logo Use CA10RAM to get 10%* Discount.
Order Nowlogo

Blockchain technology is rapidly changing businesses\' transaction behavior and efficiency in recent years.


Abstract—Blockchain technology is rapidly changing businesses' transaction behavior and efficiency in recent years. Data privacy and system reliability are critical issues that must be addressed in the Blockchain environment. However, anomaly in- trusion poses a significant threat to a Blockchain. Therefore, this article proposes a collaborative clustering-characteristic-based data fusion approach for intrusion detection in a Blockchain-based system. A mathematical data fusion model is designed, and an AI model is used to train and analyze the data clusters in Blockchain networks. The abnormal characteristics in a Blockchain dataset are identified, a weighted combination is carried out, and the weighted coefficients among several nodes are obtained after multiple rounds of mutual competition among clustering nodes. When the weighted coefficient and a similarity-matching relationship follow a standard pattern, abnormal intrusion behavior is accurately and collaboratively detected. Experimental results show that the proposed algorithm has high recognition accuracy and promising performance in detecting attacks in a Blockchain.

Index Terms—Blockchain, intrusion detection, weighted combination, data fusion, similarity matching



HE Blockchain network user community has witnessed a rapid exponential growth   along   with   the   develop- ment of Blockchain technology. Therefore, ensuring the se- curity of Blockchain networks has become imperative[1][2]. A Blockchain is a point-to-point distributed ledger based on cryptography and a network-sharing system characterized by its disintermediation, transparency, and openness[3]. The security issue caused by the trust-based centralization model adopted by this technology needs to be addressed beforehand.

This work is supported by the National Natural Science Foundation of China under Grants 62072170, 61976087, 61872130, and 61872138, the Fundamental Research Funds for the Central Universities under Grant 531118010527, the Key Research and Development Project of China Hu- nan Science and Technology Department under Grant 2020GK2006, and 2020SK2066, and the Open Research Fund of Hunan Provincial Key Lab- oratory of Network Investigational Technology under Grant 2020WLZC001.

W. Liang is with the College of Computer Science and Electronic Engi- neering, Hunan University, Changsha 410082, China, and also with Hunan Provincial Key Laboratory of Blockchain Infrastructure and Application, Changsha 410082, China (e-mail:weiliang99@hnu.edu.cn)

L. Xiao is with Big Data Development and Research Center, Guangzhou College of Technology and Business, Guangzhou 528138, China

K. Zhang is with the Department of School of Information and Communica- tion Engineering, University of Electronic Science and Technology of China, Chengdu 611731, China

M. Tang is with School of Information Science and Technology, Guangdong University of Foreign Studies, Guangzhou 510006, China

D. He is with the College of Computer Science and Electronic Engineering, Hunan University, Changsha 410082, China

K.-C. Li is with the Department of Computer Science and Infor- mation Engineering, Providence University, Taichung 43301, Taiwan (e- mail:kuancli@pu.edu.tw) (Corresponding Author)


Transaction signatures, consensus algorithm, and cross-chain technology are utilized to ensure consistency in the distributed ledger of each transaction party, to achieve an automatic information disclosure, and to realize the principle of “account agrees with a document, account agrees with another account, and account agrees with the physical inventory.” In this way, the credit rating of a tradable product can be significantly improved, and its cost can be sharply reduced. Information users can obtain global information on company operations in real time, and their global access to such data signifies the large-scale growth of information. In this case, storing and extracting the value of information are critical. However, anomaly intrusion in Blockchain significantly threatens such information’s security and privacy; therefore, secure intrusion detection approaches should be developed.

Security detection technologies for Blockchain data have been widely used in various AI fields[4]. Nevertheless, the global financial system is exposed to security threats that may result in massive losses. For instance, some vulnerabilities have been detected in the function call of the smart contract in DAO, a crowdfunding project run by an Ethereum-based decentralized organization where 3,641,694 Ethereum coins (approximately $7.9 million USD) were transferred to private accounts in 2016[5].

