The pre processed UAVCAN dataset.

The processing of this dataset involves the following steps. First, create a list of file paths that includes the paths of 10 binary files. Then, traverse the files one by one, check if each file exists, and if it does, read the file content. Next, for each line of data, use regular expressions to extract relevant information of the CAN message, including the label (either \Normal\ or \Attack\), the CAN ID (convert the hexadecimal value to an integer), the data length (DLC), and the data payload (Payload, convert it to a list of integers). For cases where the DLC is less than 8, fill the empty positions with 0 (which can be changed to - 1). At the same time, perform a logarithmic transformation on the CAN ID to avoid overly large values. After that, combine the information of each CAN message into a feature vector and store it in a list. Finally, convert the processed data into a Pandas DataFrame, specify the column names, and save it as a CSV file. Ultimately, the datasets are merged and saved as the can_data.csv file, which contains all the processed CAN message information.