applicable to a practical device.