A flexible method to defend against computationally resourceful miners in blockchain proof of work

Research output: Contribution to journalArticle


  • Wei Ren
  • Jingjing Hu
  • Tianqing Zhu
  • Yi Ren
  • Kim-kwang Raymond Choo

Organisational units


Blockchain is well known as a decentralized and distributed public digital ledger, and is currently used by most cryptocurrencies to record transactions. One of the fundamental differences between blockchain and traditional distributed systems is that blockchain's decentralization relies on consensus protocols, such as proof of work (PoW). However, computation systems, such as application specific integrated circuit (ASIC) machines, have recently emerged that are specifically designed for PoW computation and may compromise a decentralized system within a short amount of time. These computationally resourceful miners challenge the very nature of blockchain, with potentially serious consequences. Therefore, in this paper, we propose a general and flexible PoW method that enforces memory usage. Specifically, the proposed method blocks computationally resourceful miners and retains the previous design logic without requiring one to replace the original hash function. We also propose the notion of a memory intensive function (MIF) with a memory usage parameter k (kMIF). Our scheme comprises three algorithms that construct a kMIF Hash by invoking any available hash function which is not kMIF to protect against ASICs, and then thwarts the pre-computation of hash results over a nonce. We then design experiments to evaluate memory changes in these three algorithms, and the findings demonstrate that enforcing memory usage in a blockchain can be an effective defense against computationally resourceful miners.


Original languageEnglish
Pages (from-to)161-171
Number of pages11
JournalInformation Sciences
Early online date13 Aug 2019
Publication statusPublished - 1 Jan 2020


    Research areas

  • ASICs, Blockchain, Hash, Pow

View graph of relations

ID: 165087729

Related by author
  1. Invariant Deep Compressible Covariance Pooling for Aerial Scene Categorization

    Research output: Contribution to journalArticle

  2. On Optimizing Signaling Efficiency of Retransmissions for Voice LTE

    Research output: Contribution to conferencePaper

  3. ID-Free multigroup cardinality estimation for massive RFID Tags in IoT

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

Related by journal
  1. Dual-Verification Network for Zero-shot Learning

    Research output: Contribution to journalArticle

  2. PATCH-IQ: A patch based learning framework for blind image quality assessment

    Research output: Contribution to journalArticle

  3. Content-based retrieval of human actions from realistic video databases

    Research output: Contribution to journalArticle