Dataset for Categorizing Unicode and Non-Unicode Transmitted Via H.323 Protocol

To secure data transmitted through the H.323 protocol, a new data security algorithm has been introduced. This algorithm combines hybrid cryptography and real-time steganography. The hybrid cryptography includes a multilayer encryption algorithm and the RSA algorithm, while steganography involves embedding encrypted data in the 3rd LSBs of video frame pixels. To ensure data integrity and assess the reliability of the proposed algorithm using a Machine Learning model, a dataset containing 50,523 rows and 10 columns was generated. These rows and columns contain information about 50,523 Unicode and Non-Unicode characters (including all special characters) from around 159 languages. Columns were created by following the steps of the multi-layer symmetric cryptography algorithm. This means that the dataset was created by encrypting each of the 50,523 characters separately using the multi-layer cryptography algorithm and then filling up the 10 columns of each row with the encrypted result. The dataset features include Unicode_Numbers, Decimal_Values, Hex_Addition_Result, Decimal_of_Hex_Operation, XORed_Values, Linear_Arithmetic_Result_1, Linear_Arithmetic_Result_2, Type, Character_Type_in_Numerals, and Class, with Class being the dependent variable. The 'Type' column indicates the character type based on Unicode and Non-Unicode distinctions, while the 'Character_Type_in_Numerals' column contains a numeric float value based on each character's type. Class 0 is assigned to Non-Unicode characters, and Class 1 is assigned to Unicode characters. To obtain the encrypted result, a key generation algorithm creates 7 corresponding keys from a 256-bit randomly generated hexadecimal string for use in multi-layer encryption.