MUMBAI, India, June 22 -- Intellectual Property India has published a patent application (202641070855 A) filed by Dr M Madhusudhan Reddy; S. Venkata Anand,; Vijayasri Domadula; Dr. A. Vani; Dr. Nagesh Mantravadi; Dr. Arun Raj S. R; Dr. B. Ramu; Mr. Mohammad Shabeeh; Dr. M. Adiseshaiah; and Mr. C. Sudhakar on June 07, 2026, for A Smart Adaptive Noise Cancellation System For Real-Time Wireless Communication Using Ai-Based Signal Processing.
Inventors include Dr M Madhusudhan Reddy; S. Venkata Anand; Vijayasri Domadula; Dr. A. Vani; Dr. Nagesh Mantravadi; Dr. Arun Raj S. R; Dr. B. Ramu; Mr. Mohammad Shabeeh; Dr. M. Adiseshaiah; and Mr. C. Sudhakar.
The application for the patent was published on June 12, 2026, under issue no. 24/2026.
Abstract: ABSTRACT OF THE INVENTION The present invention provides a smart adaptive noise cancellation system for real-time wireless communication using AI-based signal processing. The system comprises dual microphones—one for primary speech capture and one for ambient noise reference—connected to a low-power digital signal processor running a lightweight recurrent neural network with long short-term memory (LSTM) units. Unlike conventional adaptive filters such as LMS or fixed active noise cancellation, the proposed AI model learns dynamic noise patterns in real time. The system is pre-trained on over 10,000 hours of real-world noisy speech data from 50 environments including traffic, crowd chatter, wind, cafes, and moving vehicles. Once deployed, it performs real-time frame-by-frame noise prediction and subtraction with a latency under 10 milliseconds. A voice activity detector ensures only noise is canceled, preserving voice harmonics and naturalness. The system also includes an on-device unsupervised learning capability, allowing it to adapt to new noise types without cloud dependency. In prototype testing, the system achieved over 28 dB noise reduction, 92% speech intelligibility (STOI), and consumed only 42 mW of power, making it suitable for wireless earbuds, hearing aids, and mobile phones. The invention solves the limitations of traditional ANC, which fails under non-stationary noise and causes voice distortion. The system continuously improves over time via feedback-based parameter adaptation. This smart adaptive solution offers a significant advancement in wireless communication clarity, especially in dynamic acoustic environments.
Disclaimer: Curated by HT Syndication.
