Meta-Material Sensor-Based Internet of Things for Environmental Monitoring by Deep Learning: Design, Deployment, and Implementation
- 11 October 2022
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 22 (4) , 2462-2476
Abstract
Using widely deployed Internet of Things (IoT) sensors to perceive the environmental distribution is crucial in many IoT applications, such as intelligent healthcare and smart home. As using traditional sensors will lead to high costs and maintenance, it is important to design the next-generation IoT sensor to reduce the cost of ubiquitous deployment. For this purpose, we propose a novel IoT system based on low-cost and fully passive meta-material sensors. Specifically, the meta-material sensors can sense multiple environmental conditions such as temperature and humidity levels, and transmit back the information by signal reflection, simultaneously. With the information contained in the received signals, a wireless receiver can obtain detailed environmental distributions. However, it is not trivial to achieve high sensing accuracy in the meta-material sensor based IoT system because the structure of the meta-materials, the deployment positions of sensors, and the reconstruction function for environmental distributions need to be jointly optimized. To handle this challenge, we propose an algorithm to design the meta-material based IoT system with the help of a deep learning approach. Simulation results verify that the proposed algorithm effectively maximizes the sensing accuracy. Experimental evaluations also show that the proposed scheme can obtain humidity distribution with an accuracy of over 93%.Keywords
Funding Information
- National Key R&D Project of China (2020YFB1804800)
- Clinical Medicine Plus X - Young Scholars Project, Peking University, the Fundamental Research Funds for the Central Universities and the National Natural Science Foundation (62271012, 61941101)
- Beijing Natural Science Foundation (L212027, 4222005)
This publication has 32 references indexed in Scilit:
- A Fully Inkjet-Printed Wireless and Chipless Sensor for CO2and Temperature DetectionIEEE Sensors Journal, 2014
- Metamaterial-based microfluidic sensor for dielectric characterizationSensors and Actuators A: Physical, 2013
- Simple and Accurate Analytical Model of Planar Grids and High-Impedance Surfaces Comprising Metal Strips or PatchesIEEE Transactions on Antennas and Propagation, 2008
- Non-line-of-sight identification in ultra-wideband systems based on received signal statisticsIET Microwaves, Antennas & Propagation, 2007
- A Tutorial on the Cross-Entropy MethodAnnals of Operations Research, 2005
- Pattern Search Algorithms for Bound Constrained MinimizationSIAM Journal on Optimization, 1999
- A comparison of theoretical and empirical reflection coefficients for typical exterior wall surfaces in a mobile radio environmentIEEE Transactions on Antennas and Propagation, 1996
- Universal approximation to nonlinear operators by neural networks with arbitrary activation functions and its application to dynamical systemsIEEE Transactions on Neural Networks, 1995
- Generalization of back-propagation to recurrent neural networksPhysical Review Letters, 1987
- Simulated annealing1987