On the relationship between the lower order of coefficients and the growth of solutions of differential equations (Q530313)

From MaRDI portal
scientific article
Language Label Description Also known as
English
On the relationship between the lower order of coefficients and the growth of solutions of differential equations
scientific article

    Statements

    On the relationship between the lower order of coefficients and the growth of solutions of differential equations (English)
    0 references
    0 references
    0 references
    29 July 2016
    0 references
    Let \(\rho(f)=\limsup_{r\to\infty} \frac{\log \log M(r,f)}{\log r}\), \(\mu(f)=\liminf_{r\to\infty} \frac{\log \log M(r,f)}{\log r}\) denote the order and the lower of an entire function \(f\), \(f\not\equiv \mathrm{const}\), respectively, where \(M(r,f)=\max\{|f(z)|: |z|=r\}\). Consider a second-order linear differential equation \[ f''+ A(z)f'+B(z)f=0, \eqno(1) \] where \(A\) and \(B\) are entire functions. There is a heuristic principle that if \(B(z)\) dominates over \(A(z)\) on a large set, then \(B\) defines the growth of non-trivial solutions. If, in addition, \(B\) is of positive order, then any nontrivial solution must be of infinite order. The domination condition can have different forms, for instance: 1) \(\rho(A)<\rho(B)\) [\textit{G. G. Gundersen}, Trans. Am. Math. Soc. 305, No. 1, 415--429 (1988; Zbl 0634.34004); \textit{I. Laine} and \textit{P. Wu}, Proc. Am. Math. Soc. 128, No. 9, 2693--2703 (2000; Zbl 0952.34070); \textit{K.-H. Kwon} and \textit{J.-H. Kim}, Kodai Math. J. 24, No. 3, 344--351 (2001; Zbl 1005.34079)]; 2) \(\log|A(z)|=o(\log |B(z)|)\) if \(z\) belongs to some angle \(\theta_1<\arg z< \theta_2\) (see the same paper by Gundersen and [the second author et al., Ann. Acad. Sci. Fenn., Math. 40, No. 2, 985--1003 (2015; Zbl 1337.34089)]) or if \(|z|\) belongs to a set with large lower density [\textit{K.-H. Kwon}, Bull. Korean Math. Soc. 33, No. 3, 487--496 (1996; Zbl 0863.34007)]. A characteristics feature of the results obtained in the present paper is that instead of restrictions on the orders \(\rho(A) \) and \(\rho(B)\) assumptions on the lower orders \(\mu(A) \) and \(\mu(B)\) are required. Let \(\{\overline{\text{log dens}}(E)=\limsup_{r\to\infty} \frac{\int_{E\cap [1,r)} \frac {dt}t}{\log r},\) where \(E\) is a measurable subset of positive numbers. Let \(T(r,f)=\int_0^{2\pi} \log^+|f(re^{i\theta})|\, d\theta\) be the Nevanlinna characteristic of an entire function \(f\). We formulate a typical result. Theorem 1.5. Suppose that \(A(z)\) and \(B(z)\) are two entire functions with \(\mu(B)< \mu( A)<\infty\), and \(T(r, A)\sim \alpha \log M(r,A) \), \(r\to\infty\), \(r\not \in E\), where \(\overline{\text{log dens}}(E)=0\). Then every nontrivial solution \(f\) of (1) satisfies \[ \rho(f)\geq \frac{\mu(A)-\mu(B)}{21(\mu(A)+ \mu(B))\sqrt{2\pi(1-\alpha) }}-1, \quad \alpha \in (0, 1]. \]
    0 references
    0 references
    0 references
    0 references
    0 references
    complex differential equation
    0 references
    growth of solutions
    0 references
    density
    0 references
    lower order
    0 references
    0 references