THE WAKEUP PROBLEM

We study a new problem-the wakeup problem-that seems to be fundamental in distributed computing. We present efficient solutions to the probl em and show how these solutions can be used to solve the consensus pro blem, the leader-election problem, and other related problems. The mai n question we try to answer is ''How much memory is needed to solve th e wakeup problem?'' We assume a model that captures important properti es of real systems that have been largely ignored by previous work on cooperative problems.