Collecting mobile-agent garbage using actor-based weighted reference counting

In the past decade, the number of distributed computing paradigms and systems has been grown up rapidly due to the reduced cost of computing resources and the popularization of the Internet. Among those computing paradigms and systems, mobile agent-based systems are becoming prevalent because they are more flexible and more reconfigurable at runtime. Applications running on a mobile-agent system may produce mobile-agent garbage, which wastes system resources and eventually becomes harmful to system performance and reliability. The major challenge of the problem is that, the mobility property and the decentralization nature make it hard to trace and to collect mobile-agent garbage automatically. In this paper, we use the actor garbage collection model to theoretically depict how mobile-agent garbage is collected. We propose an efficient distributed garbage collecting mechanism called actor-based weighted reference counting algorithm to collect garbage of mobile agents. The experimental results show that the proposed algorithm is efficient than a recent similar method.