Pioneering automated vulnerability detection for smart contracts in blockchain using KEVM: Guardian ADRGAN
Rohini G. Pise,Sonali Patil
DOI: https://doi.org/10.1007/s10207-024-00817-z
2024-02-29
International Journal of Information Security
Abstract:Smart contracts function like specialized computer programs on the blockchain. Many of these contracts are on Ethereum, but sometimes these contracts have problems with security. These problems caused big money losses and made the blockchain less stable. Smart contracts are self-executing with predefined rules and are at the core of many blockchain applications. However, they are susceptible to various vulnerabilities and security risks. Automated vulnerability detection helps identify and mitigate these issues efficiently. Smart Contracts (SC) have become really popular lately. People think they are the future for making deals on blockchains. Smart contracts are like automatic agreements. They work by themselves using special computer programs. They follow the rules of the deal and keep track of everything. The main idea with smart contracts is to get rid of the need for traditional trusted middlemen-like authorities or organizations. Instead, we use code that runs on a secure and unchangeable system. In this manuscript, pioneering automated vulnerability detection for smart contracts in blockchain using KEVM: Guardian ADRGAN (ADRGAN-SCB-KEVM) is proposed. Here, K framework's Ethereum virtual machine (KEVM) is a computation engine used in this research. From this KEVM, smart contracts data are provided to feature extraction phase. Feature extraction is done using Nested patch-based feature extraction. Then the extracted features are fed to attentive dual residual generative adversarial network (ADRGAN), identifying KEVM smart contracts vulnerabilities. Finally, by using ADRGAN, it is classified as Vulnerable and Non-Vulnerable in smart contracts. The proposed ADRGAN-SCB-KEVM method employed on Python and efficiency of proposed method evaluated with different metrics like Accuracy, Computation Time, Precision, Recall, F1 Score, Specificity, RoC are evaluated. The simulation outcomes prove that the proposed ADRGAN-SCB-KEVM technique attains 41.34%, 31.28%, and 36.38% higher Accuracy for Vulnerable; 32.44%, 38.45%, and 29.47% higher Accuracy for Non-Vulnerable while compared with the existing methods such as Utilizing fault injection to evaluate blockchain systems in the presence of faulty smart contracts (UFI-BS-ESVD), State-of-the-Art Blockchain-Enabled Smart Contract Applications in the University (SA-BESC-ESVD), enhancing Ethereum smart contracts static analysis by computing precise Control-Flow Graph of Ethereum bytecode (EESC-CPC-EBD), respectively.
computer science, information systems, theory & methods, software engineering