A Novel Data-Aided Channel Estimation with Reduced Complexity for TDS-OFDM Systems

TL;DR

This paper introduces a data-aided channel estimation method for TDS-OFDM systems that reduces complexity by avoiding decoding feedback, utilizing both PN sequences and data symbols for improved accuracy in challenging conditions.

Contribution

The paper presents a novel, low-complexity data-aided channel estimation technique combining PN-based and data-based estimates without turbo decoding, enhanced by filtering and iterative refinement.

Findings

Effective in harsh channel conditions like SFN

Reduces complexity compared to turbo estimation

Improves MSE and BER performance

Abstract

In contrast to the classical cyclic prefix (CP)-OFDM, the time domain synchronous (TDS)-OFDM employs a known pseudo noise (PN) sequence as guard interval (GI). Conventional channel estimation methods for TDS-OFDM are based on the exploitation of the PN sequence and consequently suffer from intersymbol interference (ISI). This paper proposes a novel dataaided channel estimation method which combines the channel estimates obtained from the PN sequence and, most importantly, additional channel estimates extracted from OFDM data symbols. Data-aided channel estimation is carried out using the rebuilt OFDM data symbols as virtual training sequences. In contrast to the classical turbo channel estimation, interleaving and decoding functions are not included in the feedback loop when rebuilding OFDM data symbols thereby reducing the complexity. Several improved techniques are proposed to refine the data-aided channel estimates, namely one-dimensional (1-D)/two-dimensional (2-D) moving average and Wiener filtering. Finally, the MMSE criteria is used to obtain the best combination results and an iterative process is proposed to progressively refine the estimation. Both MSE and BER simulations using specifications of the DTMB system are carried out to prove the effectiveness of the proposed algorithm even in very harsh channel conditions such as in the single frequency network (SFN) case.