This paper presents a reliable model for mobile codes in distributed networks, which represents reliable mobile agent execution. The model ensures non-blocking mobile agent execution and forces the once property without relying on correct fault detection. A mobile agent More
This paper presents a reliable model for mobile codes in distributed networks, which represents reliable mobile agent execution. The model ensures non-blocking mobile agent execution and forces the once property without relying on correct fault detection. A mobile agent execution is blocking if a fault of agent prevents the agent from continuing in its execution. The once problem is related to non-blocking in the sense that solutions to the latter may lead to multiple executions of the mobile agent. A solution to reliable mobile agent execution needs to ensure both the non-blocking and once properties. The analytical results show new theoretical perceptions into the statistical behaviors of mobile agents and provide useful tools for executing mobile agents in networks. The results show that agents' behavior is influenced by places' characteristics and the agents' behavior can be managed to network. In this paper, we analyzed the average time consuming of mobile agents between two places. The approach, Fault-Tolerant approach for mobile codes offers a user-transparent fault tolerance which can be selected by the user for every single application given to the environment. Thereby, the user can decide for every application weather it has to be treated fault-tolerant or not. We proposed a reliable execution model of mobile codes and analyzed the life expectancy, including the average time consuming of mobile agents between two places, the average number of places agents will visit, and the agents' life expectancy.
Manuscript profile
