Adaptive Demodulation in Differentially Coherent Phase Systems: Design and Performance Analysis

J. David Brown, Jamshid Abouei, Konstantinos N. Plataniotis, Subbarayan Pasupathy

TL;DR

This paper introduces Adaptive Demodulation (ADM), a novel rate-adaptive communication system that doesn't require the transmitter to know channel conditions. By intelligently adapting the receiver's decision boundaries and using rateless codes, ADM achieves efficient data transmission over challenging wireless channels, specifically for differentially coherent phase systems like M-DPSK and M-DAPSK.

What Problem Does This Paper Address?

Traditional adaptive communication systems often rely on the transmitter knowing the Channel State Information (CSI) to adjust modulation or coding rates. However, obtaining and transmitting this CSI, especially in fast-changing mobile environments, is a significant challenge, adding complexity and overhead. Existing rate-adaptive techniques like adaptive modulation require CSI at the transmitter, while incremental redundancy systems, though CSI-free at the transmitter, can lead to complex receiver decoding and large buffers. This paper tackles the problem of achieving rate adaptation without transmitter CSI and without the complexity of incremental redundancy, specifically for differentially coherent systems where a precise phase reference is not available at the receiver.

Methods

The core of this work is the design and analysis of Adaptive Demodulation (ADM) for two differentially coherent schemes: M-DPSK and M-DAPSK.

  • Adaptive Demodulation (ADM): The transmitter sends data using a fixed rate and standard constellation, encoded with a rateless code. The receiver, using observed Channel State Information (CSI), adaptively adjusts its decision regions to demodulate bits with varying reliability. Reliable bits are buffered until enough are collected to decode the original message.
  • Differentially Coherent Detection: This approach avoids the need for a precise phase reference at the receiver, which is crucial in dynamic wireless environments. The paper focuses on M-DPSK (M-ary Differential Phase Shift Keying) and M-DAPSK (M-ary Differential Amplitude and Phase Shift Keying).
  • Optimal β-Decision Schemes: The paper presents a method to determine the "most reliable" $\beta$ bits from a pair of received symbols. This involves calculating Log-Likelihood Ratios (LLRs) for each bit and selecting the $\beta$ bits with the largest LLR magnitudes.
  • Near-Optimal/Simple Implementations: For practical implementation, the paper derives simplified, near-optimal decision rules for 16-DPSK and 16-DAPSK. These rules aim to reduce computational complexity while maintaining good performance. For 16-DAPSK, heuristic approaches are used to create simpler decision schemes with non-linear boundaries.
  • Bit Error Analysis and Spectral Efficiency: The paper provides analytical methods for calculating the Bit Error Rate (BER) and spectral efficiency of the proposed ADM schemes over AWGN and Rayleigh fading channels.

Datasets

Not discussed in the paper. The evaluation is based on theoretical analysis and simulations.

Key Insights and Ideas

  • CSI-Free Rate Adaptation: The most significant insight is the demonstration that effective rate adaptation is achievable without requiring CSI at the transmitter, a major bottleneck in many existing systems.
  • Leveraging Receiver-Side CSI: ADM cleverly shifts the adaptation burden to the receiver, which can estimate channel conditions and adjust its demodulation strategy accordingly.
  • Differentially Coherent Advantage: Applying ADM to differentially coherent schemes is a practical advancement, as it bypasses the complexity and potential unreliability of coherent phase recovery in dynamic wireless environments.
  • Trade-offs in M-DAPSK: While M-DAPSK offers advantages in certain scenarios, the paper reveals that for recovering fewer bits (1, 2, or 3), 16-DPSK can outperform 16-DAPSK due to its ability to selectively erase error-prone bits for significant gains.
  • Ring Ratio Optimization for M-DAPSK: The optimal ring ratio (R) for 16-DAPSK is shown to be dependent on the number of bits being recovered and the operating SNR, with R=2 being optimal for 4-DS but lower values being better for fewer bits.

Key Results

  • Design of Near-Optimal Receivers: The paper successfully designs and analyzes near-optimal receivers for 16-DPSK and 16-DAPSK within the ADM framework.
  • Performance Comparison:
    • For 16-DAPSK, simplified decision schemes show performance very close to optimal schemes, with a loss of only about 0.6dB at high SNR for $\beta=3$.
  • For 16-DPSK, the derived simple rules lead to straightforward implementations.
  • In Rayleigh fading, ADM with 16-DPSK outperforms ADM with 16-DAPSK for low rate transmission (up to ~2.5 bits/symbol).
  • For higher rates, 16-DPSK still performs competitively, not suffering a large deficit.
  • Spectral Efficiency: Figure 9 shows that ADM with 16-DPSK achieves higher spectral efficiency than ADM with 16-DAPSK over a significant range of SNR.
  • Ring Ratio Impact: For 16-DAPSK, ring ratios greater than 2 are never beneficial for spectral efficiency. Lower ring ratios improve performance at low average SNR, while higher ratios are better at high average SNR.

Limitations and Future Work

  • Complexity of M-DAPSK Schemes: While simplified schemes are proposed, the paper notes that the optimal decision schemes for 16-DAPSK can still involve non-linear boundaries, making them less straightforward to implement than those for 16-DPSK.
  • Ring Ratio Selection for M-DAPSK: The paper highlights that the optimal ring ratio for 16-DAPSK in ADM is not a single value but depends on the desired system performance and operating range. Further investigation into optimal ring ratio selection for different ADM scenarios could be beneficial.
  • Future Work: The paper implicitly suggests further exploration into optimizing the trade-offs between different modulation schemes (like 16-DPSK vs. 16-DAPSK) within the ADM framework for various channel conditions and performance objectives. The impact of different rateless codes could also be a future research direction.

Summary updated

Abstract

Adaptive Demodulation (ADM) is a new rate-adaptive system that operates without requiring Channel State Information (CSI) at the transmitter, instead using adaptive decision region boundaries at the receiver and encoding the data with a rateless code. This paper addresses the design and performance of an ADM scheme for two common differentially coherent schemes: M-DPSK and M-DAPSK. The optimal method for determining the most reliable bits for a given differential detection scheme is presented. In addition, simple (near-optimal) implementations are provided for recovering the most reliable bits from a received pair of differentially encoded symbols for systems using 16-DPSK and 16-DAPSK. The new receivers offer the advantages of a rate-adaptive system, without requiring CSI at the transmitter or a coherent phase reference at the receiver. Bit error analysis for the ADM system in both cases is presented along with numerical results of the spectral efficiency for the rate-adaptive systems operating over a Rayleigh fading channel.
BibTeX
@article{brown2011adaptive,
  author = {J. David Brown and Jamshid Abouei and Konstantinos N. Plataniotis and Subbarayan Pasupathy},
  title = {{Adaptive Demodulation in Differentially Coherent Phase Systems: Design and Performance Analysis}},
  journal = {IEEE Trans. Communications},
  year = {2011},
  url = {https://openreview.net/forum?id=TonjCBAZhb},
}
Figure