With the development of the Internet of things (IoT) and wearable devices, the sensor-based human activity recognition (HAR) has attracted more and more attentions from researchers due to its outstanding characteristics of convenience and privacy. Meanwhile, deep learning algorithms can extract high-dimensional features automatically, which makes it possible to achieve the end-to-end learning. Especially the convolutional neural network (CNN) has been widely used in the field of computer vision, while the influence of environmental background, camera shielding, and other factors are the biggest challenges to it. However, the sensor-based HAR can circumvent these problems well. Two improved HAR methods based on Gramian angular field (GAF) and deep CNN are proposed in this paper. Firstly, the GAF algorithm is used to transform the one-dimensional sensor data into the two-dimensional images. Then, through the multi-dilated kernel residual (Mdk-Res) module, a new improved deep CNN network Mdk-ResNet is proposed, which extracts the features among sampling points with different intervals. Furthermore, the Fusion-Mdk-ResNet is adopted to process and fuse data collected by different sensors automatically. The comparative experiments are conducted on three public activity datasets, which are WISDM, UCI HAR and OPPORTUNITY. The optimal results are obtained by using the indexes such as accuracy, precision, recall and F-measure, which verifies the effectiveness of the proposed methods.

Human activity recognition (HAR) utilizing wearable sensors is a prominent topic in academia, and has been widely used in health monitoring, medical treatment, motion analysis, and other fields. Although HAR technology based on deep learning has made some progress, there are still some problems such as incomplete feature extraction and low utilization rate of features, which may result in false recognition. To settle the problems, we propose a novel deep multi-feature extraction framework based on attention mechanism (DMEFAM), which includes the temporal attention feature extraction layer (TAFEL), the channel and spatial attention feature extraction layer (CSAFEL), and the output layer. The TAFEL is comprised of bidirectional gated recurrent units (Bi-GRU) and self-attention (SA) mechanism, and the CSAFEL is comprised of convolutional block attention module (CBAM) and residual network-18 (ResNet-18). In this framework, the combination of deep learning neural networks, SA mechanism, and CBAM assigns different important degrees to features, increases the diversities of feature extraction, and improves the accuracy of HAR. To improve the practicability of the proposed framework, a daily and aggressive activity dataset (DAAD) is collected by our laboratory for performance evaluation. The experiments are conducted on the wireless sensor data mining laboratory (WISDM) dataset, the University of California Irvine-HAR (UCI-HAR) dataset, and the collected DAAD dataset. The results show that the proposed DMEFAM can upgrade the recognition performance effectually, compared to other advanced HAR frameworks, with recognition accuracies of 97.9%, 96.0% and 99.2% on the WISDM dataset, the UCI-HAR dataset, and the DAAD dataset, respectively.

Integrating reconfigurable intelligent surface (RIS) and non-orthogonal multiple access (NOMA) provides a promising solution to support large-scale secure communications. However, reducing the impact of transceiver hardware impairments (HWI) on the performance of RIS-aided NOMA secure networks remains a challenge. In this paper, we propose a robust transmission scheme for RIS-aided NOMA secure networks with transceiver HWI under two proposed eavesdropping scenarios and two artificial noise methods. A closed-form expression for the distortion noise power caused by transceiver HWI in NOMA networks is derived to quantify transceiver HWI. The sum secrecy rate of NOMA users with imperfect successive interference cancellation (SIC) is maximized under distortion noise for more robust security. Further, to tackle the resulting non-convex problem, we decompose it into two sub-problems. Both of them are converted into convex problems using the multi-dimensional quadratic transform (MDQT) method, semi-definite relaxation (SDR) technique, and the successive convex approximation (SCA) algorithm, and then employing the alternate optimization (AO) algorithm solve them. Numerical results demonstrate that the proposed scheme has superior security performance over orthogonal multiple access (OMA) networks, time division multiple access (TDMA) networks, and traditional NOMA networks, achieving more robust performance than the conventional scheme that ignores HWI and imperfect SIC.

The advances in deep learning with the ability to automatically extract advanced features have achieved a bright prospect for human activity recognition (HAR). However, the traditional HAR methods still have the deficiencies of incomplete feature extraction, which may lead to incorrect recognition results. To resolve the above problem, a novel framework for spatiotemporal multi-feature extraction on HAR called CapsGaNet is propounded, which is based on capsule and gated recurrent units (GRU) with attention mechanisms. The proposed framework involves a spatial feature extraction layer consisting of capsule blocks, a temporal feature extraction layer consisting of GRU with attention mechanisms, and an output layer. At the same time, considering the actual demands for recognizing aggressive activities in some specific scenarios like smart prison, we constructed a daily and aggressive activity dataset (DAAD). Moreover, based on the acceleration characteristics of aggressive activity, a threshold-based approach for aggressive activity detection is propounded to meet the needs of high real-time and low computational complexity in prison scenarios. The experiments are performed on the wireless sensor data mining (WISDM) dataset and the DAAD dataset, and the results verify that the propounded CapsGaNet could effectually improve the recognition accuracy. The proposed threshold-based approach for aggressive activity detection provides a more effective HAR way by using smart sensor devices in smart prison scenarios.

Human activity recognition (HAR) has become a research hotspot in the field of artificial intelligence and pattern recognition. However, the HAR system still has some deficiencies in the aspects of platform algorithms and wireless access technologies. On the one hand, some state-of-the-art frameworks such as convolutional neural network (CNN) and recurrent neural network (RNN) have been proven successfully in classification tasks of HAR, while those frameworks just identify the feature data of activity but ignore the spatial relationship among features, which may lead to incorrect recognition. On the other hand, some existing transmission modes, such as Bluetooth and 4G, are difficult to realize real-time transmission in the case of a large range and low-power consumption. In this paper, a real-time human activity recognition system based on capsule and “long range" (LoRa) is presented, which pioneers the application of capsule to HAR. The capsule framework encapsulates the multiple convolution layers in parallel to solve the defect that current frameworks cannot identify the spatial relationship among features. Simultaneously, the combination of long-distance transmission and low-power consumption is achieved by using LoRa networking technology instead of other existing transmission modes. The experiments are performed on the dataset WISDM that is collected by the Wireless Sensor Data Mining Lab in Fordham University, and the results demonstrate that the proposed capsule framework achieves a higher classification result than CNN and RNN, and the proposed system makes the real-time HAR based on intelligent sensor devices possible in some special scenarios such as smart prison.