Sequential recommendation(SR) aims to provide personalized recommendations by capturing behavioral intents from existing user interaction sequences. Most previous studies are based on attention mechanisms; however, these approaches suffer from inherent over-smoothing issues that limit their ability to capture transient behavioral signals reflecting the user's immediate intents in interaction sequences. Recently, frequency-domain analysis methods based on the Fourier transform have garnered significant attention in the sequential recommendation domain. By applying the Fourier transform, interaction sequences can be mapped to the frequency domain, enabling direct analysis and targeted manipulation of distinct frequency components. In addition to the inherent limitations of self-attention mechanisms, sequential recommendation faces persistent challenges such as data sparsity and noise. To address these issues, we propose Frequency-Domain Disentanglement-Fusion and Dual Contrastive Learning for Sequential Recommendation (FDCLRec). FDCLRec replaces self-attention mechanisms with a frequency-domain adaptive filtering module, which decouples sequence patterns into distinct high-/low-frequency components and synthesizes comprehensive sequence representations through adaptively weighted fusion. In addition, two auxiliary contrastive learning tasks(augmented-view contrasting and same-target sequence contrastive learning) are strategically integrated to alleviate data sparsity and interaction noise. Extensive experiments on four real-world datasets demonstrate that our model outperforms baseline methods.

Ensuring safety and operational continuity in underground coal mines requires robust mine monitoring. Traditional methods based on fixed sensors and manual inspections suffer from limited coverage, high cost, and poor real-time performance. Mobile Crowd Sensing (MCS), enabled by miner-carried devices, offers flexible coverage but introduces challenges such as data sparsity, noise, and heterogeneity due to device variability and electromagnetic interference. This article proposes a Bidirectional Long Short-Term Memory/Gated Recurrent Unit-based Truth Discovery (BLGTD) method for mine MCS. The model integrates spatiotemporal sequence modeling with Monte Carlo Dropout-based uncertainty quantification, enabling adaptive fusion of multi-source data. Experimental results show that BLGTD achieves a mean absolute error (MAE) of 0.19 ± 0.01 ppm in CH4concentration estimation when 90% of data comes from reliable miners, yielding a 57.8% improvement over traditional weighted averaging. The method demonstrates strong robustness under conditions of data incompleteness, device heterogeneity, and signal interference.

The Vehicular Edge Computing (VEC) paradigm significantly reduces task processing latency in Internet of Vehicles (IoV) and Intelligent Transportation Systems (ITS) by deploying computational resources at Roadside Units (RSUs). However, the high mobility of vehicles and dynamic task arrivals lead to uneven load distribution among RSUs, severely impacting system performance. Actually, load balancing as an important evaluation metric for VEC system greatly affects the performance of individual edge servers in terms of latency, energy consumption, and task completion rates. In view of this, we propose a Proximal Policy Optimization (PPO) based deep reinforcement learning (DRL) approach to determine the task offloading and migration decisions and incorporate the fairness into the constraint, aiming to achieve efficient load-balanced task offloading in VEC. Particularly, we introduce a metric named Load Balancing Metric (LBM) to optimize RSU resource allocation and employ dynamical task migration strategies to optimize the metric. Simulation results demonstrate that this approach significantly enhances load balancing performance, reduces average latency and energy consumption, and provides an efficient resource scheduling solution for VEC systems.

With the increasing awareness of health, using wearable sensors to monitor individual activities and accurately estimate energy expenditure has become a current research focus. However, existing research encounters challenges including low estimation accuracy, a deficiency of frequency domain features, and difficulty in integrating time domain and frequency domain features. To address these issues, we propose an innovative framework called the Dual-Stream Fusion Network (DSFN). This framework combines the Time Domain Encoding (TDE) module, the Frequency Domain Hierarchical-Split Encoding (FDHSE) module, and a Two-Stage Feature Fusion (TSF) module. Specifically, the temporal stream of the framework employs the TDE module to capture deep temporal features that reflect the complex dynamic variations in time-series data. The frequency domain stream introduces the FDHSE module, which extracts frequency domain features using a multi-level, multi-scale approach, ensuring a comprehensive and diverse representation of frequency information. Through this dual-stream architecture, our model effectively learns both time and frequency domain features, addressing the limitations of frequency domain features observed in prior studies. Additionally, we propose the TSF module to fully integrate time and frequency domain features, effectively overcoming the challenge of fusing these two types of features. We conducted experiments on two public datasets, namely the GOTOV dataset (elderly people) and the JSI dataset (young people). Experimental results demonstrate that our method achieves excellent performance across different age groups. Compared to the baseline models, the proposed DSFN significantly improves the accuracy of human energy expenditure estimation.

