BLAS: A Blockchain-Enabled Efficient and Verifiable Log Audit System With Hybrid Storage

A reliable log audit system is a fundamental tool for efficient security management and attack detection. Blockchain has emerged as a prominent technology for building log audit systems, thanks to its non-repudiation and immutability properties. However, current blockchain-based solutions are computationally intensive and typically rely on coarse-grained queries, making them impractical for real-world systems. Therefore, we propose a blockchain-based verifiable log audit system, BLAS, which adopts three novel techniques. Firstly, we propose a novel data structure called the Index-Object Merkle Forest (IOMF), which combines modified Merkle Tree data structures and keyword-range bitmap indexes to support efficient log auditing. Secondly, we propose a hierarchical extension of IOMF to create an Authenticated Layered Index Structure (ALIS). ALIS enables fine-grained auditing at the entry level. Finally, we propose two optimization techniques in ALIS to reduce computational and communication costs. The performance evaluation confirms that BLAS is consistently faster than the baseline solutions with similar settings in various experiments, achieving speedups of hundreds to over ten thousand times for the most challenging workloads.