ESSL: Enhanced Sliding Skip List index with adaptive dynamic weights for blockchain data

Abstract

Blockchain provides trustworthy storage for the Industrial Internet of Things (IIoT), yet its inefficient data retrieval remains a critical bottleneck for real-world IIoT applications. Addressing the time-sensitive, uneven, and cyclical nature of IIoT data queries, this paper proposes an Enhanced Sliding Skip List (ESSL) index with adaptive dynamic weights for on-chain data retrieval. ESSL integrates time, frequency, and long-term trend weights into a unified dynamic framework and introduces a weight-driven reconstruction mechanism that replaces static randomness with normalized probabilistic promotion. This design achieves a balance between real-time responsiveness and long-term stability, making ESSL particularly well-suited for dynamic IIoT environments. Compared to the original Sliding Skip List (SSL), ESSL improves retrieval efficiency by 28%, enhances index stability by 30%, and reduces redundant layer adjustments by 55%. Extensive experiments across diverse query patterns — including random, range, hotspot, sequential, and hybrid queries — demonstrate that ESSL consistently outperforms SSL and other baseline methods. The proposed dynamic weight model effectively adapts to varying access patterns and ensures robust performance under fluctuating workloads, offering a promising solution for efficient and resilient data retrieval in blockchain-based IIoT systems.