Constant-time snapshots with applications to concurrent data structures

Abstract

This research explores the concurrent implementation of range search operations on quadtrees using Versioning of the data structure. The proposed approach utilizes time-stamped versioned lists to maintain the state of each object, introducing a novel mechanism for timestamp attachment during updates. The algorithm employs a clone object as a global clock and Versioned quadtree. The linearizable implementation ensures constant-time operations and effective handling of concurrent snapshots. Experimental analyses compare the performance of the proposed method with state-of-the-art alternatives, showcasing its scalability, efficiency, and low overhead across diverse workloads. The results demonstrate the competitiveness of the Versioning-based approach in enabling efficient and concurrent range search operations on quad trees.