With worldwide deployment of LoRa/LoRaWAN LPWAN networks in a large variety of applications, it is crucial to improve the robustness of LoRa channel access which is largely ALOHA-like to support environments with higher node density. This article presents extensive experiments on LoRa Channel Activity Detection and Capture Effect property in order to better understand how a competition-based channel access mechanisms can be optimized for LoRa LPWAN radio technology. In the light of these experimentation results, the contribution continues by identifying design guidelines for a channel access mechanism in LoRa and by proposing a channel access method with a lightweight collision avoidance mechanism that can operate without a reliable Clear Channel Assessment procedure. The proposed channel access mechanism has been implemented and preliminary tests show promising capabilities in increasing the Packet Delivery Rate in dense configurations.

Recently, Low-Power Wide Area Networks (LPWAN) play a key role in the Internet-of-Things (IoT) maturation process. Under the LPWAN broad term are a variety of technologies enabling power efficient wireless communication over very long distances. For instance, technologies based on ultra-narrow band modulation (UNB)–for example SigFox–or Chirp Spread Spectrum modulation (CSS)–for example LoRa–have become de facto standards in the IoT ecosystem. Given the incredible worldwide uptake of LPWAN networks for a large variety of innovative IoT applications, including multimedia sensors, it is important to understand the challenges behind large scale and dense LPWAN deployment, especially because both Sigfox and LoRa networks are currently deployed in unlicensed bands. This situation is most likely not going to change, at least in the next few years, as working in the unlicensed band allows for a much quicker uptake of the technology. This chapter has a particular focus on LoRa technology as it can be deployed in a private and ad-hoc manner, making experimental deployments much easier. Existing studies on LoRa scalability and radio channel access mechanisms for LoRa LPWAN will be reviewed and promising approaches will be presented in more details. The chapter will also provide to the readers useful information on the LoRa physical layer, as well as on promising interference mitigation techniques that can be applied such as capture effect and successive interference cancellation. The chapter will also give a large part on experimental results based on real-world deployments of both IoT test-beds and IoT production networks in the context of 3 R&D projects (2 EU H2020 projects–WAZIUP & WAZIHUB–and 1 national ANR project–PERSEPTEUR).

An abstract is not available.

Image sensor node's activity is defined based on the application's criticality level and sentry nodes with faster capture rates have higher probability to detect intrusions and will alert neighbor nodes. At the MAC level, we consider duty-cycled approaches to periodically set nodes in sleep mode for energy preservation. However, in doing so, care must be taken to also preserve the quality of event detection and sentry nodes must still be able to quickly propagate alert messages to meet the strong requirements of mission-critical surveillance applications in communication delays. We propose an original approach to dynamically determine the duty-cycle length of image sensor nodes to increase the probability of matching active period between nodes, thus reducing the alert latency while globally reducing the energy consumption.

Mission-critical surveillance applications such as intrusion detection or disaster response have strong requirements in communication delays. We consider a Wireless Image Sensor Network (WISN) with a scheduling of image sensor node’s activity based on the application criticality level. Sentry nodes capable of detecting intrusions with a higher probability than others will alert neighbor nodes as well as activating cover sets member for image disambiguation or situation-awareness purposes. At the access level, we consider duty-cycled Medium Access Control to periodically set nodes in sleep mode for energy preservation. However, in doing so, care must be taken to also preserve the quality of event detection and sentry nodes must still be able to propagate quickly alert messages. We propose an original approach to dynamically determine the duty-cycle values of image sensor nodes to increase the probability of matching active period between nodes, thus reducing the alert latency while globally reduce the energy consumption.