Adaptive Learning Based Active Noise Cancellation
Yu Zhang, Jiangtao Wen, Yuxing Han · 2018
Synchronous audio system is a broadcasting system that speakers from different locations can play the same source simultaneously. In addition to the synchronous playback, the system also can be used to cancel the sound of another speaker. However, the time synchronization accuracy requirement of the active noise cancellation is very high, and general time synchronization protocols can not reach the demand. Therefore, we propose an adaptive learning based active noise cancellation method, which can improve the synchronization rate between speakers to achieve effective active noise cancellation. In order to improve the accuracy of the synchronization, we use WSN (i.e. wireless sensor network) synchronization protocols and the genetic algorithm to learn and adjust parameters for the noise cancellation. Besides, adaptive learning based active noise cancellation does not require high accuracy microphones and dedicated chips, and its scope of applications is not limited to the handset. We also introduce an use case and make a comprehensive discussion.