Energy Efficiency in AI for 5G and Beyond: A DeepRx Case Study

TL;DR

This paper evaluates the energy consumption of DeepRX, a deep learning receiver, and demonstrates how knowledge distillation can create smaller, energy-efficient models with comparable or better performance in 5G signal processing.

Contribution

It introduces the application of knowledge distillation to develop energy-efficient DeepRX models that maintain high performance, extending understanding of energy dynamics in AI-based 5G receivers.

Findings

Knowledge distillation reduces energy consumption while maintaining performance.

Distilled models show lower error floors across SINR levels.

Energy consumption correlates with memory access patterns.

Abstract

This study addresses the challenge of balancing energy efficiency with performance in AI/ML models, focusing on DeepRX, a deep learning receiver based on a fully convolutional ResNet architecture. We evaluate the energy consumption of DeepRX, considering factors including FLOPs/Watt and FLOPs/clock, and find consistency between estimated and actual energy usage, influenced by memory access patterns. The research extends to comparing energy dynamics during training and inference phases. A key contribution is the application of knowledge distillation (KD) to train a compact DeepRX student model that emulates the performance of the teacher model but with reduced energy consumption. We experiment with different student model sizes, optimal teacher sizes, and KD hyperparameters. Performance is measured by comparing the Bit Error Rate (BER) performance versus Signal-to-Interference & Noise Ratio (SINR) values of the distilled model and a model trained from scratch. The distilled models demonstrate a lower error floor across SINR levels, highlighting the effectiveness of KD in achieving energy-efficient AI solutions.