Abstract
We design a soft-in soft-out (SISO) decision feedback equalizer (DFE) that performs better than its linear counterpart in turbo equalizer (TE) setting. Unlike previously developed SISO-DFEs, the present DFE scheme relies on extrinsic information formulation that directly takes into account the error propagation effect. With this new approach, both error rate simulation and the extrinsic information transfer (EXIT) chart analysis indicate that the proposed SISO-DFE is superior to the well-known SISO linear equalizer (LE). This result is in contrast with the general understanding today that the error propagation effect of the DFE degrades the overall TE performance below that of the TE based on a LE. We also describe a new extrinsic information combining strategy involving the outputs of two DFEs running in opposite directions, that explores error correlation between the two sets of DFE outputs. When this method is combined with the new DFE extrinsic information formulation, the resulting "bidirectional" turbo-DFE achieves excellent performance-complexity tradeoffs compared to the TE based on the BCJR algorithm or on the LE. Unlike turbo LE or turbo DFE, the turbo BiDFE's performance does not degrade significantly as the feedforward and feedback filter taps are constrained to be time-invariant.
Introduction
In high-speed digital communications, signals often overlap, causing Intersymbol Interference (ISI). To mitigate this, receivers use Turbo Equalization, where an equalizer and a decoder iteratively exchange "soft" information (probabilities) to refine the signal estimate. While the BCJR algorithm is the optimal standard for this task, its exponential complexity makes it impractical as channel lengths grow.
Engineers typically turn to simpler Linear Equalizers (LE) or Decision Feedback Equalizers (DFE). Historically, the DFE was considered a poor fit for turbo equalization because it relies on "hard" decisions that can cause error propagation, potentially poisoning the soft information sent to the decoder. This research challenges that consensus by rethinking how soft information is calculated.
Methodology and Mathematical Framework
The core innovation of this paper is a Modified Extrinsic Information Formulation. Standard SISO-DFEs assume feedback decisions are always correct, which leads to inaccurate Log-Likelihood Ratio (LLR) calculations when errors occur. The authors use a first-order Taylor expansion to reformulate the LLR mapping, which directly accounts for the probability that a feedback decision might be wrong. This effectively "shields" the decoder from the effects of error propagation.
As shown in the notation table above, the framework defines critical parameters such as the DFE feedforward ($\mathbf{c}_n$) and feedback ($\mathbf{d}_n$) filter coefficients, and the extrinsic LLR ($L_e(x_n)$) of the transmitted symbol. By modeling the uncertainty of past decisions (causal symbols $\mathbf{x}_n^c$), the DFE becomes a powerful tool within a turbo loop.
Bi-directional DFE (BiDFE)
Building on the improved SISO-DFE, the authors introduce the Bi-directional DFE (BiDFE). This system processes received data twice: once in the normal forward direction and once in a time-reversed direction.
Because errors in a forward DFE and a backward DFE occur at different times and have different statistical distributions, they can be combined to extract a much cleaner signal. The authors use a novel noise-whitening combiner that estimates the statistical correlation ($\rho_n$) between the noise and errors of the two directions. This strategy suppresses noise and error propagation more effectively than either direction could achieve alone.
Key Results and Performance
The researchers evaluated their algorithms using simulated communication channels with severe ISI (e.g., 5-tap and 7-tap channels with spectral nulls). The results demonstrate a significant performance-complexity advantage:
- Near-Optimal Performance: In simulations, the proposed time-varying BiDFE performed almost identically to the optimal MAP (BCJR) equalizer.
- Outperforming the LE: Unlike traditional DFE implementations, this proposed version consistently beat the Linear Equalizer. In severe channel conditions where the LE failed to converge, the BiDFE maintained a low Bit Error Rate (BER).
- Efficiency: The BiDFE provides a "sweet spot" in design. It is slightly more complex than a standard linear equalizer but significantly less complex than the BCJR algorithm. Furthermore, the time-invariant BiDFE (TIV-BiDFE) outperformed the time-varying Linear Equalizer, proving that smart algorithm design can compensate for simpler hardware implementations.
Conclusion
By mathematically modeling the uncertainty of feedback decisions and leveraging directional diversity, the proposed BiDFE overcomes the long-standing limitations of decision feedback in turbo settings. It offers a robust and efficient alternative to optimal algorithms, providing high performance even when constrained by time-invariant hardware filters.
Summary updated