In this paper, we present a learning support system DASE-E to help students understand fundamental concepts of distributed algorithms in computer science. DASE-E is a collaborative learning system, in which the task of students is to devise a distributed algorithm. DASE-E offers a set of small wireless terminals with accelerometers. Each student plays the role of a process with a terminal, according to the algorithm that students devised. Each terminal enables a student to take physical actions that control the behavior of a process in the simulator. After the role playing simulation is finished, students discuss their activity played back on a screen. We imp1emented the system for learning distributed leader election algorithms, had a trial exercise in our research group, and we confirmed that it is effective to learn the critical part of distributed systems and algorithms.