Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

CVNN-based Channel Estimation and Equalization in OFDM Systems Without Cyclic Prefix

Heitor dos Santos Sousa, Jonathan Aguiar Soares, Kayol Soares Mayer, Dalton Soares Arantes

TL;DR

This paper tackles the spectral efficiency loss in OFDM caused by cyclic prefixes by introducing a complex-valued neural approach for CP-free joint channel estimation and equalization. It proposes a phase-transmittance radial basis function (PT-RBF) CVNN that operates on FFT outputs and sidesteps CP decoding, reducing computational load while maintaining performance. Empirical results show the PT-RBF outperforms traditional LS and MMSE estimators in CP-free scenarios (e.g., improvements of about 2.68 dB and 10.51 dB at a pre-FEC BER of $2\times10^{-2}$), and incurs roughly 30% less real-multiplication complexity than a competing CVNN-based method. The work supports CP-free OFDM viability and suggests FFT keeps outside the neural network to preserve efficiency, with potential applicability to 5G-and-beyond channels and dynamics.

Abstract

In modern communication systems operating with Orthogonal Frequency-Division Multiplexing (OFDM), channel estimation requires minimal complexity with one-tap equalizers. However, this depends on cyclic prefixes, which must be sufficiently large to cover the channel impulse response. Conversely, the use of cyclic prefix (CP) decreases the useful information that can be conveyed in an OFDM frame, thereby degrading the spectral efficiency of the system. In this context, we study the impact of CPs on channel estimation with complex-valued neural networks (CVNNs). We show that the phase-transmittance radial basis function neural network offers superior results, in terms of required energy per bit, compared to classical minimum mean-squared error and least squares algorithms in scenarios without CP.

CVNN-based Channel Estimation and Equalization in OFDM Systems Without Cyclic Prefix

TL;DR

This paper tackles the spectral efficiency loss in OFDM caused by cyclic prefixes by introducing a complex-valued neural approach for CP-free joint channel estimation and equalization. It proposes a phase-transmittance radial basis function (PT-RBF) CVNN that operates on FFT outputs and sidesteps CP decoding, reducing computational load while maintaining performance. Empirical results show the PT-RBF outperforms traditional LS and MMSE estimators in CP-free scenarios (e.g., improvements of about 2.68 dB and 10.51 dB at a pre-FEC BER of ), and incurs roughly 30% less real-multiplication complexity than a competing CVNN-based method. The work supports CP-free OFDM viability and suggests FFT keeps outside the neural network to preserve efficiency, with potential applicability to 5G-and-beyond channels and dynamics.

Abstract

In modern communication systems operating with Orthogonal Frequency-Division Multiplexing (OFDM), channel estimation requires minimal complexity with one-tap equalizers. However, this depends on cyclic prefixes, which must be sufficiently large to cover the channel impulse response. Conversely, the use of cyclic prefix (CP) decreases the useful information that can be conveyed in an OFDM frame, thereby degrading the spectral efficiency of the system. In this context, we study the impact of CPs on channel estimation with complex-valued neural networks (CVNNs). We show that the phase-transmittance radial basis function neural network offers superior results, in terms of required energy per bit, compared to classical minimum mean-squared error and least squares algorithms in scenarios without CP.
Paper Structure (7 sections, 13 equations, 4 figures, 3 tables)

This paper contains 7 sections, 13 equations, 4 figures, 3 tables.

Table of Contents

  1. Introduction
  2. OFDM System Model
  3. Channel Estimation and Equalization
  4. Proposed CVNN Channel Estimation and Equalization
  5. Computational Complexities
  6. Results
  7. Conclusions

Figures (4)

  • Figure 1: Orthogonal frequency division multiplexing (OFDM) communication architecture.
  • Figure 2: Orthogonal frequency division multiplexing (OFDM) communication architecture with the proposed CVNN-based channel estimation and equalization. The CVNN block represents the PT-RBF neural network composed of three layers.
  • Figure 3: BER results of the proposed approach, MMSE, and LS channel estimation algorithms with and without CP. Solid lines regard results without CP and dashed lines with CP.
  • Figure 4: PT-RBF joint channel estimation and equalization results depending on the number of neurons $N_n$. Results magnified for a BER $=2\times10^{-2}$ to highlight performance.