AdaChain: A Learned Adaptive Blockchain

Chenyuan Wu; Bhavana Mehta; Mohammad Javad Amiri; Ryan Marcus; Boon Thau Loo

doi:10.14778/3594512.3594531

AdaChain: A Learned Adaptive Blockchain

Chenyuan Wu
, Bhavana Mehta
, Mohammad Javad Amiri
, Ryan Marcus
, Boon Thau Loo

Research output: Journal Publications and Reviews › RGC 21 - Publication in refereed journal › peer-review

18 Citations (Scopus)

3 Downloads (CityUHK Scholars)

Abstract

This paper presents AdaChain, a learning-based blockchain framework that adaptively chooses the best permissioned blockchain architecture to optimize effective throughput for dynamic transaction workloads. AdaChain addresses the challenge in Blockchainas- a-Service (BaaS) environments, where a large variety of possible smart contracts are deployed with different workload characteristics. AdaChain supports automatically adapting to an underlying, dynamically changing workload through the use of reinforcement learning. When a promising architecture is identified, AdaChain switches from the current architecture to the promising one at runtime in a secure and correct manner. Experimentally, we show that AdaChain can converge quickly to optimal architectures under changing workloads and significantly outperform fixed architectures in terms of the number of successfully committed transactions, all while incurring low additional overhead.

Copyright is held by the owner/author(s). Publication rights licensed to the VLDB Endowment.

Original language	English
Pages (from-to)	2033-2046
Journal	Proceedings of the VLDB Endowment
Volume	16
Issue number	8
DOIs	https://doi.org/10.14778/3594512.3594531
Publication status	Published - Apr 2023
Externally published	Yes
Event	49th International Conference on Very Large Data Bases (VLDB 2023) - Vancouver, Canada Duration: 28 Aug 2023 → 1 Sept 2023 https://vldb.org/2023/

Funding

We thank the anonymous reviewers for their insightful feedback and suggestions. This work is funded by NSF grants CNS-2104882, and CNS-1703936 and by NSF/Intel joint grant #2011168.

Publisher's Copyright Statement

This full text is made available under CC-BY-NC-ND 4.0. https://creativecommons.org/licenses/by-nc-nd/4.0/

Access Output

10.14778/3594512.3594531

283055872.pdfFinal Published version, 1.32 MBLicence: CC BY-NC-ND

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

@article{03c8bfd6152e41c7ac890e1be4bf1b6a,

title = "AdaChain: A Learned Adaptive Blockchain",

abstract = "This paper presents AdaChain, a learning-based blockchain framework that adaptively chooses the best permissioned blockchain architecture to optimize effective throughput for dynamic transaction workloads. AdaChain addresses the challenge in Blockchainas- a-Service (BaaS) environments, where a large variety of possible smart contracts are deployed with different workload characteristics. AdaChain supports automatically adapting to an underlying, dynamically changing workload through the use of reinforcement learning. When a promising architecture is identified, AdaChain switches from the current architecture to the promising one at runtime in a secure and correct manner. Experimentally, we show that AdaChain can converge quickly to optimal architectures under changing workloads and significantly outperform fixed architectures in terms of the number of successfully committed transactions, all while incurring low additional overhead. Copyright is held by the owner/author(s). Publication rights licensed to the VLDB Endowment.",

author = "Chenyuan Wu and Bhavana Mehta and Amiri, \{Mohammad Javad\} and Ryan Marcus and Loo, \{Boon Thau\}",

year = "2023",

month = apr,

doi = "10.14778/3594512.3594531",

language = "English",

volume = "16",

pages = "2033--2046",

journal = "Proceedings of the VLDB Endowment",

issn = "2150-8097",

publisher = "Very Large Data Base Endowment Inc.",

number = "8",

note = "49th International Conference on Very Large Data Bases (VLDB 2023), VLDB 2023 ; Conference date: 28-08-2023 Through 01-09-2023",

url = "https://vldb.org/2023/",

}

TY - JOUR

T1 - AdaChain

T2 - 49th International Conference on Very Large Data Bases (VLDB 2023)

AU - Wu, Chenyuan

AU - Mehta, Bhavana

AU - Amiri, Mohammad Javad

AU - Marcus, Ryan

AU - Loo, Boon Thau

PY - 2023/4

Y1 - 2023/4

N2 - This paper presents AdaChain, a learning-based blockchain framework that adaptively chooses the best permissioned blockchain architecture to optimize effective throughput for dynamic transaction workloads. AdaChain addresses the challenge in Blockchainas- a-Service (BaaS) environments, where a large variety of possible smart contracts are deployed with different workload characteristics. AdaChain supports automatically adapting to an underlying, dynamically changing workload through the use of reinforcement learning. When a promising architecture is identified, AdaChain switches from the current architecture to the promising one at runtime in a secure and correct manner. Experimentally, we show that AdaChain can converge quickly to optimal architectures under changing workloads and significantly outperform fixed architectures in terms of the number of successfully committed transactions, all while incurring low additional overhead. Copyright is held by the owner/author(s). Publication rights licensed to the VLDB Endowment.

AB - This paper presents AdaChain, a learning-based blockchain framework that adaptively chooses the best permissioned blockchain architecture to optimize effective throughput for dynamic transaction workloads. AdaChain addresses the challenge in Blockchainas- a-Service (BaaS) environments, where a large variety of possible smart contracts are deployed with different workload characteristics. AdaChain supports automatically adapting to an underlying, dynamically changing workload through the use of reinforcement learning. When a promising architecture is identified, AdaChain switches from the current architecture to the promising one at runtime in a secure and correct manner. Experimentally, we show that AdaChain can converge quickly to optimal architectures under changing workloads and significantly outperform fixed architectures in terms of the number of successfully committed transactions, all while incurring low additional overhead. Copyright is held by the owner/author(s). Publication rights licensed to the VLDB Endowment.

UR - https://www.scopus.com/pages/publications/85161625769

UR - https://www.scopus.com/pages/publications/85161625769#tab=impact

U2 - 10.14778/3594512.3594531

DO - 10.14778/3594512.3594531

M3 - RGC 21 - Publication in refereed journal

SN - 2150-8097

VL - 16

SP - 2033

EP - 2046

JO - Proceedings of the VLDB Endowment

JF - Proceedings of the VLDB Endowment

IS - 8

Y2 - 28 August 2023 through 1 September 2023

ER -

AdaChain: A Learned Adaptive Blockchain

Abstract

Funding

Publisher's Copyright Statement

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

Cite this