Scaling TrustChain to One Million Blocks on Mobile Devices
Storage Performance Evaluation and Benchmarking
M.S. Bakker (TU Delft - Electrical Engineering, Mathematics and Computer Science)
Johan Pouwelse – Mentor (TU Delft - Data-Intensive Systems)
B. Nasrulin – Mentor (TU Delft - Data-Intensive Systems)
K.G. Langendoen – Graduation committee member (TU Delft - Embedded Systems)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
Smartphones offer limited storage and memory, con- straints that conventional blockchains struggle to meet, yet they are also the devices where user- owned transaction chains promise the most value. We present the first empirical study of TrustChain, a DAG-based per-peer ledger running entirely on mobile hardware. A Rust implementation and open- source benchmark evaluate how flush-interval batch- ing (k) and lossless compression affect a single node scaling from 103 to 106 blocks.
On a Galaxy S8 and Pixel-6 emulator, RAM stays below 600 MB and compressed disk use below 0.5 GB at one million blocks (128 B payload). Insert la- tency remains interactive (< 8 ms) with disk:100 flushing; moderate batching (k ≈ 500) cuts CPU load by roughly 45 % without harming durabil- ity. Lightweight compression (LZ4-1, Zstd-1) trims space by 20 to 30 % at a sub-10 ms cost, with diminishing returns at higher levels. End-to-end tests show storage is never the bottleneck, raw UDP achieving a 7 ms median RTT.
Taken together, the implementation and measure- ment dataset provide a concrete reference for deploy- ing DAG-based chains on smartphones and highlight opportunities for advancing mobile blockchain tech- nology.