How Can Wake-up Radio Reduce LoRa Downlink Latency for Energy Harvesting Sensor Nodes?

Nour El Hoda Djidi, Matthieu Gautier, Antoine Courtay, Olivier Berder, Michele MagnoSensors 2021 · DOI

TL;DR

This paper tackles the high downlink latency in LoRaWAN networks, a common issue for IoT applications. They propose a novel MAC protocol, LoRa-WuR, that combines LoRa's long-range capabilities with the low-latency of Wake-up Radio (WuR) and integrates energy harvesting to ensure node sustainability. The results show a significant reduction in downlink latency, by up to 90%, making LoRaWAN more responsive for critical command delivery.

What Problem Does This Paper Address?

The Internet of Things (IoT) relies heavily on Low-Power Wide-Area Networks (LPWANs) for long-range, low-power communication. LoRa, and its standard MAC protocol LoRaWAN, are popular choices due to their extensive range and ability to communicate in noisy environments. However, LoRaWAN faces a significant challenge: high downlink latency. This is particularly problematic for applications requiring timely command delivery from a gateway to sensor nodes, such as actuator control or urgent data rate adjustments. This latency is a direct consequence of LoRaWAN's energy-saving mechanisms, which often involve nodes sleeping for extended periods and only waking up at scheduled intervals or after transmitting their own data.

Furthermore, for widespread, sustainable IoT deployments, especially in remote or hard-to-access locations, battery replacement is impractical. Energy harvesting (EH) offers a solution by enabling nodes to scavenge energy from their environment (e.g., solar power). However, integrating EH introduces variability in node power availability, which can impact Quality of Service (QoS) and further complicate latency management. This paper aims to bridge this gap by proposing a solution that reduces downlink latency in LoRaWAN while simultaneously addressing the challenges of energy harvesting and ensuring node self-sustainability.

Methods

The core of this paper is the LoRa-WuR MAC protocol, which cleverly combines LoRa and Wake-up Radio (WuR) technologies. Here's how it works:

  • Heterogeneous Architecture: Sensor nodes are equipped with two radio modules: one for LoRa (long-range, used for communication with the gateway) and another for WuR (short-range, always-on, ultra-low power).
  • Opportunistic Cluster Heads (OCH): Nodes use LoRa in Class A mode for uplink communication with the gateway. When a node transmits data, it opens two short receive windows. During these windows, any node can act as an Opportunistic Cluster Head (OCH).
  • WuR for Low-Latency Commands: If the gateway has a command for a specific "targeted node," it can send this command during one of the OCH's receive windows. The OCH then forwards this command to the targeted node using its short-range WuR module as a Wake-Up Beacon (WUB). The WUB contains the targeted node's address, ensuring only the intended node is woken up.
  • Energy Harvesting Integration: Nodes are equipped with energy harvesting systems (e.g., solar panels). A novel energy manager is introduced to adapt the node's uplink Quality of Service (QoS) – specifically, its packet generation rate – based on the harvested energy. This manager aims to minimize fluctuations in QoS between day and night.
  • Hybrid Modeling: The authors developed latency and power consumption models that leverage microbenchmarks on real hardware platforms to accurately reflect the influence of energy harvesting conditions.

Datasets

Not discussed in the paper. The authors used experimental measurements from their own platform for evaluation.

Key Insights and Ideas

  • WuR as a Latency Killer for Downlink: The most significant insight is that WuR, typically known for its low-power listening, can be repurposed to drastically reduce downlink latency in LoRaWAN. By having an OCH relay gateway commands via WuR, the need for the targeted node to wait for its own scheduled receive window is eliminated.
  • Cooperative Networking for Latency Reduction: The LoRa-WuR protocol fosters cooperation. Even nodes with lower uplink QoS (due to less harvested energy) can contribute by acting as OCHs, forwarding commands for others. This collective effort is key to the latency reduction.
  • Energy Harvesting for Sustainability, Not Just Performance: While energy harvesting is crucial for node longevity, the paper highlights its role in tuning performance. The proposed energy manager dynamically adjusts uplink QoS based on harvested energy, ensuring a more consistent experience despite environmental variations.
  • Trade-off Management is Crucial: The paper explicitly models and evaluates the trade-off between downlink latency and power consumption, showing how LoRa-WuR can achieve lower latency at comparable or even lower power consumption than traditional LoRa, especially for shorter latencies.

Key Results

  • Up to 90% Latency Reduction: For a cluster of 10 nodes, the LoRa-WuR protocol achieved a downlink latency reduction of almost 90% compared to the traditional LoRa scheme.
  • Improved Downlink QoS: The LoRa-WuR scheme demonstrated a downlink QoS that was up to 10 times higher than traditional LoRa, as all nodes could contribute to forwarding commands.
  • Reduced Power Consumption for Lower Latencies: For achieving similar downlink latencies, LoRa-WuR showed reduced power consumption compared to LoRa, particularly when targeting lower latency values.
  • Consistent QoS with Energy Manager: The proposed energy manager (RHE) showed minimal fluctuation in downlink QoS between day and night, outperforming other tested energy management strategies in this regard.

Limitations and Future Work

  • Short-Range WuR Bottleneck: The effectiveness of the LoRa-WuR protocol is inherently limited by the short range of the WuR. While the paper mentions this, the maximum number of nodes that can effectively form a cluster is naturally constrained by WuR's communication distance.
  • Energy Harvesting Variability: While the paper models energy harvesting and its impact, real-world deployments might face more complex and unpredictable energy availability patterns.
  • Future Work: The authors suggest further investigation into the uplink QoS tuning based on harvested energy to fully leverage the proposed MAC protocol. They also imply that exploring different energy harvesting sources and more advanced energy management strategies could be beneficial. The paper also implicitly points towards optimizing the cluster size and formation based on WuR range and node density.

Summary updated

BibTeX
@article{djidi2021how,
  author = {Nour El Hoda Djidi and Matthieu Gautier and Antoine Courtay and Olivier Berder and Michele Magno},
  title = {{How Can Wake-up Radio Reduce LoRa Downlink Latency for Energy Harvesting Sensor Nodes?}},
  journal = {Sensors},
  year = {2021},
  doi = {10.3390/s21030733},
}
Figure