Search

A Mobile agent is a software program with so high mobility that it is difficult to transmit and receive messages accurately. At present, the location management of mobile agent consists of tracking phase and message delivery phase, both of which dutifully create or incur plenty of update messages and interaction messages separately in order to guarantee the effectiveness of mobile agent location management scheme. However, the transmission of messages not only affects the overall performance of mobile agent location management scheme, but also causes the network overhead. Therefore, this paper proposes a ratio-based update scheme that can minimize the overall transmission cost of produced messages. A number of simulation experiments are conducted to evaluate the proposed scheme. And, as a consequence, the simulation results reveal that this scheme is general, adaptable, and possessed of higher granularity in the network latency; moreover, the results also indicate that this scheme is better than previous schemes in optimizing the caused overall performance overhead.

The location management ability of a mobile agent system which enables a mobile agent to interact with others accurately by message passing beyond doubts stands important for the mobile agent system. At present, the location management of mobile agent consists of tracking phase and message delivery phase, both of which dutifully create or incur plenty of update messages and interaction messages separately in order to guarantee the effectiveness of mobile agent location management scheme. However, the transmission of messages not only affects the overall performance of mobile agent location management scheme, but also causes the network overhead. Previous studies have devoted to achieve a balance of the trade-off between the costs in tracking phase and message delivery phase in designing a mobile agent location management scheme. This paper argues that the interaction messages in message delivery phase produce the major overhead of overall performance. Therefore, it focuses on investigating how to reduce the performance overhead in message delivery phase and proposes an intelligent update algorithm, the I-Update, for this purpose. Simulation results show that keeping the length of a channel as short as possible leads to the optimum of overall performance.

This paper proposed a fault tolerant mechanism to suppress and exterminate file pollution attacks; the idea has been originated from the concept that P2P file sharing network is now in a form of global file storage. The fault tolerant mechanism harnesses an approach that is adapted from the EVENODD Coding Scheme used in the disk storage technology, which tolerates effectively disk failures in the RAID (Redundant Array of Independent Disks) system. The Fluid Models are extended in this paper to analyze and evaluate the proposed anti-pollution mechanism. The accuracy (correctness) of the model has been proved by performing several simulation experiments, which yield the results that show the proposed mechanism could effectively suppress the pollution and could shorten 40~60% of the polluted time that a P2P file sharing network may be exposed to.

A location management scheme is critical for a mobile agent system to trace mobile agents and support on-the-go communications between peer agents. For the time being, most mobile agent location management schemes consist of both a tracking phase and a message delivery phase, and dutifully create or incur plenty of update messages and interaction messages in order to guarantee the effectiveness of the schemes. Excessive transmission of messages not only downgrades the overall performance of mobile agent location management scheme, but also causes network overhead. This paper proposes a ratio-based update scheme that can minimize the overall transmission costs of produced messages by intelligently adjusting the ratio between update messages and forwarding pointers. Experiments are conducted to evaluate the proposed scheme, which reveal that this scheme is general, adaptable, and possessed of higher granularity in the network latency. Moreover, the results also indicate that this scheme is better than previous schemes in optimizing the overall performance.