RFC: Block-Range Filters (a.k.a. Hierarchical filters)

The proposal to introduce block-range filters as an optimization technique for compact block filters (BIP-158) aims to reduce the total download size required when seeking transactions within blockchain blocks. The concept involves creating filters for a specified range of blocks, such as every 256 blocks, which are downloaded initially. If these block-range filters indicate potential matches with the sought scripts, only then are the individual block filters downloaded. This method shows promise in significantly reducing download sizes, with preliminary simulations suggesting possible reductions to less than 25% of the total.

In detailed simulation scenarios, varying the range size demonstrated that total block-range filter size decreases as the range size increases. For instance, for a 256-block range, the filter size was about 14.8% relative to the total size of individual block filters, with further reductions observed in larger ranges. However, there are practical limits to the optimal range size; too large a range could lead to inefficiencies and management issues, indicating that there might be a sweet spot that maximizes efficiency without adding undue complexity.

The impact on download sizes varies depending on the transaction count within the scripts being analyzed. In cases with very low transaction counts, the download size could be reduced to around 18.2% for a 256-block range. Simulations also highlighted that while increasing the range size generally reduces the filter size, it eventually reaches a point where the benefits diminish or reverse, suggesting the existence of an optimal range size for efficiency gains.

Further exploration into this topic could involve experimenting with different filter parameters to find the most effective settings for block-range filters. Additionally, refining the approach to use optimized block filters, potentially by incorporating script indexes instead of original script hashes, could offer further reductions in size and enhance overall effectiveness.

For more details on alternative filtering techniques and discussions related to BIP-158, interested readers can refer to the discussion on Binary fuse filters, which provides insights into other potential optimizations (Binary Fuse Filters Discussion). These ongoing explorations and discussions underscore the dynamic nature of blockchain technology development, where efficiencies are continually sought through innovative technical strategies.