Mobile agent based security enhancements in ad hoc grid environment

The mobile agent paradigm and security issues in the grid has been successfully used in large number of research areas separately, and specifically contributes to the research endeavors. The proposed paper combines both the technologies and provides avenues for further research. Particularly in ad hoc grid environment, resources are not always available as the nodes can spontaneously connect and disconnect at any time. Thus, these environments demand the correct execution of process for improving the performance criteria. However, there are some intruders that affect the normal operation of ad hoc grids. Therefore, it is essential to use punishment procedures based on trust models. The existing work has proven that the RETENTION a reactive trust based mechanism detected and punished malicious nodes in ad hoc grid environments, without generating any false-positives. We propose a mobile agent based trust model in ad hoc grid environment for improving the performance considerations and security issues. Simulation results demonstrate the effectiveness of the proposed work, to detect and punish nodes having malicious behavior, such as selfish nodes, injected nodes and so on. We observe that the behavior of the proposed model is comparatively good in a scenario where the types of attack implemented in different nodes.