Denoising Diffusion Probabilistic Models for Hardware-Impaired Communications

Mehdi Letafati,Samad Ali,Matti Latva-aho

2023-10-05

Abstract:Generative AI has received significant attention among a spectrum of diverse industrial and academic domains, thanks to the magnificent results achieved from deep generative models such as generative pre-trained transformers (GPT) and diffusion models. In this paper, we explore the applications of denoising diffusion probabilistic models (DDPMs) in wireless communication systems under practical assumptions such as hardware impairments (HWI), low-SNR regime, and quantization error. Diffusion models are a new class of state-of-the-art generative models that have already showcased notable success with some of the popular examples by OpenAI1 and Google Brain2. The intuition behind DDPM is to decompose the data generation process over small ``denoising'' steps. Inspired by this, we propose using denoising diffusion model-based receiver for a practical wireless communication scheme, while providing network resilience in low-SNR regimes, non-Gaussian noise, different HWI levels, and quantization error. We evaluate the reconstruction performance of our scheme in terms of mean-squared error (MSE) metric. Our results show that more than 25 dB improvement in MSE is achieved compared to deep neural network (DNN)-based receivers. We also highlight robust out-of-distribution performance under non-Gaussian noise.

Information Theory,Signal Processing

What problem does this paper attempt to address?

### Problems Addressed by the Paper

This paper aims to investigate the application of Denoising Diffusion Probabilistic Models (DDPM) in wireless communication systems with hardware impairments. Specifically, the authors propose using DDPM to enhance the robustness of wireless communication systems under conditions of low Signal-to-Noise Ratio (SNR), non-Gaussian noise, various levels of hardware impairments, and quantization errors.

#### Main Contributions:
1. **Proposing the DDPM Model**: The authors use the DDPM model at the receiver end (without relying on any other traditional autoencoders) to improve the wireless system's ability to counteract real-world non-ideal conditions such as hardware impairments and quantization errors.
2. **Performance Evaluation**: The proposed scheme is evaluated using the Mean Squared Error (MSE) metric, demonstrating a performance improvement of over 25dB compared to receivers based on Deep Neural Networks (DNN).
3. **Robustness Analysis**: The method's robustness is validated under low SNR environments, non-Gaussian noise, and various levels of hardware impairments, showcasing its superior performance under these conditions.

### Summary
By introducing the DDPM framework, this paper addresses practical issues in traditional wireless communication systems, such as hardware impairments and channel noise, and demonstrates the effectiveness and robustness of the proposed method through experiments.