Predictive Forwarding Rule Caching for Latency Reduction in Dynamic SDN

Area of Science:

• Computer Science
• Networking
• Robotics

Background:

• Unmanned vehicles in industrial/military settings require reliable wireless communication.
• Wireless links cause data loss and delays, impacting control and sensor data.
• Existing distributed routing methods increase network traffic and interference.

Purpose of the Study:

• To address the challenge of rapid route creation in Software-Defined Networking (SDN) for mobile unmanned systems.
• To reduce end-to-end latency and data loss in wireless networks for unmanned vehicles.

Main Methods:

• Introduced the CRIMSON (Cashing Routing Information in Mobile SDN Network) algorithm.
• Implemented a caching technique for forwarding rules based on predicted link states in SDN.
• Detected network link state changes due to node mobility and cached new forwarding rules based on predicted topology changes.

Main Results:

• The CRIMSON algorithm significantly reduces end-to-end latency.
• Achieved an average latency reduction of 88.96% compared to conventional reactive modes.
• Achieved an average latency reduction of 59.49% compared to conventional proactive modes.

Conclusions:

• CRIMSON enhances SDN performance in dynamic wireless environments for unmanned vehicles.
• The algorithm effectively minimizes latency and data loss, crucial for mission-critical operations.
• Predictive caching of routing rules is a viable strategy for improving network efficiency.