Quasi-convexity of the asymptotic channel MSE in regularized semi blind estimation

TL;DR

This paper proves the quasi-convexity of a sum of quadratic fractions, ensuring a unique global minimum, which is crucial for optimizing regularized semi-blind channel estimation and related estimation problems.

Contribution

It provides the first theoretical proof of the quasi-convexity and uniqueness of the minimum for a specific sum of quadratic fractions used in semi-blind estimation.

Findings

The sum of quadratic fractions is quasi-convex on the positive real axis.

The function has a unique global minimum, confirmed theoretically.

This result supports optimization in semi-blind channel identification.

Abstract

In this paper, the quasi-convexity of a sum of quadratic fractions in the form $\sum_{i=1}^n \frac{1+c_i x^2}{\left(1+d_ix\right)^2}$ is demonstrated where $c_i$ and $d_i$ are strictly positive scalars, when defined on the positive real axis $\mathbb{R}^{+}$. It will be shown that this quasi-convexity guarantees it has a unique local (and hence global) minimum. Indeed, this problem arises when considering the optimization of the weighting coefficient in regularized semi-blind channel identification problem, and more generally, is of interest in other contexts where we combine two different estimation criteria. Note that V. Buchoux {\it et.al} have noticed by simulations that the considered function has no local minima except its unique global minimum but this is the first time this result, as well as the quasi-convexity of the function is proved theoretically.