A Trust Model for SEUGrid: a Grid Platform for Monte Carlo Simulation

The trust in grid is becoming increasingly more and more popular. Trust has been recognized as an important factor for grid security. In the SEUGrid, the security strategies are almost static, contrary to the dynamicity of the grid environment. Considering this problem, this paper proposes a trust model imitating the trust system in real world. The primary application goal of SEUGrid is to run AMS-02 Monte Carlo (MC) simulation through the cooperative work of the resources in the Southeast University. The alternative sides of MC simulation: computing and storing, lay different emphasis on the capability of node. So the trust model presents different computing methods of trust value in the two contexts based on the updating strategy. The model also computes indirectly the trust value between grid entities via VO. Finally, the simulation results present that the trust model can make the SEUGrid more secure and flexible