Exponential stability tests for linear delayed differential systems depending on all delays (Q2333296)

The paper presents a sufficient condition for the uniform exponential stability of the system of linear delay differential equations \[ \dot x_i = \sum_{j=1}^{m} \sum_{k=1}^{r_{ij}} a_{ij}^k (t) x_j \left( h_{ij}^k(t) \right), \] where \(a_{ij}^k:[0,\infty)\to\mathbb{R}\) and the time-varying delays \(h_{ij}^k\) are Lebesgue measurable functions such that \(t-h_{ij}^k(t) \le \tau_{ij}^k\) with \(\tau_{ij}^k>0\) for all \(t\ge 0\). One of the ingredients of the obtained sufficient condition is that the logarithmic norm \(\mu_A (t)\) of the matrix \(A(t) = \left( \sum_{k=1}^{r_{ij}} a_{ij}^k (t) \right)_{i,j=1}^m\) is negative and uniformly bounded from zero. Another ingredient is more technical, but it holds for sufficiently small \(\tau_{ij}^k\). Roughly speaking, the theorem extends the exponential stability result for zero delays, when \(\tau_{ij}^k=0\), to the case of sufficiently small delays.