Design And Implementation Of IPv6 Address Scanning Target Generation System Based On Blockchain And Federated Learning

IP address scanning is to obtain the network information of the target address by accessing the common ports.With this network information,people can view IP availability,subnet utilization,dynamic IP assignment,and switch port mapping.It plays an important role in communication data encryption,vulnerability repair,and tracking malicious attackers.But the space of available IPv6 address is very large,and the density of active IPv6 address is very low.Current technology infer a relatively high density of active IPv6 address areas based on known active IPv6 address.Then they generate the IPv6 address of scanning target.However,due to the small size of the dataset,they perform poorly under the multiple prefixes of IPv6 address.In response to the problems described above,thesis proposes an algorithm for IPv6 address scanning target generation based on blockchain and federated learning.It can federate multiple enterprises with users’ data of IPv6 for distributed machine learning.It solves the problem that most enterprises have insufficient data to support the training of efficient machine learning models.By using blockchain and federated learning technology,multiple participants can jointly participate in active IPv6 address model training without uploading the datasets of active IPv6 address owned by others.It can protect the original data privacy and prevent the central server that federated learning is highly dependent on overload or attack.It prevents malicious parties from uploading false information and enhances the security and stability of the system.Thesis uses the address datasets of public data sources as input to test the method.The experimental results show that the training effect of the algorithm model can be improved by increasing the number of datasets by adding participants.When each participant has a low amount of data,compared to other techniques using their own dataset,the algorithm can increase the number of active IPv6 addresses generated by up to 9.36 times and new active IPv6 addresses by up to 9.12 times.Its hit rate increased by 12.02%,and the new address generation rate increased by 8.54%.Based on the above algorithm,thesis designs and implements an IPv6 address scanning target generation system based on blockchain and federated learning.This system is mainly divided into data acquisition layer,data processing layer and data display layer.The system collects active IPv6 address data through the data collection layer.And its data processing layer realizes the generation of IPv6 address scanning targets.It realizes the interactive operation with the user through the data display layer and displays the output information of each stage in the process of generating the IPv6 address scanning target in a visual form.