Abstract: This chapter describes the evolution of, and state of the art in, energy‐efficient techniques for wirelessly communicating networks of embedded computers, such as those found in wireless sensor network (WSN), Internet of Things (IoT) and cyberphysical systems (CPS) applications. Specifically, emphasis is placed on energy efficiency as critical to ensuring the feasibility of long lifetime, low‐maintenance and increasingly autonomous monitoring and control scenarios. A comprehensive summary of link layer and routing protocols for a variety of traffic patterns is discussed, in addition to their combination and evaluation as full protocol stacks.
Introduction
The "fourth industrial revolution" envisions a world where billions of sensors and actuators instrument industrial processes. However, a significant hurdle remains: power. Many of these devices are deployed in environments without mains electricity, making energy the primary constraint on their lifetime and maintenance costs. Since the radio transceiver is typically the most power-hungry component of an embedded system, communication is the primary drain on a device's limited energy budget. This paper explores the design of communication protocols—from the physical link to network routing—that minimize power consumption while maintaining the reliability and latency required by industrial applications.
Link Layer Strategies: Radio Duty Cycling
The core of energy efficiency at the link layer is Radio Duty Cycling (RDC), where the radio is powered down for as long as possible. Protocols are generally categorized into three types based on their wake-up mechanisms:
- Scheduled (TDMA): Nodes wake up at precise, synchronized times (e.g., TSMP).
- Preamble Sampling (LPL): Transmitters "strobe" the channel to wake up receivers (e.g., B-MAC, ContikiMAC).
- Hybrid: These combine contention-based access with scheduled slots to balance flexibility and efficiency.
Network Layer and Routing
Routing protocols must efficiently move data through the network based on specific traffic patterns: Collection (many-to-one), Peer-to-Peer (any-to-any), and Dissemination (one-to-many). A major turning point in routing was the shift from simple "hop count" metrics to the Expected Transmission Count (ETX). By accounting for link quality and the energy cost of retransmissions, ETX-based protocols significantly boost network efficiency.
Standardized Protocol Stacks
The field is converging on standardized stacks that bridge the gap between battery-constrained devices and the Internet. 6LoWPAN uses clever header compression to fit IPv6 packets into the small frames (127 bytes) allowed by low-power wireless standards like IEEE 802.15.4. Furthermore, the 6TiSCH architecture combines the reliability of Time-Synchronized Channel Hopping (TSCH) with the interoperability of IPv6, creating a robust framework for industrial IoT.
Performance and Key Results
Research indicates that standardization does not necessarily come at a high energy cost. For instance, the standard IPv6 routing protocol (RPL) consumes only about 3% more energy than the highly optimized, non-IP Collection Tree Protocol (CTP). Additionally, 6LoWPAN can compress a standard 40-byte IPv6 header down to as little as 4 bytes. State-of-the-art implementations of TSCH have demonstrated sub-1% duty cycles under reasonable traffic loads, moving closer to the goal of "energy neutrality," where devices harvest as much energy as they consume.
Conclusion and Future Directions
Despite significant progress, challenges remain. There is a lack of comprehensive, standardized evaluations for entire protocol stacks, and a need for reproducible experiments on real-world facilities. Future work must move toward holistic co-design, where energy harvesting, ultra-low-power wake-up circuits (WuR), and predictive sensing are designed in tandem with communication protocols to enable truly autonomous IoT systems.
Summary updated