Abstract
LPWANs are networks characterized by the scarcity of their radio resources and their limited payload size. To extend the efficiency of the data transmission by decreasing the traffic sent from sensors, this paper proposes a lossy compression method using known ML techniques. We embedded a pre-trained neural network directly on constrained LoRaWAN devices and we tested the trade-off between compression ratio and accuracy of the compression algorithm. This paper studies multiple aspects of the system-energy consumption, error rate due to the lossy compression, compression ratio and the impact of LSTM parameter quantization-to measure the possible strengths and weaknesses of using a dual prediction system in order to reduce transmission costs. Surprisingly, machine learning used in this context does not consume a lot of energy and it even leads to energy saving in the very constrained devices which are the sensors.
To improve data transmission efficiency by reducing sensor traffic, the article “Embedding ML Algorithms onto LPWAN Sensors for Compressed Communications” introduces a lossy compression method using known ML techniques.
We embedded a pre-trained neural network in constrained LoRaWAN devices and tested the balance between compression rate and accuracy.
The study covers system aspects like energy consumption, error rates due to lossy compression, compression ratio, and LSTM parameter quantization impact, aiming to evaluate the strengths and weaknesses of using a predictive system for reducing transmission costs.
Surprisingly, ML in this context is energy-efficient, even saving energy in highly constrained devices like sensors.
Antoine BERNARD
Postdoctoral Fellow @ Polytechnique Montréal
Tech enthousiast, interested in computer networks, distributed processing and a bit of AI.