[Paper Review] Interference Mitigation via Interference-Aware Successive Decoding

Hyukjoon Kwon, Jungwon Lee+1 more|arXiv (Cornell University)|Sep 18, 2012

Advanced Wireless Communication Techniques23 references4 citations

TL;DR

This paper proposes Interference-Aware Successive Decoding (IASD) and its extension, Interference-Aware Parallel Decoding (IAPD), for MIMO interference channels using point-to-point codes and BICM. By successively decoding desired and interference signals while updating a priori information, IASD significantly outperforms interference non-decoding and interference detection-only schemes, especially at high SIR and high code rates, demonstrating that interference decoding enhances performance without increasing complexity unduly.

ABSTRACT

In modern wireless networks, interference is no longer negligible since each cell becomes smaller to support high throughput. The reduced size of each cell forces to install many cells, and consequently causes to increase inter-cell interference at many cell edge areas. This paper considers a practical way of mitigating interference at the receiver equipped with multiple antennas in interference channels. Recently, it is shown that the capacity region of interference channels over point-to-point codes could be established with a combination of two schemes: treating interference as noise and jointly decoding both desired and interference signals. In practice, the first scheme is straightforwardly implementable, but the second scheme needs impractically huge computational burden at the receiver. Within a practical range of complexity, this paper proposes the interference-aware successive decoding (IASD) algorithm which successively decodes desired and interference signals while updating a priori information of both signals. When multiple decoders are allowed to be used, the proposed IASD can be extended to interference-aware parallel decoding (IAPD). The proposed algorithm is analyzed with extrinsic information transfer (EXIT) chart so as to show that the interference decoding is advantageous to improve the performance. Simulation results demonstrate that the proposed algorithm significantly outperforms interference non-decoding algorithms.

Motivation & Objective

• To address the growing challenge of inter-cell interference in dense cellular networks with small cells.
• To develop a practical receiver algorithm that enables joint decoding of desired and interference signals without intractable computational complexity.
• To improve spectral efficiency and throughput in interference-limited environments using successive decoding with updated a priori information.
• To demonstrate that decoding interference signals—rather than treating them as noise or detecting them only—leads to significant performance gains.

Proposed method

• Proposes Interference-Aware Successive Decoding (IASD), which successively decodes desired and interference signals while iteratively updating a priori information using extrinsic information from both signals.
• Extends IASD to Interference-Aware Parallel Decoding (IAPD) for multiple decoder support, enabling parallel processing of desired and interference signals.
• Employs extrinsic information transfer (EXIT) charts to analyze convergence and performance gain, assuming Gaussian-distributed a priori information.
• Uses bit-interleaved coded modulation (BICM) with iterative detection and decoding (IDD) to exchange log-likelihood ratios (LLRs) between detector and decoder.
• Incorporates modulation and coding rate information of interference signals as side information to improve decoding accuracy.
• Validates the approach through simulations across various SNRs, code rates (0.33–0.83), modulation schemes (16/64 QAM), and SIR levels (±3 dB).

Experimental results

Research questions

• RQ1 Can interference decoding improve performance in MIMO interference channels when using practical point-to-point codes and BICM?
• RQ2 Does successive decoding of both desired and interference signals, with updated a priori information, outperform treating interference as noise or detecting it only?
• RQ3 How does the performance of IASD and IAPD vary with code rate, modulation order, and signal-to-interference ratio (SIR)?
• RQ4 What is the impact of inaccurate a priori information on interference decoding, and does it degrade the performance of the desired signal decoding?
• RQ5 Can EXIT chart analysis accurately predict the convergence and performance of the proposed interference-aware decoding schemes?

Key findings

• IASD and IAPD achieve significant performance gains over interference non-decoding (IW) and interference-aware detection (IA-Detection), especially at high SIR levels and high code rates.
• At code rate 0.83 and 16 QAM, the performance gain of IASD over IW increases with decreasing SIR, indicating robustness to interference strength.
• The proposed IASD outperforms IIAD (IASD without interference decoding), proving that interference decoding provides measurable performance gain even with imperfect a priori information.
• Simulation results closely match EXIT chart predictions, validating the analytical model’s accuracy under finite-length and discrete modulation assumptions.
• IAPD fails to support code rate 0.75 at 6 dB SNR with 2 iterations, but achieves reliable performance with 3 iterations, confirming the importance of iterative convergence.
• The PER of the desired signal is consistently lower than that of the interference signal in IASD, confirming the method’s design focus on enhancing desired signal reliability.