Unified Stochastic Geometry Modeling and Analysis of Cellular Networks in LOS/NLOS and Shadowed Fading

Imene Trigui, Sofiène Affes, Ben Liang

Abstract

Statistical characterization of the signal-to-interference-plus-noise ratio (SINR) via its cumulative distribution function (CDF) is ubiquitous in cellular network research, primarily because it simplifies the averaging of the Laplace transform of aggregate interference. However, this convenience often comes at the cost of being confined to simplistic Rayleigh fading models. In this work, we address the limitations of traditional models by exploiting the moment generating function (MGF) of the SINR. This approach allows us to analytically assess cellular network performance over shadowed $\kappa$-$\mu$, $\kappa$-$\mu$, and $\eta$-$\mu$ fading models. These models offer high flexibility, capturing diverse fading channels including Rayleigh, Nakagami-$m$, Rician, and Rician shadow fading. Our framework integrates coverage, achievable rate, and bit error probability (BEP) into a single analytical structure, providing a unified toolkit for analyzing dense urban environments, future femtocells, and device-to-device (D2D) shadowed channels.

Introduction

For decades, the wireless research community has relied heavily on Rayleigh fading to model communication channels. Its popularity stems from its mathematical tractability within the framework of Stochastic Geometry, which is used to model the random spatial distribution of base stations. However, as we transition toward 5G and beyond, Rayleigh fading is increasingly insufficient. Modern network deployments—such as femtocells, D2D communications, and dense urban small cells—are characterized by complex Line-of-Sight (LOS) components, non-Line-of-Sight (NLOS) paths, and significant shadowing.

While sophisticated models like $\kappa$-$\mu$ and $\eta$-$\mu$ can capture these nuances, they have historically been difficult to integrate into stochastic geometry frameworks. Previous analytical attempts often resulted in unwieldy infinite series or required heavy reliance on Monte Carlo simulations, which lack the intuitive design insights needed by network engineers. This paper proposes a shift in the analytical paradigm to bridge this gap.

Methodology: The MGF-Based Framework

The core innovation of this work is the move away from the standard CDF-based analysis of the SINR. Instead, we utilize the Moment Generating Function (MGF). This shift provides a significant mathematical advantage: it allows for a cleaner dissociation between the "tagged" user’s signal link and the aggregate interference generated by the rest of the network.

Unified Fading Modeling

We employ the shadowed $\kappa$-$\mu$ distribution as a "master" model. By adjusting its parameters, this distribution can be simplified to represent nearly any common fading scenario:

  • Rayleigh and Nakagami-$m$: Standard models for multi-path fading.
  • Rician: Captures LOS components.
  • Hoyt ($\eta$-$\mu$): Useful for satellite and mobile radio links.

Stochastic Geometry and Mathematical Tools

Base station locations are modeled as a homogeneous Poisson Point Process (PPP). To derive exact expressions for the MGF within this spatial framework, we utilize advanced mathematical constructs, including the Humbert function and Appell hypergeometric functions. This allows us to maintain analytical rigor while accounting for the randomness of node placement and the complexity of the fading environment.

Integrated Performance Metrics

Unlike previous literature that treats network metrics in isolation, this framework provides a unified approach to three critical performance indicators:

  1. Coverage Probability: The probability that a user achieves a minimum SINR threshold.
  2. Ergodic Achievable Rate: The average data rate supported by the channel.
  3. Bit Error Probability (BEP): A measure of the reliability of the transmitted data.

By deriving these from a single MGF-based foundation, we ensure consistency across different layers of network performance analysis.

Key Insights and Results

Validation and Accuracy

Extensive Monte Carlo simulations were conducted to validate the derived closed-form expressions. The results show an "excellent agreement" across a wide range of parameters, including base station density ($\lambda$), path-loss exponents ($\alpha$), and varying fading intensities. This confirms that the complex hypergeometric functions used in the framework accurately reflect physical reality.

Asymmetric Fading and Shadowing

A significant finding of this research is the framework's ability to handle asymmetric fading. In urban environments, a user may have a clear LOS to their serving base station (Rician fading) while being bombarded by interference from distant, obstructed base stations (Rayleigh or shadowed fading). Our model captures this disparity accurately. Furthermore, we observed that heavy shadowing on dominant interference components can, counter-intuitively, improve the SINR by reducing the variance of the aggregate interference.

Density Invariance

The analysis confirms that in interference-limited regimes—where thermal noise is negligible compared to the signals from neighboring cells—the average data rate becomes independent of base station density. This suggests that in dense 5G deployments, simply adding more hardware will not improve spectral efficiency without the simultaneous implementation of interference coordination or management techniques.

Conclusion and Future Directions

This work provides a robust mathematical "Swiss Army knife" for the analysis of modern cellular networks. By unifying diverse fading models under a single MGF-based stochastic geometry framework, it offers engineers a powerful tool for predicting performance in realistic environments.

While the current model is comprehensive for single-tier networks, future work will focus on extending this analysis to multi-tier heterogeneous networks (HetNets) and Multiple-Input Multiple-Output (MIMO) systems. These extensions will be vital as the industry moves toward more complex, multi-layered architectural configurations.

Summary updated

Abstract

Statistical characterization of the signal-to-interference-plus-noise ratio (SINR) via its cumulative distribution function (CDF) is ubiquitous in a vast majority of technical contributions in the area of cellular networks since it boils down to averaging the Laplace transform of the aggregate interference, a benefit accorded at the expense of confinement to the simplistic Rayleigh fading. In this work, to capture diverse fading channels that appear in realistic outdoor/indoor wireless communication scenarios, we tackle the problem differently. By exploting the moment generating function (MGF) of the SINR, we succeed in analytically assessing cellular networks performance over the shadowed {\kappa}-{\mu}, {\kappa}-{\mu} and {\eta}-{\mu} fading models. The latter offer high flexibility by capturing diverse fading channels including Rayleigh, Nakagami-m, Rician, and Rician shadow fading distributions. These channel models have been recently praised for their capability to accurately model dense urban environments, future femtocells and device-to-device (D2D) shadowed channels. In addition to unifying the analysis for different channel models, this work integrates, for the first time, the coverage, the achievable rate, and the bit error probability (BEP), which are largely treated separately in the literature. The developed model and analysis are validated over a broad range of simulation setups and parameters.
BibTeX
@article{trigui2017unified,
  author = {Imene Trigui and Sofiène Affes and Ben Liang},
  title = {{Unified Stochastic Geometry Modeling and Analysis of Cellular Networks in LOS/NLOS and Shadowed Fading}},
  journal = {IEEE Trans. Communications},
  year = {2017},
  doi = {10.1109/tcomm.2017.2732444},
}
Figure