Second order linear differential equations with a basis of solutions having only real zeros (Q6565890)

The main result of the paper is Theorem 1.3. It is geometric in nature and its assumptions are purely topological. Its statement is the following:\N\NLet \(F:\mathbb{C}\longrightarrow{\overline{\mathbb{C}}} \) be a symmetric local homeomorphism with all zeroes and poles real. Suppose that the number \(m\) of singularities of \(F^{-1}\) over points in \(\mathbb{C}^{*}\) is finite, but \(F^{-1}\) has infinitely many singularities over \(0\) and \(\infty \). Then there exists a symmetric homeomorphism \(\phi: \mathbb{C}\longrightarrow\mathbb{C} \) such that \(F_0=F\circ \phi \) is a meromorphic function, so that \(E=\frac{F_0}{F_0^{\prime}}\) is entire and has the following properties\N\N(i) If \(F \) has only finitely many zeros and poles, then \(m\geq 1\) and the order \(\rho(E)\) of \(E\) satisfies \(\rho(E) = m\).\N\N(ii) If the zeros and poles of \(F\) form an infinite one-sided sequence, then \(m \geq 2\) and the order of convergence \(\lambda (E)\) of the zeroes of \(E\) satisfy \(\lambda(E) =\rho(E) = m -1/2\).\N\N(iii) If the zeros and poles of \(F \) form an infinite two-sided sequence, then \(m\) is even, \(m \geq 4 \) and \(\lambda(E) = \rho (E) = m - 1\).\N\N(iv) The function \(A\) defined by \(4A = -2\frac{E^{\prime\prime}}{E}+\left(\frac{E^{\prime}}{E}\right)-\frac{1}{E^2}\) and the function \(E\) have the same order \(\rho = \rho(E)\), and they are of completely regular growth in the sense of Levin-Pfluger. For \(\vert \theta \vert \pi \), the indicator of \(E\) is given in case (i) by \(h_E(\theta) = c \cos \rho\theta \) with \(c\in R \ {0}\), while in case (ii) we have \( h_E(\theta) = c \sin(\rho\vert \theta\vert)\) with \(c > 0\) if the zeros are positive and \(h_E(\theta) = c \sin(\tilde{n}(\delta -\vert \theta\vert))\) with \( c > 0\) if the zeros are negative. In case (iii), \(h_E\) is given by the (now coinciding) formulae of case (ii). In all cases we have \N\[ \Nh_A = 2\max{-h_E, 0}.\N\] with some \(c > 0\).\N\N(v) All values of \(m\) indicated in (i)--(iii) can actually occur.\N\NOne of the important consequences of the above theorem concerns the number of linearly independent entire solutions \(w_1\) and \(w_2\), that have only real zeroes, of differential equation \( w^{\prime\prime}+Aw=0\), where \(A\) is a transcendental entire function. The authors prove Theorem 1.2, which strengthens some recent results of \textit{J. K. Langley} [J. Anal. Math. 141, No. 1, 225--246 (2020; Zbl 1476.30117)]. This theorem proves that given the entire function \(E=w_1w_2\) and the entire function \(A\) described in (iv) then their orders \(\rho(E)\) and \(\rho(A)\) satisfy either\N\N1) \(\rho(E)=\rho(A)=\lambda (E)=n-\frac{1}{2}\) for some natural number \(n\geq 2\) and where \(\lambda (E)\) is the exponent of convergence of one-sided, infinite sequence of real zeroes of \(E\), or\N\N2) \(\rho(E)=\rho(A)=\lambda (E)=2n-1\) for some natural number \(n\geq 2\), non-constant \(A\), and where \(\lambda (E)\) is the exponent of convergence of two-sided, infinite sequence of real zeroes of \(E\).