On deviation matrices for birth-death processes

From MaRDI portal





Consider an irreducible, aperiodic, time-homogeneous and positive recurrent Markov chain with \(n\)-step transition probability \(\{p_{ij}^{(n)}\}\) and stationary distribution \(\{\pi_j\}\). Then the deviation matrix is defined by \(D_{ij}= \lim_{\alpha \uparrow 1} \sum_{n=0}^{\infty}(p_{ij}^{(n)}-\pi_j)\alpha^n\). This concept is useful in the control of multidimensional queueing systems. The authors give an algorithm for computing deviation matrices for birth-death processes. As an application, they obtain explicitly for the \(M/M/s/N\) and \(M/M/s/\infty\) queues.











This page was built for publication: On deviation matrices for birth-death processes

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5890264)