A load balancing skyline query algorithm in high bandwidth distributed systems

Skyline queries, which retrieve the points that are not dominated by any other points in a given dataset, are well recognized as a powerful tool in multi-criteria decision making. Most of the previous works focus on computing skylines on centralized environments typically with one CPU. In this paper, to scale up skyline computation on large datasets, we propose a load balancing parallel skyline query algorithm for multi-processor clusters or high-speed network environments. Our algorithm utilizes the monotonic ordering and clustering properties of Z-order to reduce dominance tests and significantly shortens the response time by performing parallel processing over multiple processors. Our algorithm is progressive, fast convergent and scalable. Extensive experiments were conducted to demonstrate the feasibility and effectiveness of the proposed algorithm.