Signal Detection Under Short-Interval Sampling of Continuous Waveforms for Optical Wireless Scattering Communication

Difan Zou, Chen Gong, Zhengyuan Xu

TL;DR

This paper tackles the challenge of reliable signal detection in optical wireless scattering communication systems that use short-interval sampling of continuous waveforms. The authors propose new methods for calculating transmission rate bounds and develop efficient signal detection and threshold selection strategies, demonstrating that optimal detection can be achieved with reduced computational complexity and robustness to noise.

What Problem Does This Paper Address?

Optical wireless communication (OWC) offers a promising alternative to traditional radio frequency (RF) communication, especially in environments where RF is restricted. Non-line-of-sight (NLOS) OWC, which relies on scattering, is particularly attractive for its wider application range. However, OWC systems, especially NLOS ones, suffer from significant path loss, leading to very weak received signals characterized by discrete photoelectrons. Practical receivers, unlike ideal photon-counting ones, use continuous waveform processing (WP) with photomultiplier tubes (PMTs) and finite sampling intervals. This introduces complexities in signal detection and limits the achievable transmission rate. Specifically, the paper addresses how to model and analyze the performance of such practical WP receivers, focusing on determining transmission rate bounds and developing efficient signal detection and threshold selection methods that balance performance with computational cost.

Methods

The paper employs a multi-faceted approach:

  • Transmission Rate Bounds: It derives upper and lower bounds on the transmission rate for the WP receiver with finite sampling intervals. These bounds are shown to converge as noise variances decrease.
  • Signal Detection:
    • MAP Detection: A Maximum A Posteriori (MAP) detector is formulated for symbol detection.
  • Piecewise Approximation: To reduce the computational complexity of the MAP detector, piecewise linear and cubic polynomial approximations are developed for the log-likelihood ratio function.
  • Threshold Selection: For a simplified non-ideal photon-counting receiver (hard-decision based), two threshold selection criteria are proposed: minimizing detection error probability and maximizing the Kullback-Leibler (KL) distance. The KL distance criterion is highlighted for its lower computational complexity.

Datasets

Not discussed in the paper. The evaluation is based on numerical simulations.

Key Insights and Ideas

  • Finite Sampling Matters: The paper highlights that the theoretical capacity of Poisson channels, achieved with infinitesimally small sampling intervals, is not directly achievable with practical WP receivers that use finite sampling.
  • Approximation is Key for Efficiency: The proposed piecewise linear and cubic polynomial approximations for the log-likelihood ratio function significantly reduce computational complexity while maintaining high detection performance, especially the cubic approximation.
  • KL Distance as a Practical Metric: The KL distance-based threshold selection is shown to be a robust and computationally efficient alternative to minimizing the exact detection error probability, with minimal performance loss.
  • Noise Sensitivity: The study demonstrates that for small thermal and shot noise variances, the detection threshold can be selected over a wide range without significant degradation in performance (KL distance), making the system more practical.

Key Results

  • The derived upper and lower bounds on the transmission rate are shown to be tight for PMTs with large amplification factors and small spreading factors.
  • The piecewise cubic polynomial approximation offers significantly better performance than the piecewise linear approximation, especially in high thermal noise regimes.
  • The threshold selection rule based on KL distance is shown to be close to the optimal threshold derived from minimizing the exact detection error probability, with substantially lower computational complexity.
  • Numerical results indicate that the non-ideal photon-counting receiver with hard-decision can serve as a good approximation to the optimal receiver, offering reduced complexity and performance that is not overly sensitive to threshold selection under low noise conditions.

Limitations and Future Work

  • Idealized Noise Models: The paper assumes specific noise models (Gaussian for thermal and shot noise). Real-world scenarios might involve more complex noise distributions.
  • Specific Modulation Scheme: The analysis is primarily focused on On-Off Keying (OOK) modulation. Extending the framework to other modulation schemes would be valuable.
  • Experimental Validation: While numerical results are presented, direct experimental validation of the proposed methods in a real OWC system is not included.

Future work could involve exploring more sophisticated noise models, investigating other modulation schemes, and conducting experimental validation to confirm the practical efficacy of the proposed techniques.

Summary updated

Abstract

In optical wireless scattering communication, the received signal in each symbol interval is captured by a photomultiplier tube and then sampled through very short but finite interval sampling. The resulting samples form a signal vector for symbol detection. The upper and lower bounds on transmission rate of such a processing system are studied. It is shown that the gap between two bounds approaches zero as the thermal noise and shot noise variances approach zero. In order to reduce the computational cost in maximum a posteriori (MAP) receiver, the threshold-based signal detection is also studied, where two threshold selection rules are proposed based on the detection error probability and the Kullback-Leibler (KL) distance. For the latter, it is shown that the KL distance is not sensitive to the threshold selection for small shot and thermal noise variances, and thus, the threshold can be selected among a wide range without significant loss from the optimal KL distance. The performances of the transmission rate bounds, the signal detection, and the threshold selection approaches are evaluated by the numerical results.
BibTeX
@article{zou2018signal,
  author = {Difan Zou and Chen Gong and Zhengyuan Xu},
  title = {{Signal Detection Under Short-Interval Sampling of Continuous Waveforms for Optical Wireless Scattering Communication}},
  journal = {IEEE Trans. Wireless Communications},
  year = {2018},
  url = {https://openreview.net/forum?id=KbOQquOsGw},
}
Figure