[Paper Review] Some extensions of Eneström-Kakeya Theorem

N. A. Rather, Suhail Gulzar+1 more|arXiv (Cornell University)|Nov 7, 2015

Analytic Number Theory Research1 references3 citations

TL;DR

This paper extends the Eneström-Kakeya theorem for polynomials with complex coefficients by introducing a refined zero-bound estimate using parameterized inequalities on coefficient magnitudes and arguments. It improves upon prior results by providing tighter circular regions containing all zeros, especially for polynomials where the coefficient sequence changes behavior at index $k \leq n-3$, offering a sharper bound than earlier theorems through a novel combination of trigonometric and recursive coefficient constraints.

ABSTRACT

In this paper we obtain some refinements of a well-known result of Eneströ-Kakeya concerning the bounds for the moduli of the zeros of polynomials with complex coefficients which improve upon some results due to Aziz and Mohammad, Govil and Rahman and others.

Motivation & Objective

• To refine existing bounds on the moduli of polynomial zeros for complex coefficients beyond classical Eneström-Kakeya results.
• To address limitations in prior extensions that do not account for coefficient sequences with non-monotonic behavior beyond a certain index $k$.
• To improve the sharpness of zero-containing regions by incorporating both magnitude and argument constraints on coefficients.
• To generalize earlier results by Aziz, Rahman, and others through a unified framework using recursive inequalities and trigonometric bounds.

Proposed method

• Introduces a two-parameter system $t_1 \geq t_2 \geq 0$ to model coefficient decay and growth patterns in polynomial sequences.
• Applies trigonometric constraints $|\arg a_j - \beta| \leq \alpha \leq \pi/2$ to control the angular distribution of complex coefficients.
• Uses recursive inequalities: $t_1 t_2 |a_r| + (t_1 - t_2)|a_{r-1}| - |a_{r-2}| \geq 0$ for $r \leq k+1$ and $\leq 0$ for $r \geq k+2$, to segment coefficient behavior at index $k$.
• Derives a new zero-containing disk centered at $-\frac{a_{n-1}}{a_n} + (t_1 - t_2)$, with radius combining trigonometric and coefficient-weighted terms.
• Employs complex analysis techniques, including majorization and triangle inequality estimates on auxiliary functions, to bound the modulus of polynomial roots.
• Validates the bound’s superiority by showing containment of previous theorems’ zero disks within the new region for $k \leq n-3$.

Experimental results

Research questions

• RQ1 Can the Eneström-Kakeya theorem be extended to polynomials with complex coefficients while maintaining tighter zero bounds?
• RQ2 How can coefficient magnitude and argument constraints be jointly used to refine zero localization regions?
• RQ3 What is the impact of a change-point $k$ in coefficient sequence behavior on the sharpness of zero bounds?
• RQ4 Does the new bound improve upon Theorem 1.5 (Rathod et al.) for $k \leq n-3$?
• RQ5 Can the zero-containing region be expressed as a disk centered at a complex shift of the leading coefficient ratio?

Key findings

• The proposed bound (1.3) strictly contains the zero-containing region of Theorem 1.5 for $k \leq n-3$, proving a significant improvement in sharpness.
• The new bound incorporates both the argument $\alpha$ of coefficients and their magnitude distribution via $t_1, t_2$, yielding a more adaptive zero localization.
• For $|z| \leq t_1$, the zero set lies within a disk whose radius is bounded by a sum of $\cos\alpha$ and $\sin\alpha$ terms weighted by coefficient magnitudes and inverse powers of $t_1$.
• The center of the zero-containing disk is shifted from the origin to $-\frac{a_{n-1}}{a_n} + (t_1 - t_2)$, reflecting the influence of the second-highest coefficient.
• The bound remains valid even when coefficient magnitudes are not monotonic, provided the recursive inequalities on $|a_r|$ are satisfied in two segments.
• The derived inequality (3.9) ensures all zeros lie within the new disk, and the proof confirms containment of both high- and low-modulus roots under the new region.