In recent years, railway construction has been plagued by frequent safety accidents, with workers’ safety during the construction process remaining a major concern. To mitigate this issue, intelligent monitoring of railway workers’ activity has been proposed as a means of improving the safety coefficient of construction. Human activity recognition (HAR) based on wearable devices holds significant application value in areas such as health monitoring, motion analysis, and intelligent assistance. Recently, convolutional neural networks (CNNs) have gained extensive adoption and demonstrated outstanding performance in HAR. However, current HAR research still faces some challenges, including problems with establishing spatial–temporal dependencies and addressing the demand for lightweight models. To address the above issues, we propose a lightweight dual-stream convolution model (LDSC) based on deformable convolution and hierarchical segmentation. The model adaptively captures significant variations in sensor readings over time from portable cards of railway personnel through a temporal stream and learns the interactive information among sensor channels over a spatial stream. LDSC consists of three lightweight convolutional modules that combine deep convolution and point convolution to reduce model parameters, thus meeting the demand for a lightweight model. Experiments and ablation studies are conducted on three available datasets (UCI-HAR, UniMiB-SHAR, and WISDM) to evaluate the proposed model. The experimental results indicate that our model outperforms existing state-of-the-art methods in terms of recognition accuracy, validating the effectiveness and feasibility of LDSC. In addition, theoretical analysis and ablation experiments demonstrate that the proposed LDSC embodies lightweight characteristics.

In Vehicular Edge Computing (VEC), apart from the Road Side Units (RSUs) that can undertake the computation, smart vehicles that incorporate high-end multi-core processors into On-Board Units (OBU) can also contribute their computing resources for vehicular tasks in a pay-as-you-go fashion. Designing an appropriate pricing strategy for vehicles with abundant computing resources is essential yet challenging, as it requires balancing profit-seeking objectives with the needs of service requestors. On the other hand, considering the perspective of vehicles with offloading requests, task offloading should strike a balance between achieving ultra-low task latency and minimizing the associated offloading costs. To tackle these issues, we propose a Collaborative Computation Offloading Scheme (CCOS) for the VEC system. In particular, we take into account the fluctuation of service pricing, to cater to the monetary constraints of service requesters. A Mixed-Integer Nonlinear Programming (MINLP) problem is formulated to minimize the weighted sum of task completion latency and the offloading costs. The optimization problem is decomposed into two subproblems, i.e., the task offloading problem and the computing resource allocation problem, respectively. The task offloading problem is essentially a combinatorial optimization problem that necessitates exponential time complexity for determining the optimal solution. Hence, a Deep Reinforcement Learning (DRL)-based algorithm is put forward to solve this subproblem. The resource allocation problem, however, has been proven to be a convex optimization problem, and the scheduling and allocation of computing resources can be performed in parallel, since each edge node is aware of its own task offloading requests. Simulation results demonstrate that our strategy outperforms other approaches in terms of the convergence rate, task completion rate, and optimal values.

Mobile Edge Computing (MEC) revolutionizes the traditional cloud-based computing paradigm by moving resources in proximity to the network edge, aiming to cater to the rigorous requirements of emerging latency-sensitive applications. However, the escalating resource demands intensify the competition among user devices (UDs). Thus, it is essential to coordinate task offloading and resource scheduling while ensuring fairness among users in MEC. Despite the crucial role of user fairness in motivating task offloading in MEC, it is often overlooked in existing literature. Therefore, we in this paper propose a caching-enhanced MEC framework and formulate a collaborative service caching, task offloading, and multi-resource allocation problem to maximize average user satisfaction. Multiple factors contribute to the difficulty in solving the optimization problem, including constrained resource capabilities, user mobility, service heterogeneity, and spatial demand coupling. Consequently, we transform the origin problem into two distinct subproblems – the service caching and task offloading problem, and the multi-resource allocation problem, respectively. Then, the Advantage Actor-Critic (A2C) based approach is proposed to address the former problem, while a Lagrangian duality-based approach is adopted to tackle the latter problem. The simulation results demonstrate the superior performance of the proposed solution in comparison to several baseline methods.

The cloud-edge cooperative network presents both opportunities and challenges for latency-sensitive and computation-intensive tasks. Effectively harnessing the strengths of edge computing and cloud computing enables real-time task handling, thus reaching a win-win situation where not only the stated quality of service (QoS) is delivered from the angle of service providers, but also the quality of experience (QoE) is improved from the angle of service requestors. However, due to the unpredictable task generation and time-varying environments, it is challenging to achieve optimal task scheduling and effective resource management and allocation. To address this issue, we propose an innovative cloud-edge framework that incorporates task offloading, service caching, and resource allocation in this paper. In this framework, we can determine where to offload the task, e.g., locally, at the edge, or in the cloud center. In view of the importance of the superior user experience, we aim to maximize the user satisfaction regarding task offloading in this framework. The problem is actually a mixed-integer nonlinear programming (MINLP) problem that entails simultaneously addressing cache decisions, offloading decisions, and resources allocation in a dynamic cloud-edge computing system. Owing to the NP-hardness, our original problem is decomposed into two layers of alternating problems. Specifically, we adopt a genetic algorithm (GA) based approach to jointly make cache and offloading decisions, and then iteratively optimize the communication and computing resources allocation. Extensive experimentation has demonstrated the feasibility and effectiveness of the proposed approach.

