By offloading computation-intensive and latency-sensitive tasks from mobile devices to the mobile edge server, mobile edge computing (MEC) has been considered as a promising technology in 5 G and beyond to reduce task delay and energy consumption. This paper considers an optimization framework of computation offloading in a MEC system with multiple devices. We aim to minimize the energy consumption of all devices with their task delay constraints by jointly optimize the communication and computation resource allocation at both the devices and the mobile edge server. Specifically, dynamic voltage scaling (DVS) technology is considered in each device to adjust the operating frequency according to its delay constraint. The formulated problem is a non-convex problem due to a couple among multiple variables. To tackle the non-convex problem, we first decompose the original problem by optimizing the offloading ratio and transmission power iteratively. Then, we proposed a joint communication and computation optimization algorithm based on the difference of convex function algorithms (DCA) to solve the optimization problem. Finally, simulation results show that the proposed joint communication and computation scheme significantly improves the energy efficiency of the devices comparing with the local computing scheme and server computing scheme.

Characteristic gas detection can be an efficient way to predict the degree of thermal runaway of a lithium battery. In this work, a sensor array consisting of three commercial MOS sensors was employed to discriminate between three target gases, CO, H₂ and a mixture of the two, which are characteristic gases released during the thermal runaway of lithium batteries. In this work, an integrated model that makes the classification stage results one of the feature inputs for the concentration regression stage was employed, successfully reducing the RMSE of the concentration regression results. In addition, we also explored the influence of the selection of the response time length on the classification and regression tasks, achieving the best results in a short time through the optimum algorithm. To assess the impact of time duration sensor data on the results, we selected four time windows of different length and extracted the corresponding sensor response data for subsequent processing. Initially, principal component analysis (PCA) was used to visualise the clustering of the three target gas samples at room temperature, providing a preliminary data analysis. For the classification phase, we chose three classification algorithms—MLP (Multilayer Perceptron), ELM (Extreme Learning Machine), and SVM (Support Vector Machine)—and performed a comprehensive comparison of their classification and generalisation abilities using grid search for hyperparameter optimisation and five-fold cross-validation. The results demonstrated that MLP achieved 99.23% classification accuracy during the 20 s response period. In the concentration regression phase, we combined the classification results with the raw features to create a new feature set, which was then input into a multi-output MLP regression model. The root mean square error (RMSE) employing the new feature set was used to measure the prediction error. Ultimately, the findings showed that the input of combined features significantly reduced the regression error for the mixed gas.

The emergence of low-cost depth sensors opens up new potentials for skeleton-based human action recognition. The recent methods for this task have made significant progress by incorporating graph convolution. However, they (1) have limitations in modeling the complex and variable temporal dynamics, and (2) cannot maximize the complementarity of the spatial and temporal features. Besides, (3) the loss function of these methods has an inherent weakness in optimizing the intraclass compactness. To this end, we propose a pyramidal graph convolutional network (PY-GCN) in this paper. Specifically, (1) an effective yet efficient single-oriented pyramidal convolution is proposed. It involves multiple kernels with varying sizes and depths that are capable of capturing different levels of the temporal dynamics at multiple scales. (2) A pseudo-two-stream structure for the basic block of the network is proposed to comprehensively aggregate discriminative cross-spatiotemporal features. Moreover, (3) a pairwise Gaussian loss together with the cross-entropy loss is introduced to the model, which can focus on both intraclass compactness and interclass separability. Our PY-GCN achieves state-of-the-art performance on three challenging large-scale datasets.

A deployment strategy for antennas is proposed to enhance the field coverage of wireless devices in underground coal mines, which uses the multimodal approach. Based on the attenuation coefficient of each mode, the propagation loss of the vertically polarized modes with n = 1, 2, and 3 and that of the horizontally polarized modes with m = 1, 2, and 3 are less. For the signal propagation in the near-field region of the source, these modes play significant role. According to the coupling of the antenna to the modes, the significant modes can be effectively excited by mounting horizontally polarized antenna on the ceiling or the vertically polarized antenna on the sidewall. For the signal propagation in the far-field region, the (1, 1) mode is significant, and the horizontally polarized excitation is preferred due to smaller attenuation. By comparing the total coupling efficiency of the antenna-radiated power, stronger excitation power can be obtained by mounting the antenna near the center of the ceiling or that of the sidewall. The field strength in the near-field and far-field region coverage can be maximized by optimizing the antenna polarization and position. For verification, a grid-based test system was designed and a series of measurements inside an underground coal mine tunnel were conducted. The measured results have a good agreement with the theoretical ones. In addition, the EM interference caused by the infrastructure in real mines on the signal received by the antennas at different positions was also measured and compared. Results show that the influences of cables with long-term stable loads, such as power cables and communication cables, can be neglected.

Wireless underground sensor network (WUSN) has potential application value in environmental monitoring, disaster detection and personnel rescue in the coal mine collapsed zone. Due to the absorption and reflection of rocks and minerals, the attenuation law of the electromagnetic signal in the coal mine collapsed area is complicated and the path loss is large. Understanding the signal attenuation characteristics and estimating path loss are the basis for WUSN design, deployment, and optimization. In this paper, we conduct extensive survey on the geological structures and materials of coal mine collapsed areas. We propose a path loss model for wireless communications in coal mine collapsed areas. The signal attenuation under different geological and communication conditions is discussed. Based on the model results, the basic design suggestions of WUSN in coal mine collapsed areas are put forward. When the material in the collapsed area is shale or mudstone, the WUSN design should focus on solving the problem of large signal path loss. When the collapsed area is sandstone or limestone, the WUSN design should focus on solving the multipath fading problem. The signal with operating frequency of 300-600 MHz has small path loss in the mine collapsed area, but the signal power fluctuates greatly. Signal fluctuation can be reduced by shortening the communication distance, placing the wireless node near the roof or the floor of the collapsed zone, and using vertically polarized antennas.

In traditional underwater wireless sensor networks (UWSNs), it is difficult to establish reliable communication links as the acoustic wave experiences severe multipath effect, channel fading, and ambient noise. Recently, with the assistance of magnetic induction (MI) technique, cooperative multi-input-multi-output (MIMO) is utilized in UWSNs to enable the reliable long range underwater communication. Compared with the acoustic-based UWSNs, the UWSNs adopting MI-assisted acoustic cooperative MIMO are referred to as heterogeneous UWSNs, which are able to significantly improve the effective cover space and network throughput. Due to the complex channel characteristics and the heterogeneous architecture, the connectivity of underwater MI-assisted acoustic cooperative MIMO networks is much more complicated than that of acoustic-based UWSNs. In this paper, a mathematical model is proposed to analyze the connectivity of the networks, which considers the effects of channel characteristics, system parameters, and synchronization errors. The lower and upper bounds of the connectivity probability are also derived, which provide guidelines for the design and deployment of underwater MI-assisted acoustic cooperative MIMO networks. Monte Carlo simulations were performed, and the results validate the accuracy of the proposed model.