Whisper: Programmable and Flexible Control on Industrial IoT Networks

Area of Science:

• Computer Science
• Network Engineering

Background:

• Software Defined Networking (SDN) offers centralized network control but faces challenges in low power and lossy networks (LLNs) common in the Internet of Things (IoT).
• Key challenges include reliable in-band connectivity and accommodating resource-constrained IoT devices.

Purpose of the Study:

• To introduce Whisper, a novel system enabling SDN capabilities in LLNs.
• To demonstrate Whisper's ability to remotely control node forwarding and cell allocation in IoT networks.

Main Methods:

• Whisper utilizes a centralized controller that sends routing and scheduling messages compatible with existing LLN protocols.
• Algorithms minimize message overhead, reducing the burden on resource-constrained IoT devices.
• Implementation is demonstrated within current IoT standards (RPL and 6TiSCH) without requiring firmware modifications.

Main Results:

• Whisper successfully controls both the scheduling and routing planes of LLNs.
• Experimental results show significantly lower overhead, no additional latency, and no packet loss compared to other SDN-IoT solutions.
• Whisper supports use cases like rerouting around low-battery devices and runtime defense against jamming attacks.

Conclusions:

• Whisper provides an effective SDN-like control for distributed LLNs, compatible with legacy and next-generation IoT devices.
• The system enhances network programmability and flexibility without specific SDN hardware support.
• Whisper offers a lightweight, efficient solution for managing complex IoT networks.