SWAN: Swarm-Based Low-Complexity Scheme for PAPR Reduction

TL;DR

This paper introduces SWAN, a swarm intelligence-based optimization method for PAPR reduction in MIMO-OFDM systems, achieving near-optimal performance with significantly lower computational complexity than traditional exhaustive search methods.

Contribution

The paper adapts CS-PTS for MIMO systems and proposes SWAN, a novel swarm intelligence algorithm that reduces complexity while maintaining near-optimal PAPR reduction performance.

Findings

SWAN achieves near-optimal PAPR reduction performance.

SWAN significantly reduces computational complexity.

SWAN balances optimality and efficiency effectively.

Abstract

Cyclically shifted partial transmit sequences (CS-PTS) has conventionally been used in SISO systems for PAPR reduction of OFDM signals. Compared to other techniques, CS-PTS attains superior performance. Nevertheless, due to the exhaustive search requirement, it demands excessive computational complexity. In this paper, we adapt CS-PTS to operate in a MIMO framework, where singular value decomposition (SVD) precoding is employed. We also propose SWAN, a novel optimization method based on swarm intelligence to circumvent the exhaustive search. SWAN not only provides a significant reduction in computational complexity, but it also attains a fair balance between optimality and complexity. Through simulations, we show that SWAN achieves near-optimal performance at a much lower complexity than other competing approaches.