On the generalized dining philosophers problem

From MaRDI portal




Abstract: We consider a generalization of the dining philosophers problem to arbitrary connection topologies. We focus on symmetric, fully distributed systems, and we address the problem of guaranteeing progress and lockout-freedom, even in presence of adversary schedulers, by using randomized algorithms. We show that the well-known algorithms of Lehmann and Rabin do not work in the generalized case, and we propose an alternative algorithm based on the idea of letting the philosophers assign a random priority to their adjacent forks.









This page was built for publication: On the generalized dining philosophers problem

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2787669)