An MGF-Based Unified Framework to Determine the Joint Statistics of Partial Sums of Ordered i.n.d. Random Variables

Sung Sik Nam, Hong-Chuan Yang, Mohamed-Slim Alouini, Dong In Kim

Introduction

In the design and analysis of wireless communication systems, diversity combining techniques are essential for mitigating the effects of fading. These techniques often involve ranking available signal paths by strength and selecting a subset for processing—a process governed by order statistics. While the joint statistics of ordered independent and identically distributed (i.i.d.) random variables (RVs) are well-documented, the i.i.d. assumption rarely holds in real-world environments. Factors such as varying distances between transmitters and receivers, physical obstacles, and unbalanced hardware lead to signal paths that are independent but non-identically distributed (i.n.d.).

This paper presents a unified analytical framework using Moment Generating Functions (MGFs) to determine the joint statistics of partial sums of ordered i.n.d. random variables. By moving beyond the i.i.d. constraint, this research provides a more accurate mathematical toolset for modeling complex, frequency-selective fading channels.

The Mathematical Challenge of i.n.d. Variables

The primary difficulty in analyzing ordered variables lies in their inherent dependence. Even if the underlying random variables are independent, the act of ordering them (e.g., $X_{(1)} \geq X_{(2)} \geq \dots \geq X_{(N)}$) introduces a statistical relationship where the value of one rank constrains the possible values of another.

In the i.i.d. case, symmetry simplifies the integration process. However, in the i.n.d. case, researchers must account for every possible permutation of which specific variable occupies which rank. Traditionally, this leads to multi-fold nested integrals that are computationally prohibitive and analytically opaque. This paper addresses this "combinatorial explosion" by introducing a streamlined MGF-based approach.

The Unified MGF-Based Framework

The authors propose a two-step methodology to derive the Probability Density Functions (PDFs) of partial sums without relying on direct multi-dimensional integration.

Step I: The MGF Domain and Mixture Functions

Instead of working directly with spatial distributions, the framework first derives the joint MGF. The authors introduce a set of "mixture functions" ($c_{i_l}, e_{i_l}, \mu_{i_l}$) that combine Cumulative Distribution Functions (CDFs) or Exceedance Distribution Functions (EDFs) with exponential MGF terms. This allows the complex interaction between different non-identical distributions to be represented as compact summations. To manage the complexity of non-identical variables, the framework utilizes power set notation, which systematically tracks the contributions of different variable subsets to the ranked sums.

Step II: Inverse Laplace Transform

Once the joint MGF is established in a compact form, the framework applies an inverse Laplace transform to return to the probability domain. This step yields the joint PDF of the partial sums. For the exponential distribution—which is the standard model for Rayleigh fading—the authors provide ready-to-use closed-form expressions, eliminating the need for numerical integration entirely.

Key Insights and Applications

Flexibility in Summation

Unlike previous models that might only consider the sum of the "top $N$" paths, this framework is generalized. It can determine the statistics for arbitrary partial sums, such as the sum of the 2nd through 5th strongest paths. This flexibility is critical for analyzing interference-cancellation schemes and specific sub-optimum combining receivers.

Performance Analysis of GSC RAKE Receivers

To demonstrate the framework's utility, the authors apply it to Generalized Selection Combining (GSC) RAKE receivers. These receivers operate over frequency-selective fading channels with non-uniform power delay profiles (a classic i.n.d. scenario). The paper derives the exact capture probability for these systems, providing engineers with a concrete method to predict performance in urban environments where signal paths are naturally unbalanced.

Limitations and Future Directions

While the framework significantly simplifies the analytical process, it is not without challenges:

  • Computational Scaling: As the number of variables ($N$) increases, the number of terms in the power set summations grows significantly, which may impact computational efficiency for very high-order diversity systems.
  • Distribution Constraints: While the framework is unified, finding closed-form solutions for distributions other than the exponential may still require numerical contour integration (the Bromwich integral) during the inverse Laplace transform step.

Future research may extend this framework to multiuser scheduling and adaptive transmission techniques, where the ability to handle non-identical path statistics is increasingly vital for next-generation network optimization.

Summary updated

Abstract

The joint statistics of partial sums of ordered random variables (RVs) are often needed for the accurate performance characterization of a wide variety of wireless communication systems. A unified analytical framework to determine the joint statistics of partial sums of ordered independent and identically distributed (i.i.d.) random variables was recently presented. However, the identical distribution assumption may not be valid in several real-world applications. With this motivation in mind, we consider in this paper the more general case in which the random variables are independent but not necessarily identically distributed (i.n.d.). More specifically, we extend the previous analysis and introduce a new more general unified analytical framework to determine the joint statistics of partial sums of ordered i.n.d. RVs. Our mathematical formalism is illustrated with an application on the exact performance analysis of the capture probability of generalized selection combining (GSC)-based RAKE receivers operating over frequency-selective fading channels with a non-uniform power delay profile.
BibTeX
@article{nam2014mgf,
  author = {Sung Sik Nam and Hong-Chuan Yang and Mohamed-Slim Alouini and Dong In Kim},
  title = {{An MGF-Based Unified Framework to Determine the Joint Statistics of Partial Sums of Ordered i.n.d. Random Variables}},
  journal = {IEEE Trans. Signal Process.},
  year = {2014},
  url = {https://openreview.net/forum?id=IbOUw_51SOi},
}
Figure