Real-Time Adaptive Noise and Echo Cancellation on a Compact Microcontroller

Abstract

This paper provides an analysis of current advancements made in acoustic noise and echo cancellation (ANEC) and describes a real-time acoustic noise and feedback cancellation implementation that is low-cost and effective. Challenges in developing this compact mobile ANEC system included integrating a high order normalized least mean square (NLMS) adaptive algorithm with an analog synchronized twochannel interface. The paper also presents electrical and mechanical details of this small footprint ANEC prototype along with test results. The system was evaluated in terms of performance and computational complexity using objective and subjective evaluations. Metrics used to assess real-time performance included the Echo Return Loss Enhancement (ERLE), the Added Stable Gain (ASG) as well as algorithm convergence curves.