An Accurate Sample Rejection Estimator for the Estimation of Outage\n Probability of EGC Receivers

Nadhir Ben Rached, Abla Kammoun, Mohamed‐Slim Alouini, Raúl TemponearXiv 2019 · DOI · arXiv

Abstract

In this work, we evaluate the outage probability (OP) for L-branch equal gain combining (EGC) diversity receivers operating over fading channels, i.e. equivalently the cumulative distribution function (CDF) of the sum of the L channel envelopes. In general, closed form expressions of OP values are unobtainable. The use of Monte Carlo (MC) simulations is not considered a good alternative as it requires a large number of samples for small values of OP, making MC simulations very expensive. In this paper, we use the concept of importance sampling (IS), being known to yield accurate estimates using fewer simulation runs. Our proposed IS scheme is essentially based on sample rejection where the IS probability density function (PDF) is the truncation of the underlying PDF over the L dimensional sphere. It assumes the knowledge of the CDF of the sum of the L channel gains in a closed-form expression. Such an assumption is not restrictive since it holds for various challenging fading models. We apply our approach to the case of independent Rayleigh, correlated Rayleigh, and independent and identically distributed Rice fading models. Next, we extend our approach to the interesting scenario of generalised selection combining receivers combined with EGC under the independent Rayleigh fading environment. For each case, we prove the desired bounded relative error property. Finally, we validate these theoretical results through some selected experiments.

Introduction

In modern wireless diversity systems, receivers utilize multiple antennas to mitigate the effects of fading. Equal Gain Combining (EGC) is a widely adopted technique where signals from different branches are summed to improve signal quality. To design reliable communication systems, engineers must accurately calculate the Outage Probability (OP)—the probability that the combined signal strength falls below a required threshold.

Mathematically, the OP for EGC involves calculating the cumulative distribution function (CDF) of a sum of random variables representing channel envelopes. For most realistic fading models, such as Rayleigh or Rice, no simple closed-form formula exists for this sum. Traditionally, engineers rely on Monte Carlo (MC) simulations. However, for high-reliability systems where outages are rare (e.g., $10^{-9}$), standard MC simulations require trillions of samples to achieve precision, making them computationally prohibitive.

Methodology: Sample Rejection Importance Sampling

To overcome the limitations of standard MC simulations, this research proposes a variance reduction technique known as Importance Sampling (IS). Instead of sampling from the true distribution where outage events are rare, the authors sample from a "biased" distribution where outages occur more frequently, then apply a mathematical correction.

The core innovation lies in the selection of the biased distribution through Sample Rejection. While the CDF of EGC (sum of envelopes) is difficult to solve, the CDF of Maximum Ratio Combining (MRC, sum of squared envelopes) often has a known closed-form solution. The authors define a biased probability density function (PDF) by truncating the original PDF over a multidimensional hypersphere defined by the MRC outage region. This forces the simulation to generate samples only within the region where an EGC outage is likely to occur.

This approach was applied to several complex scenarios:

  • Independent and non-identically distributed (i.n.i.d.) Rayleigh fading.
  • Correlated Rayleigh fading.
  • i.i.d. Rice fading.
  • Generalized Selection Combining (GSC) paired with EGC.

Key Results and Performance

The proposed estimator demonstrates a Bounded Relative Error property. This means that as the outage event becomes increasingly rare, the number of samples required to maintain a specific accuracy level remains stable rather than increasing to infinity.

Figure 1: Comparison of the number of simulation runs required for a 5% relative error across different thresholds.

As shown in Figure 1, the efficiency gains are massive. For a 4-branch receiver with an outage probability of $10^{-9}$, a naive MC simulation requires approximately $1.5 \times 10^{12}$ runs. In contrast, the proposed estimator achieves the same 5% relative error in just $1.5 \times 10^5$ runs—a speedup of seven orders of magnitude.

Figure 2: Estimated outage probability versus threshold $\gamma_{th}$ for various fading scenarios.

Figure 2 validates the accuracy of the estimator. The theoretical "bounded error" proofs were consistently matched by experimental simulation results across all tested fading models, including those with correlated channels which typically pose significant challenges for rare-event estimators.

Conclusion

The proposed sample rejection estimator provides a highly efficient and accurate framework for predicting outage probabilities in EGC receivers. While the method requires a closed-form expression for the corresponding MRC receiver's OP, such expressions are available for most standard fading models. Future work may extend this framework to more complex distributions like Nakagami-m or $\kappa-\mu$, which are critical for 5G and 6G network research.

Summary updated

Abstract

In this work, we evaluate the outage probability (OP) for L-branch equal gain\ncombining (EGC) diversity receivers operating over fading channels, i.e.\nequivalently the cumulative distribution function (CDF) of the sum of the L\nchannel envelopes. In general, closed form expressions of OP values are\nunobtainable. The use of Monte Carlo (MC) simulations is not considered a good\nalternative as it requires a large number of samples for small values of OP,\nmaking MC simulations very expensive. In this paper, we use the concept of\nimportance sampling (IS), being known to yield accurate estimates using fewer\nsimulation runs. Our proposed IS scheme is essentially based on sample\nrejection where the IS probability density function (PDF) is the truncation of\nthe underlying PDF over the L dimensional sphere. It assumes the knowledge of\nthe CDF of the sum of the L channel gains in a closed-form expression. Such an\nassumption is not restrictive since it holds for various challenging fading\nmodels. We apply our approach to the case of independent Rayleigh, correlated\nRayleigh, and independent and identically distributed Rice fading models. Next,\nwe extend our approach to the interesting scenario of generalised selection\ncombining receivers combined with EGC under the independent Rayleigh fading\nenvironment. For each case, we prove the desired bounded relative error\nproperty. Finally, we validate these theoretical results through some selected\nexperiments.\n
BibTeX
@misc{rached2019accurate,
  author = {Nadhir Ben Rached and Abla Kammoun and Mohamed‐Slim Alouini and Raúl Tempone},
  title = {{An Accurate Sample Rejection Estimator for the Estimation of Outage\textbackslash\{\}n Probability of EGC Receivers}},
  year = {2019},
  eprint = {1903.05481},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/1903.05481},
}
Figure