The currently available security technologies, such as iden- tity authentication[6] [7], resource protection[8][9][10][11], and machine learning[12] can effectively address the security issues in Blockchain. The tamper-proofing environment of a Blockchain network requires a joint verification among all anonymous participants in any digital capital transaction. Many encryption algorithms are also utilized in Blockchain systems, and the transaction data in these systems are linked together to make the records traceable and unchangeable. Fig.1 shows the data transaction chart of consortium Blockchain, which has been used in various fields, including finance, traf- fic, and communication, to identify the normal and abnormal behavior of users. In these Blockchain-based applications, malicious third-party can invade the systems for their purposes. Nevertheless, illegal attacks can use deception to terminate the transmission of data in high-frequency data transactions. Specifically, the miner’s calculation ability is required after data consensus for huge rewards, and greedy miners always attempt to enhance their mining ability through the system. In other words, many security vulnerabilities aim to improve miners’ calculation ability and increase their profit. With the increasing number of information leakage and security events over the past years, developing a secure way for third parties to collect and control a massive amount of private data has


trusion detection scheme to address data privacy and reliability issues in Blockchain-based systems. The data fusion charac- teristic is utilized for training the datasets in Blockchain-based systems.

The remaining of this article is organized as follows. Sec- tion II presents the related work, Section III introduces the matching detection model, and the data fusion approach for collaborative anomaly intrusion detection in Section IV. Sec- tion V evaluates the performance of the proposed algorithm, and finally, Section VI summarizes the study and presents directions for future work.



Fig. 1: Data transaction chart of consortium Blockchains


become imperative.

The available intrusion detection methods for Blockchain mainly consider data characteristic attributes, data character- istic models, and joint detection. Attribute-based detection usually creates a standard network behavior model and deter- mines whether the current behavior accords with the standard model. An intrusion alarm is sent if the difference between these models exceeds a certain threshold. Despite being able to detect many new attacks, this method has a high false- alarm probability and cannot detect those intrusion attacks that pretend to be normal. Meanwhile, data characteristic models store the characteristics of known intrusion attacks in a database for misuse detection, which can lead to high detection accuracy and short response time. However, this method is unable to detect unknown intrusion attacks. The intrusion library should also be updated in real-time to ensure detection ability. Joint intrusion detection combines the advantages of misuse and anomaly detection and achieves a fairly accurate result[13]. Anomaly detection can recognize unknown attacks with a higher false alarm rate compared with misuse detection.

The analysis of these Blockchain-based intrusion detection technologies[14] reveals the following challenges:

1) The redundant transaction information of intrusion de- tection should be reduced to minimize the cost of decentralization. The optimal recognition of data clusters with high security and decentralization is selected in the Blockchain. Given that the Blockchain always requires a highly secure and energy-efficient data transaction veri- fication at the cost of resources, obtaining the minimum number of clusters and consensus costs in intrusion detection presents a challenge.

2) The intrusion detection algorithm has limited storage and cannot accurately recognize the clusters of a dataset in any case. Therefore, this algorithm cannot identify the clusters of a dataset with low-frequency fusion char- acteristics, thereby affecting cross-chain technologies’ security for intrusion detection.

3) Traditional network anomaly detection methods have low detection accuracy and speed. A high-speed, ac- curate anomaly detection is necessary to improve the real-time management of a Blockchain network.

This article proposes a data-fusion-based collaborative in-



Research on intrusion detection technologies for Blockchain-based systems remains in its infancy. Characteristic behavior analysis is an essential component of security detection given that the frequency and scale of data transactions are critical to the security of a Blockchain network. Fig. 2 shows a high-frequency intrusion detection model for a Blockchain network and reveals that digital- characteristic-attribute-based intrusion detection has poor data transmission precision and real-time performance.

Given that the available intrusion detection technologies for Blockchain networks incur high detection costs for large networks, continuously measuring the entire network’s per- formance incurs high communication costs and usually has poor timeliness. Previous studies have mainly focused on data consensus, completeness, privacy protection, and scalability and have, accordingly, proposed data consensus algorithms for Blockchain systems. However, the existing anomaly detection algorithms incur high calculation costs.

