D3-Tree: A Dynamic Distributed Deterministic Load - Balancer for decentralized tree structures

TL;DR

D3-Tree is a dynamic, deterministic, distributed load-balancer for decentralized tree structures that improves load balancing efficiency, fault tolerance, and performance, and introduces the ART+ structure with sub-logarithmic complexity.

Contribution

The paper introduces D3-Tree, a novel deterministic load-balancing structure for decentralized networks, and proposes ART+ with sub-logarithmic query and update performance.

Findings

D3-Tree achieves logarithmic amortized load-balancing complexity.

D3-Tree demonstrates high fault tolerance even with massive node failures.

ART+ attains sub-logarithmic performance for queries and updates.

Abstract

In this work, we propose D3-Tree, a dynamic distributed deterministic structure for data management in decentralized networks. We present in brief the theoretical algorithmic analysis, in which our proposed structure is based on, and we describe thoroughly the key aspects of the implementation. Conducting experiments, we verify that the implemented structure outperforms other well-known hierarchical tree-based structures, since it provides better complexities regarding load-balancing operations. More specifically, the structure achieves a logarithmic amortized bound, using an efficient deterministic load-balancing mechanism, which is general enough to be applied to other hierarchical tree-based structures. Moreover, we investigate the structure's fault tolerance, which hasn't been sufficiently tackled in previous work, both theoretically and through rigorous experimentation. We prove that D3-Tree is highly fault tolerant, since, even for massive node failures, it achieves a significant success rate in element queries. Afterwards we go one step further, in order to achieve sub-logarithmic complexity and propose the ART+ structure (Autonomous Range Tree), exploiting the excellent performance of D3-Tree. ART+ is a fully dynamic and fault-tolerant structure, which achieves sub-logarithmic performance for query and update operations and performs load-balancing in sub-logarithmic amortized cost.