Embedded Echo Cancellation With Integrated Noise Suppression

The patent segments the echo cancellation process into multiple stages: initial echo cancellation using a first set of coefficients, followed by residual echo cancellation using a second set of coefficients. This segmentation allows the system to achieve effective echo cancellation while dividing the computational workload into manageable parts, reducing the overall computational resource requirement compared to a single sophisticated algorithm.

The patent applies partial echo cancellation by using a first set of echo cancellation coefficients to handle the majority of echo removal, then applies a second set of coefficients to address residual echoes. This partial action approach achieves sufficient echo cancellation effectiveness without requiring the full computational power of more sophisticated algorithms, thereby reducing energy and computational resource consumption.

The patent merges multiple audio processing tasks (echo cancellation, noise suppression, filtering, equalization) into a single integrated signal processor. This consolidation reduces device complexity by eliminating the need for separate circuits for each function while maintaining the capability to perform all required audio processing operations through a unified processing architecture.

The integrated signal processor is designed with multi-functionality, capable of performing echo cancellation, noise suppression, filtering, and equalization within a single device. This universal approach allows the system to handle diverse audio processing requirements without increasing device complexity, as all functions share the same processing hardware and computational resources.