Machine Learning Based Detection of Hidden Data in Network Packets

Transfer of secret information is made possible by the use of network steganography, which benefits from the features included in standard communication protocols. Aside from its notable benefits of concealing and transmitting confidential information, network steganography has a significant drawback because hackers may alter packets to send data or interact with the command host. Network steganography at the transport and network layers enables the seamless incorporation of cutting-edge covert channel methods. Its detection is crucial for maintaining network security and preventing potential data breaches or malicious activities that may threaten the integrity and confidentiality of network communication. Detection of attacks is often very difficult, especially with traditional tools like intrusion detection system. We propose a new method to detect based on machine learning to identify anomalous conduct of steganographic packets.