Physiological signals have been widely used for emotion recognition, but current works seldom apply the feature fusion and attention technologies to ECG emotion recognition. In this paper, we propose a novel ECG emotion recognition method, which adopts a spatial and temporal ECG emotion recognition model based on dynamic feature fusion (DFF-STM) to learn spatial-temporal representations of different ECG areas. Considering the difference in roles played by the different ECG areas in ECG emotion recognition, a dynamic weight distribution layer is introduced into DFF-STM to extract ECG temporal features and learn weights to adjust (e.g., enhance or weaken) the contribution of the ECG areas at the same time. Finally, we conduct experiments using real ECG data on the AMIGOS dataset to evaluate the performance of the DFF-STM on valence and arousal labels. Experiments show that dynamic feature fusion for ECG emotion recognition is much better than those using only handcraft features and deep features.

Vehicular edge computing (VEC) as a promising computing paradigm has accelerated the reformation of existing dominating computing infrastructures, enabling resource provisioning in close proximity to resource requestors. However, several challenges still exist, including efficient resource scheduling and management, dynamic wireless channel state, and limited bandwidth usage. To address these issues, we introduce the digital twin (DT) technology into VEC, enabling DTs of physical entities in VEC to achieve real-time offloading decision-making in the DT simulation cycle. In particular, we propose a DT-empowered VEC (DT-VEC) architecture, aiming to achieve efficient task offloading while considering extra latency incurred by task migration. We further put forward an efficient algorithm to minimize the response latency for all the tasks in the optimization period. The simulation results have proven that our approach outperforms the other two greedy approaches.

Abstract Cross-domain recommendation (CDR) has become an important research direction in the field of recommender systems due to the increasing demand for personalized recommendations across different domains. However, CDR faces multiple challenges, including data sparsity, popularity bias, and long-tail problems. To address these challenges, we propose a novel framework that combines graph contrastive embedding and multi-head cross-attention transfer for cross-domain recommendation, called GCE-MCAT. Specifically, in the pre-training process, we generate more uniform user and item embeddings through contrastive learning, effectively solving the problem of inconsistent data embedding space distribution and recommendation popularity bias. Moreover, we propose a multi-head cross-attention transfer mechanism that allows the model to extract user common and specific domain features from multiple perspectives and perform cross-domain bidirectional knowledge transfer. Finally, we propose a cross-domain feature fusion mechanism that dynamically assigns weights to common user features and specific domain features. This enables the model to more effectively learn common user interests. We evaluate the proposed framework on three real-world CDR datasets and show that GCE-MCAT consistently and significantly improves recommendation performance compared to state-of-the-art methods. In particular, the proposed framework has demonstrated remarkable effectiveness in addressing long-tail distribution and enhancing recommendation novelty, providing users with more diversified recommendations and reducing popularity bias.

Wearable-sensor-based Human Activity Recognition (HAR) has long been a hot topic in ubiquitous computing, which is benefit by the success of deep learning algorithms. The critical difficulties in multimodal sensing environments are how to represent the spatial-temporal dependencies while concurrently extracting features with high characterization. In this work, we propose a self-attention based deep dilated convolution network. Our method uses two channels, named temporal channel and spatial channel, respectively, to extract the readings-over-time and time-over-readings features from sensor signals. The self-attention mechanism helps directly capture the long time dependence of sensor signals. To extract local features while expanding the receptive field and avoiding information loss caused by pooling and upsampling, we use deep dilated convolution, which expanding the receptive field and avoiding information loss caused by pooling and upsampling. Extensive experiments on a self-built dataset and two available benchmark datasets (PAMAP2, OPPORTUNITY) reveal that the effectiveness of our proposed model is more competitive than the state-of-the-art methods in HAR tasks.

