Recently researchers David Todoli Ferrandis, Javier Silvestre Blanes, Victor Sempere Paya and Salvador Santonja Climent from partners ITI and the Polytechnical University of Valencia (UPV) have published the scientific paper entitled “Adaptive Beacon Period Configurator for scalable LoRaWAN downlink applications”. Their findings groundwork the development of the real-time communications for sensor and platform integration and control tasks within the AI-PRISM human-centred collaborative robotic platform.
While implementing an industrial-end-user-driven project to provide a human-centred AI-based solutions ecosystem, the AI-PRISM researchers are investigating the best technical means to design and develop a robust wired and wireless deterministic communication network. This infrastructure is targeted to manufacturing scenarios with tasks difficult to automate and where speed and versatility are essential.
Today, we are interviewing Dr David Todoli from ITI, involved in this process of research in the AI-PRISM project Human-Centred collaborative robotic platform, to discuss their findings and impact on AI-PRISM innovations and developments. David is a senior engineer in the Industrial Ciberconectivity group at ITI, and his work involves Wireless sensor network systems and network modelling and Industrial Internet of Things (IoT) for Industry 4.0 applications.
Hello, Dr David Todoli is a pleasure to have you with us!
Could you give us further details of the challenges addressed related to networks’ requirements of IoTs?
Well, we all know that using wireless technologies comes with some challenges. We have chosen LoRaWAN because it helps overcome some of these problems, it has long range coverage, it can recover signal below noise levels, it is secure, and is energy efficient. But there are still aspects where we can enhance these technologies. Specifically, we are working on the downlink applications of LoRaWAN, which currently has 2 or 3 main drawbacks:
Scalability: LPWAN networks need to be able to scale to support a large number of devices. This can be challenging, as it requires a large number of gateways and base stations. Note that we have the limitation of channel access, that has to comply with regulations of duty cycle. For downlink this is critical, because there are fewer gateways than devices, which means that the gateway needs more time to be able to communicate with all the devices.
This takes us to the second drawback, the application-specific requirements: LPWAN networks need to meet the specific requirements of the IoT application. This can vary depending on the application, such as the data rate, latency, and reliability requirements.
That’s why our ongoing research efforts to address these challenges and improve the performance of LPWAN networks, based on dynamic channel access scheduling to adapt to different network sizes and different application requirements.
Can you walk us through the methodology conducted in your research?
First, we have simulated a scenario that can be representative of a real-life application. We selected area dimensions, channel noise models for industrial buildings, and all related parameters. We identified the main problems with the gateway buffer and proposed that a channel access scheduling that can change according to network size or traffic load can enhance the network performance, so we developed an algorithm that based on some assumptions about network size and QoS parameters, modifies the temporal parameters of the scheduled window, such as slot durations or the period of scheduling. Then we implemented the algorithm in the simulator and performed a batch of experiments to validate our proposal.
What were the results, and how will they be used for AI-PRISM developments?
We found out that this mechanism can benefit LoRaWAN networks of small to medium size for most applications, enhancing the PDR and energy efficiency or even reduced delay. But as expected, large and massive networks introduce mainly the drawback of delay. For those networks the PDR can still be guaranteed but at the cost of out of the question delay. The mechanism is configurable, so in the end the decision can rest on the user. For instance, a delay of one hour may be reasonable for configuration updates or non-critical telemetry requests, but other type of messages such as actuations would require a limit within seconds. Based on these results, network administrators can also predict how many gateways could be needed to maintain the QoS.
Finally, which other stakeholders can benefit from these findings, and what is their impact on their industries or sectors?
This communications solution can be used in many different sectors, from smart agriculture, smart cities, smart metering, remote installations control, water, oil and gas infrastructure and so on. It will enable a new batch of applications that were lacking a long-range technology with optimized bidirectional capabilities.
Thank you very much Dr Todoli.
Thanks for having me.
The full paper is available here !
Do you want to learn more about AI-PRISM and the team behind the project? Subscribe to our newsletter and follow us on LinkedIn and Twitter, so you don’t miss a thing!