Group-based adaptive result certification mechanism in Desktop Grids

In Desktop Grids, volunteers (i.e, resource providers) have heterogeneous properties and dynamically join and leave during execution. Moreover, some volunteers may behave erratically or maliciously. Thus, it is important to detect and tolerate erroneous results (i.e., result certification) in order to guarantee reliable execution, considering volatility and heterogeneity in a scheduling procedure. However, existing result certification mechanisms do not adapt to such a dynamic environment. As a result, they undergo high overhead and performance degradation. To solve the problems, we propose a new Group-based Adaptive Result Certification Mechanism (GARCM). GARCM applies different result certification and scheduling algorithms to volunteer groups that are constructed according to their properties such as volunteering service time, availability and credibility. © 2009 Elsevier B.V. All rights reserved.