The rapid development of deep reinforcement learning makes it widely used in multi-agent environments to solve the multi-agent cooperation problem. However, due to the instability of multi-agent environments, the performance is insufficient when using deep reinforcement learning algorithms to train each agent independently. In this work, we use the framework of centralized training with decentralized execution to extend the maximum entropy deep reinforcement learning algorithm Soft Actor-Critic (SAC) and proposes the multi-agent deep reinforcement learning algorithm MASAC based on the maximum entropy framework. Proposed model treats all the agents as part of the environment, it can effectively solve the problem of poor convergence of algorithms due to environmental instability. At the same time, we have noticed the shortcoming of centralized training, using all the information of the agents as input of critics, and it is easy to lose the information related to the current agent. Inspired by the application of self-attention mechanism in machine translation, we use the self-attention mechanism to improve the critic and propose the ATT-MASAC algorithm. Each agent can discover their relationship with other agents through encoder operation and attention calculation as part of the critic networks. Compared with the recent multi-agent deep reinforcement learning algorithms, ATT-MASAC has better convergence effect. Also, it has better stability when the number of agents in the environment increases.

Wearable devices provide an extremely convenient way to collect a large amount of human motion data. In this paper, the human motion recognition method based on wearable devices is studied. We smooth the data to remove the noise caused by additional motion first. After that, the characteristic values that can distinguish the types of activities can be extracted. Then, we propose a human motion recognition method based on the improved one-dimensional convolutional neural networks(1D-CNNs). Compared with other traditional classification and recognition methods, the recognition rates of 11 human motions have been greatly improved. The average accuracy of each activity identification can reach 92.8%, while the average precision and recall can reach 98.7% and 92.8%.

The development of sensor technologies and smart devices has made it possible to realize real-time data acquisition of human beings. Human behavior monitoring is the process of obtaining activity information with wearables and computer technology. In this paper, we design a data preprocessing method based on the data collected by a single three-axis accelerometer. We first use Butterworth filter as low-pass filtering to remove the noise. Then, we propose a KGA algorithm to remove abnormal data and smooth them at the same time. This method uses genetic algorithm to optimize the parameters of Kalman filter. After that, we use a threshold-based method to identify falls that are harmful to the elderly. The key point of this method is to distinguish falls from people’s daily activities. According to the characteristics of human falls, we extract eigenvalues that can effectively distinguish daily activities from falls. In addition, we use cross-validation to determine the threshold of the method. The results show that in the analysis of 11 kinds of human daily activities and 15 types of falls, our method can distinguish 15 types of falls. The recognition recall rate in our method reaches 99.1%.

Machine-to-machine (M2M) type communications (MTCs) over cellular networks feature the large number of MTC devices (MTCDs), small and time controlled data transmissions, and rigorous energy limitation. Considering full-duplex (FD) relaying can achieve high spectrum and energy efficiency, this paper proposes an MTC-enabled cellular communication scheme, where a traditional cellular user equipment (UE) is configured as an FD relaying based gateway to assist the uplink transmissions of the served MTCDs. The designed objective is to minimize the aggregate energy consumption of a group consisting of a UE and multiple MTCDs, while fulfilling their minimum throughput requirements. To this end, a convex optimization problem is formulated and a low complexity algorithm is also developed to find the optimal power allocation strategies for the UE and the MTCDs. The simulation results show that the proposed scheme can achieve most of the channel reuse gain of the FD relaying if the self-interference at the UE is controlled below a certain level.

This paper investigates the opportunities and challenges in the use of dynamic radio transmit power control for prolonging the lifetime of body-wearable sensor devices used in continuous health monitoring. We first present extensive empirical evidence that the wireless link quality can change rapidly in body area networks, and a fixed transmit power results in either wasted energy (when the link is good) or low reliability (when the link is bad). We quantify the potential gains of dynamic power control in body-worn devices by benchmarking off-line the energy savings achievable for a given level of reliability.We then propose a class of schemes feasible for practical implementation that adapt transmit power in real-time based on feedback information from the receiver. We profile their performance against the offline benchmark, and provide guidelines on how the parameters can be tuned to achieve the desired trade-off between energy savings and reliability within the chosen operating environment. Finally, we implement and profile our scheme on a MicaZ mote based platform, and also report preliminary results from the ultra-low-power integrated healthcare monitoring platform we are developing at Toumaz Technology.

Wireless sensor networks are increasingly being used for continuous monitoring of patients with chronic health conditions such as diabetes and heart problems. As biomedical sensor nodes become more wearable, their battery sizes diminish, necessitating very careful energy management. This paper proposes feedback-based closed-loop algorithms for dynamically adjusting radio transmit power in body-worn devices, and evaluates their performance in terms of energy savings and reliability as the data periodicity and feedback time-scales vary. Using experimental trace data from body worn devices, we first show that the performance of dynamic power control is adversely affected at long data periods. Next for a given data period we show that modifying the transmit power at too long timescales (around a minute) reduces the efficacy of dynamic power control, while too short a time-scale (few seconds or less) incurs a high feedback signaling overhead. We therefore advocate an intermediate range of time-scales (when permitted by the data periodicity), typically in the few tens of seconds, at which the control algorithms should adapt transmit power in order to achieve maximal energy savings in body-worn sensor devices used for medical monitoring.