An Implementation of GPU-Based Parallel Optimization for an Extended Uncertain Data Query Algorithm

To deal with users´ diversified query requirements on uncertain data, an uncertain data query semantic for requirement extension named RU-Topk is introduced. In the high-load application environment, the top-k query algorithm´s response time may be long. In order to satisfy performance requirements, with the consideration of the algorithm´s features, the design and implementation of GPU-based RU-Topk algorithm as well as a batch scheduling strategy are presented. Finally, the experimental results on GPU platform show that they can obtain optimized performance.