Abstract:
Deep Packet Inspection (DPI) has transformed network security and analysis by allowing for in-depth content study of data packets as they traverse network infrastructures. DPI goes beyond security to help with network optimization, QoS control, and policy enforcement. P4 (Programming Protocol-Independent Packet Processor) in network devices provides programmable data plane operations, making it well-suited for DPI requirements. Traditionally, DPI relied on fixed hardware, making it difficult to respond to changing threats and network dynamics. With the introduction of P4 and Software-Defined Networking (SDN), new degrees of network programmability have been provided, enabling the development of creative DPI solutions. Testbed settings are required to validate these developments. P4T, a Mininet and ONOS controller-based testbed, enables the creation of realistic network scenarios for full DPI research. Our study addresses the changing network environment, which is characterized by an increase in data-intensive applications such as video streaming, cloud services, and the Internet of Things. Protecting network infrastructure from security threats remains a high issue, which has prompted our research into dynamic network traffic detection, classification, and control, with a focus on video traffic. Our study evaluates the possibility of DPI utilizing P4T, which is implemented with Mininet and the ONOS controller. We create and test enhanced DPI solutions that combine the programmability of P4 with the flexibility of the testbed to meet security and network management concerns. Advanced video packet classification techniques, specialized video traffic analysis, and adapted P4-based solutions for current network environments that streamline video content delivery and management are among our research achievements.
