Precoded GFDM System to Combat Inter Carrier Interference : Performance Analysis

TL;DR

This paper evaluates the performance of precoded GFDM systems with MMSE receivers for 5G, proposing precoding techniques that improve spectral efficiency, reduce complexity, and lower PAPR compared to traditional GFDM.

Contribution

It introduces novel precoding schemes for GFDM that enhance performance and reduce complexity without sacrificing spectral efficiency or BER performance.

Findings

BIDFT and DFT-based precoding outperform GFDM-MMSE in BER due to frequency diversity.

Precoding schemes significantly reduce PAPR in GFDM systems.

Complexity analysis shows proposed methods are comparable to zero-forcing receivers.

Abstract

The expected operating scenarios of 5G pose a great challenge to orthogonal frequency division multiplexing (OFDM) which has poor out of band (OoB) spectral properties, stringent synchronization requirements, and large symbol duration. Generalized frequency division multiplexing (GFDM) which is the focus of this work, has been suggested in the literature as one of the possible solutions to meet 5G requirements. In this work, the analytical performance evaluation of MMSE receiver for GFDM is presented. We also proposed precoding techniques to enhance the performance of GFDM. A simplified expression of SINR for MMSE receiver of GFDM is derived using special properties related to the modulation matrix of GFDM, which are described in this work. This SINR is used to evaluate the BER performance. Precoding schemes are proposed to reduce complexity of GFDM-MMSE receiver without compromising on the performance. Block Inverse Discrete Fourier Transform (BIDFT) and Discrete Fourier Transform (DFT) based precoding schemes are found to outperform GFDM-MMSE receiver due to frequency diversity gain while having complexity similar to zero-forcing receiver of GFDM. It is shown that both BIDFT and DFT-based precoding schemes reduce peak to average power ratio (PAPR) significantly. Computational complexity of different transmitters and receivers of precoded and uncoded GFDM is also presented.