Oscillation and nonoscillation for systems of two first-order linear differential equations with delay (Q1085328)

In the first section the authors consider systems of the form \[ (1)\quad \dot x_ 1(t)=a_{11}x_ 1(t)+a_{12}x_ 2(t-\tau_ 1),\quad \dot x_ 2(t)=a_{21}x_ 1(t-\tau_ 2)+a_{22}x_ 2(t). \] Here \(a_{ij}\in R\) and \(\tau_ i>0\), \(i,j=1,2\). These systems are compared with the corresponding systems without delay: \[ (2)\quad \dot x_ 1(t)=a_{11}x_ 1(t)+a_{12}x_ 2(t),\quad \dot x_ 2(t)=a_{21}x_ 1(t)+a_{22}x_ 2(t). \] A solution \((x_ 1,x_ 2)\) of (1), or (2), is said to be oscillatory if both \(x_ 1\) and \(x_ 2\) have arbitrary large zeros in \(R_+\). Otherwise \((x_ 1,x_ 2)\) is said to be nonoscillatory. In some cases the oscillatory properties of (2) are not affected by delays, e.g., if \(a_{11}=a_{22}\) and \(a_{12}a_{21}<0\) then also all solutions of (1) are oscillatory. In many cases delay can cause or destroy oscillations. For example, if \(a_{11}+a_{22}>0\), \((a_{11}-a_{22})^ 2+4a_{12}a_{21}<0,\) then (2) is oscillatory; but for special delays, such that \(\tau_ 1+\tau_ 2\) is a given function of the coefficients \(a_{ij}\), there exists an explicitly computed nonoscillatory solution of (1). The second, larger, section of this paper deals with systems with variable coefficients \[ (3)\quad \dot x_ 1(t)=a_{11}(t)x_ 1(t)+a_{12}(t)x_ 2(t-\tau_ 1),\quad \dot x_ 2(t)=a_{21}(t)x_ 1(t-\tau_ 2)+a_{22}(t)x_ 2(t). \] The authors transform the component \(x_ 1(t)\) of (3) to a function u(t), with the same oscillatory behavior, which satisfies a selfadjoint second-order delay equation \((4)\quad (r(t)u'(t))'=P(t)r(t)u(t-\tau_ 1-\tau_ 2),\) and a similar result holds for \(x_ 2(t)\). They use and extend known results on (4) and thus obtain results on the oscillatory behavior of (3). These results are too complicated to be described here. Throughout the paper many simple examples illustrate nicely the theory.