Towards Accelerated Usage Control Based on Access Correlations.

Low run-time overhead is crucial for the practicability of usage-control mechanisms. In this article, we propose an approach to accelerate usage control by exploiting access correlations. Our approach combines two main ingredients: firstly, a technique to compute decisions ahead of time and, secondly, a method to guide selection of usage events to pre-compute decisions for. For the first, we speculatively pre-compute decisions for usage events. For the second, we exploit access correlations to identify high acceleration potential. We implemented our approach and evaluated it in a case study of security policy enforcement in a distributed storage system. Our empirical results show that the speedup is substantial. More concretely, the speedup on average is up to 61.5%.