Dataset for fuzzzing data generation based on deep advisial learning
This dataset was collected from an industrial control system running the Modbus protocol. It is used to train a deep adversarial learning model. This model is used to generate fuzzing data in the same format as the real one. The data is a sequence of hexadecimal numbers. The followed generated data is produced by the already trained model.
These decimal data can be input into the neural network for training in the form of matrix
- Training dataset raw_modbus write single register Training Data_Raw_Partial.txt (25.39 kB)
- Training dataset preprocessed_modbus write single register Training Data_Preprocessed_Partial.txt (53.59 kB)
- Generated data produced by the trained model in decimal Generated data by the trained model in decimal.txt (922.27 kB)
- Generated data produced by the trained model in hexadecimal Generated data by the trained model in hexadecimal .txt (441.77 kB)