A frequency-switch strategy is introduced into the magnetic induction-based wireless underground sensor network (MI-WUSN) for its high connectivity and network throughput, which then makes routing design more complex and challenging. To this end, we study the frequency-switchable routing design to start a discussion about the high-reliability routing design of MI-WUSN. First, we analyze the frequency-selective property and map the dynamic MI-WUSN into a multilayer network. Then, we take the network throughput and energy consumption into account and formulate the frequency switchable routing decision problem in dynamic MI-WUSN as a constrained optimization problem. Finally, we evaluate how various design parameters of our obtained protocol are affecting the network performance and explore the performance limit.

This research proposes a high performance, low-profile planar design, and demonstration of circularly polarized harmonic antenna arrays for 4 and 8 GHz harmonic radar system, which is equipped, and the antenna's circular polarization (CP) performance is tested using a passive nonlinear harmonic tag (snowflake) in a noisy environment. This work mainly concentrates on investigation of the wave polarization (circular), gain estimation and impedance matching of the antenna arrays that will be potentially assembled to the second-order harmonic radar system operating at 4 and 8 GHz, which plays a critical role in 2-D tag localization and post processing methods. The passive snowflake tag design is implemented in the simulation to determine the current distribution and gain plot patterns in comparison to a classic dipole antenna tag structure. In experiments: a combination of fundamental circular polarized (left-hand) antennas were demonstrated to find axial ratio ≤ 3 dB band. A novel tag 2-D localization measurement, combination of CP harmonic antenna arrays is utilized to perform and obtain the circular polarized received signal power in dBm versus angle rotation by every 10° on a rotating platform setup. Finally, the wireless tag tracking result is achieved by using the designed circularly polarized antennas and the passive nonlinear tag orientation setup.

This article presents a 24-GHz microfluidic sensor using frequency-locked loop (FLL) technology for detecting liquid concentrations. The sensor, based on FLL, features a microfluidic channel placed over an asymmetrical coplanar waveguide resonator (ACPWR) that functions as a sensing device. For testing purposes, we use ethanol–water mixtures and glucose–water solutions as the liquid under test. Due to the electric field distribution in media with varying dielectric constants, the phase of the signal undergoes different phase deviations. The FLL-based sensor is capable of detecting these phase deviations and, in response, produces a frequency-modulated signal. This signal is subsequently demodulated into a corresponding voltage with the aid of a frequency demodulator, realized through a phase detector. Consequently, the sensor demonstrates the capability to differentiate between tested liquids of varying concentrations and offers a linear response that correlates the output voltage with the liquid concentration. The proposed 24-GHz FLL microfluidic sensor offers advantages, such as cost effective, high sensitivity, and compact size. It has a great possibility to implement this sensor using the system-on-chip technology. As it combined with Internet of Things technologies, it may have a capability of real-time biomedical specimen sensing for daily life.

This article proposes the use of a wireless sensor developed with a microstrip patch antenna in conjunction with an open-circuited stub for remote, real-time monitoring of crack width expansion. Technically, the open-circuited stub is exploited as a sensing structure with excellent sensitivity, and the crack growth is able to be mechanically mapped to the stub length extension via an incorporation of a mirrored stub structure placed right on top of the open-circuited stub. Thanks to a very strong relationship between the stub input admittance and its electrical length, the operating frequency of the patch is able to be significantly downshifted as the stub is mechanically lengthened (i.e., the crack grows). With the proposed wireless sensing scheme, the crack width expansion can be determined by identifying the dominant frequency component of the sensing signal received at remote data stations in a real-time manner. The sensor operating at 4 GHz was developed for experimental verification and as a demonstration of effectiveness. The experimental results show that the wireless sensor is able to identify the crack expansion of up to 2 mm with a resonant frequency downshift of 110 MHz. Furthermore, the maximum expansion and the finest increment can be simply specified with the sensor operating frequency regime. Moreover, a stable wireless link can be sustained as the patch radiation patterns remain unaffected while crack grows, and the sensor can be reused since the proposed monitoring does not result in any structure destruction or deformation.

The integration of serverless edge computing and the Industrial Internet of Things (IIoT) has the potential to optimize industrial production. However, cold start latency is one of the main challenges in this area, resulting in resource waste. To address this issue, we propose a new machine learning-based resource management framework called MASTER which utilizes an extreme gradient boosting (XGBoost) model to predict the cold start latency for Industry 4.0 applications for performance optimization. Furthermore, we created a new cold start dataset using an IIoT scenario (i.e. predictive maintenance) to validate the proposed MASTER framework in serverless edge computing environments. We have evaluated the performance of the MASTER framework using a real-world serverless platform, Google Cloud Platform for single-step prediction (SSP) and multiple-step prediction (MSP) operations and compared it with existing frameworks that used deep deterministic policy gradient (DDPG) and long short-term memory (LSTM) models. The experimental results show that the XGBoost-based resource management framework is the most successful model in predicting cold start with mean absolute percentage error (MAPE) values of 0.23 in SSP and 0.12 in MSP. It has been also identified that the Linear Regression model (utilized in the MASTER framework) has the least computational time (0.03 seconds) as compared to other deep learning and machine learning models considered in this work. Finally, we compare the energy consumption and CO2 emissions of all models to emphasize resource awareness.

In this article, machine learning methods are used to assess how well wireless sensor networks transmit and receive radar signals. Measurements are done with labeled and unlabeled datasets where output functions are modified in relation to transmitted input in order to test the transceiver of radar signals. The main contribution in the proposed method is to focus on the possibility of choosing a free space model that transmits the radar signals in wireless sensor networks without any interruptions. Hence, for such type of transmissions, reference time period is selected in order to perform radar signal classification, and at the same time, separation of unnecessary interruptions is reduced using clustering procedures. Since the radar signals can be monitored with automatic transmission techniques, the outcomes are combined with supervised, unsupervised, and reinforcement learning models to increase the effect of transmissions. Therefore, the objective functions are designed with three scenarios where reinforcement learning proves to provide adequate connections for radar signals to all wireless sensor networks at reduced error of 0.3%. In addition, with reinforcement learning, the distance of radar signal transmission is maximized to a level greater than 75% at minimized noise ratio of 0.8%.

This article proposes a novel iterative vector-based localization method in a large heterogeneous sensor network, where a subset of nodes possesses the capability to measure both distance and angle information, while the others are only limited to distance measurements. Unlike conventional vector-based positioning methods that assume all nodes can measure both distance and angle, our approach tackles a more realistic scenario where some nodes are limited to distance-only measurements. To address the challenges of the node localization in a heterogeneous sensor network, the proposed positioning method calculates vector information between the nodes that are not directly communicated and aligns it with a reference coordinate. In addition, the proposed method employs an iterative calculation, such as least-squares minimization, thereby achieving high positioning accuracy. Simulation results demonstrate that the proposed positioning method outperforms the conventional distance-based positioning method in environments with low angle measurement errors, exhibiting up to 44% higher positioning accuracy. Furthermore, the proposed positioning method shows 24% higher positioning accuracy compared with the conventional vector-based positioning method.