Large-scale network anomaly detection methods based on lightweight metric restoration have also been developed[15]. The singular value decomposition result of the last iteration is used to reduce the calculation cost for the current iteration. This approach realizes fast anomaly detection and is deemed more suitable than traditional anomaly detection methods for processing data in large-scale networks. Experiments show that the proposed algorithm can precisely detect the location of an anomaly in a Blockchain system and significantly reduce the calculation cost. The schemes presented in [16][17] combine deep reinforcement learning, and the authors propose a content caching technology based on Blockchain authorization to maximize system efficiency.

Several clustering methods [18][19][20][21] have also been developed in the past. In Blockchain data fusion, different algorithms are used to generate characteristic information for a dataset. These characteristics may not be repeated and can be used to match the clustering value of various fusion algo- rithms to obtain better clustering results[22]. For instance, the generation, exchange, and storage of private data in different devices in a Blockchain can be secured via the P2P feature of this Blockchain. Several operations, such as data creation, modification, and deletion, can also be registered and verified in a Blockchain to prevent illegal intrusion behaviors, includ- ing data tampering or misappropriation. Secure access control can also be implemented by customizing the Blockchain or




Fig. 2: High-frequent intrusion detection model for Blockchain network




by employing an access mechanism. Fig. 3 shows that in a Blockchain framework, several devices securely store data in different nodes without requiring human intervention, and the features of this Blockchain ensure the decentralization, authenticity, security, and privacy of these data.

The characteristic extraction of clustering data in a Blockchain depends on the amount of quantified information. Different fusion algorithms will generate various clustering characteristics, even for the same dataset. Moreover, a fusion algorithm for the same dataset will generate different clus- tering characteristic values with varying parameters. In this case, each fusion algorithm is designed based on a certain assumption[23][24][25], thereby limiting the application of these algorithms for the extraction of clustering characteristics in high-frequency transaction environments.

Given that the global distribution pattern of clustering data can be used to recognize the characteristic category of network data, Blockchain clustering-characteristic-based intrusion detection searches for a clustering structure from the sample data without the need for classification identification. Those samples within the same cluster share similar data characteristics, whereas those samples from different clusters show different characteristics. Therefore, the target data can be identified. Some studies have applied clustering-analysis- based intrusion detection. For instance, the clustering method

[26] has been used in to connect the data in to a Blockchain network.




Related Questions

. Introgramming & Unix Fall 2018, CRN 44882, Oakland University Homework Assignment 6 - Using Arrays and Functions in C

DescriptionIn this final assignment, the students will demonstrate their ability to apply two ma

. The standard path finding involves finding the (shortest) path from an origin to a destination, typically on a map. This is an

Path finding involves finding a path from A to B. Typically we want the path to have certain properties,such as being the shortest or to avoid going t

. Develop a program to emulate a purchase transaction at a retail store. This program will have two classes, a LineItem class and a Transaction class. The LineItem class will represent an individual

Develop a program to emulate a purchase transaction at a retail store. Thisprogram will have two classes, a LineItem class and a Transaction class. Th

. SeaPort Project series For this set of projects for the course, we wish to simulate some of the aspects of a number of Sea Ports. Here are the classes and their instance variables we wish to define:

1 Project 1 Introduction - the SeaPort Project series For this set of projects for the course, we wish to simulate some of the aspects of a number of

. Project 2 Introduction - the SeaPort Project series For this set of projects for the course, we wish to simulate some of the aspects of a number of Sea Ports. Here are the classes and their instance variables we wish to define:

1 Project 2 Introduction - the SeaPort Project series For this set of projects for the course, we wish to simulate some of the aspects of a number of

Ask This Question To Be Solved By Our ExpertsGet A+ Grade Solution Guaranteed

Atharva PatilComputer science

669 Answers

Hire Me
Chrisantus MakokhaComputer science

925 Answers

Hire Me

542 Answers

Hire Me

998 Answers

Hire Me

Get Free Quote!

323 Experts Online