| The blockchain technology and its scope of application have gone beyond the scope of cryptocurrency.Through the deep integration of blockchain technology and various industries,it is possible to break through the traditional model in the field and improve the service process in multiple areas.The large-scale application of blockchain technology will bring potential to economic development and social progress,and it is even expected to become the lowest level of trust infrastructure in the digital society.However,compared with the traditional Internet,blockchain technology is still in the early stages of development,and the entire blockchain ecosystem still faces severe security threats,risks,and challenges.As the core of blockchain technology,the consensus mechanism determines the characteristics of the blockchain system such as non-tamperable,traceable,scalable,and distributed at a basic level.Through the consensus mechanism,the blockchain encourages participating nodes to contribute their own computing resources(such as storage and power)to ensure the consistency and correctness of user data on the chain.Therefore,the security of the blockchain consensus mechanism is the core issue of the entire blockchain system.The most typical blockchain consensus mechanism is the Proof of Work(Po W)mechanism proposed by Satoshi Nakamoto,which is the cornerstone of the success of blockchain platforms such as Bitcoin and Ethereum.However,the emergence of selfish mining attacks poses a threat to the security of this mechanism.This attack damages the interests of honest nodes in the system and interferes with the regular operation of the blockchain system.Therefore,it is necessary to study the characteristics of this attack and propose targeted countermeasures.First,the research on selfish mining attacks is studied,and their destructiveness to the blockchain system is analyzed.Then,based on the characteristics of selfish mining attacks and their impact on the blockchain,a selfish mining attack detection method based on machine learning models is proposed.Finally,the application scheme of the detection method is designed to provide an example of preventing selfish mining attacks and building a safe and trusted blockchain system.The main works are as follows:(1)A new attack algorithm,called Bribery Selfish Mining,that can increase the attacker’s revenue is proposed.The algorithm fuses bribery attacks with selfish mining strategies to enhance its destructive capabilities.In this paper,the blockchain is modeled as a Markov model to analyze the profitability of the attack algorithm and the impact on the state of the blockchain.Simulation experiments show that the attack not only has a lower computing power threshold but also has a higher relative benefit.Then,according to the characteristics of this kind of attack,some corresponding countermeasures are proposed.(2)A machine learning-based method for detecting selfish mining attacks is proposed.The detection method uses the classification model to detect selfish mining attacks in the blockchain.The implementation of the detection method is mainly divided into three modules: data set construction,model training,and attack detection.First,the simulator is used to simulate malicious nodes to launch selfish mining attacks on the blockchain system,and the blockchain information at the time of the episode is collected to construct a data set.Then,the classification model is trained based on the dataset,and the trained model is applied to attack detection.The realization shows that the detection method has a reasonable recognition accuracy rate,and is superior to other detection methods in terms of comprehensive performance.(3)An application scheme of selfish mining attack detection method in the medical blockchain system is designed.According to the medical field’s user needs and business characteristics,this paper constructs a medical blockchain system based on the blockchain platform.Relying on the blockchain platform,the system can realize the credible sharing of patients’ diagnoses and treatment records among different medical institutions to realize precision medicine.In addition,when building a medical blockchain system,a security control module is designed to detect attacks on the blockchain platform to prevent such attacks as selfish mining attacks and to improve the security of the blockchain system. |
