MUMBAI, India, Feb. 6 -- Intellectual Property India has published a patent application (202631010864 A) filed by C. V. Raman Global University, Bhubaneswar, Orissa, on Feb. 2, for 'a deep learning-based on precoding and companding in multi-user massive mimo-ofdm systems for papr suppression.'

Inventor(s) include Dr. Ajay Kumar Yadav; Dr. Tusar Kanti Dash; Dr. Jagannath Dayal Pradhan; Dr. Pushpendra Kumar Gupta; and Dr. Soumya Mishra.

The application for the patent was published on Feb. 6, under issue no. 06/2026.

According to the abstract released by the Intellectual Property India: "The present invention relates to a novel deep learning-based framework for simultaneous optimization of precoding and companding in multi-user massive MIMO-OFDM wireless communication systems, specifically designed to suppress Peak-to-Average Power Ratio (PAPR) while enhancing spectral efficiency and power amplifier efficiency. Traditional approaches handle precoding and companding separately, resulting in suboptimal performance and inherent trade-offs between signal quality and power efficiency. The invention introduces a unified neural network architecture that jointly learns and optimizes both precoding weights and adaptive per-user/perantenna companding parameters by training on realistic channel state information, OFDM signal characteristics, and power amplifier nonlinearity models. The method incorporates physics-aware differentiable signal processing layers including IFFT and PA nonlinearity models within the neural network training loop, enabling end-to-end gradient-based optimization of time-domain PAPR, spectral regrowth, and bit-error rate simultaneously. The architecture supports hardwarefriendly deployment on FPGA, ASIC, and DSP platforms suitable for 5G/6G base stations, enabling real-time inference with minimal computational overhead. Experimental validation demonstrates significant improvements in PAPR reduction, BER performance, and power efficiency compared to conventional separate precoding and companding techniques, while maintaining user fairness and constellation integrity across multi-user scenarios. The invention is particularly valuable for dense wireless networks where power amplifier back-off limitations and multi-user interference pose critical performance bottlenecks."

Disclaimer: Curated by HT Syndication.