IEEE Access (Jan 2020)

SELCOM: Selective Compression Scheme for Lightweight Nodes in Blockchain System

  • Teasung Kim,
  • Sejong Lee,
  • Yongseok Kwon,
  • Jaewon Noh,
  • Soohyeong Kim,
  • Sunghyun Cho

DOI
https://doi.org/10.1109/ACCESS.2020.3044991
Journal volume & issue
Vol. 8
pp. 225613 – 225626

Abstract

Read online

With the development of blockchain technology, participants need to have huge storage volumes to deal with the growing blockchain ledger size over time. This requirement leads to the conditional participation and verification of participants, thus weakening the decentralization of a blockchain system. Several compression schemes have been proposed to mitigate this storage problem by compressing a blockchain ledger based on redundancy, modular functions, and hash functions. However, these schemes have the limitation of accumulating the compression results to validate the retained blocks. The accumulation gradually reduces the storage volume for the blockchain ledger within the storage volume of nodes with limited resources, thus reducing the verification capability of the nodes. In this paper, a selective compression scheme using a checkpoint-chain is proposed to prevent the accumulation of compression results. The checkpoint-chain is a second blockchain that stores the checkpoints compressing existing blocks through a block Merkle tree. An update process is also proposed to prevent the accumulation of checkpoints by combining them. As numerous blocks can be verified with only a few updated checkpoints, blockchain nodes with limited resources can reduce the storage volume for the blockchain ledger and achieve high verification capabilities. Finally, compared with the existing compression schemes, the proposed scheme can achieve an average reduction in the storage overhead and an average increase in the verification capability of 76.02% and 13.90%, respectively. Moreover, the corresponding performance improvements are 86.14% and 15.44% when the update process is performed, respectively.

Keywords