5G • 5G-Advanced (Rel-18)
AI-Driven PAPR Reduction
Neural Network Peak-to-Average Power Ratio Mitigation
Uses deep autoencoders to optimize OFDM symbol constellations, minimizing high peak-to-average power ratios to boost power amplifier efficiency.
Technical Explanation
High Peak-to-Average Power Ratio (PAPR) in OFDM signals forces base station and handset Power Amplifiers (PAs) to operate with large back-offs, severely degrading energy efficiency. Rel-18 explores neural network-based tone reservation and active constellation extension (ACE), dynamically modifying unused subcarriers to suppress amplitude spikes without causing adjacent channel leakage.
Key Functions
- Suppresses nonlinear amplitude spikes in wideband 5G-Advanced OFDM waveforms
- Enables transmitter power amplifiers to operate closer to their saturation region
- Extends handset uplink battery life and boosts uplink transmission range
- Reduces out-of-band emissions and Adjacent Channel Leakage Ratio (ACLR)
Specifications
3GPP TR 38.843, TS 38.211 (Rel-18)
Interfaces
Uu (Physical Layer)
Related 5G Concepts
Want to memorize 5G concepts like this one?
Study it with SuperMemo SM-2 spaced repetition flashcards.