Pervasive and Mobile Computing最新文献

The current extensive deployment of IoT devices, crucial for enhancing smart computing applications in diverse domains, necessitates the utilization of essential 5G features, notably network slicing, to ensure the provision of distinct and reliable services. However, the voluminous, dynamic, and varied nature of IoT traffic introduces complexities in network flow classification, traffic analysis, and the accurate determination of network requirements. These complexities pose a significant challenge in effectively provisioning 5G network slices across various applications. To address this, we propose an innovative approach for network traffic classification, comprising a pipeline that integrates Principal Component Analysis (PCA) with KMeans clustering and the Hellinger distance measure. The application of PCA as the initial step effectively reduces the dimensionality of the data while retaining most of the original information, which significantly lowers the computational demands for the subsequent KMeans clustering phase. KMeans, an unsupervised learning method, eliminates the labor-intensive and error-prone process of data labeling. Following this, a Hellinger distance-based recursive KMeans algorithm is employed to merge similar clusters, aiding in the determination of the optimal number of clusters. This results in final clustering outcomes that are both compact and intuitively interpretable, overcoming the inherent limitations of the traditional KMeans algorithm, such as its sensitivity to initial conditions and the requirement for manually specifying the number of clusters. An evaluation of our method using a real-world IoT dataset has shown that our pipeline can efficiently represent the dataset in three distinct clusters. The characteristics of these clusters can be readily understood and directly correlated with various types of network slices in the 5G network, demonstrating the efficacy of our approach in managing the complexities of IoT traffic for 5G network slice provisioning.

Roads can develop potholes over time, posing hazards to traffic. However, regular road damage inspections is challenging due to the high cost of road surveys. By applying object detection models on footage acquired from dashboard cameras installed in garbage trucks that operate across the city, we can conduct road surveys at a low cost. In our previous work we introduced the Ensemble of Classification Mechanisms (ECM), which suppresses false positives by cross-verifying objects detected by an object detection model using a different image classification model. However, ECM faces challenges in achieving both fast inference speed and high detection performance simultaneously. It also struggles in environments where roads vary in their suitability for false positive suppression. To address these issues, we propose the Dynamic Ensemble of Classification Mechanisms (DynamicECM). This approach utilizes ECM selectively, enabling stable inference with minimal false positive suppression. To evaluate our new method, we constructed an evaluation dataset comprising objects that cause false positives in pothole detection. Our experiments demonstrate that ECM achieves higher precision, average precision (AP), and F1 scores compared to existing methods. Furthermore, DynamicECM improves the trade-off between speed and detection performance, outperforming ECM, and achieves stable inference even in challenging datasets where ECM would falter. Our method is highly scalable and expected to contribute to the stability and efficiency of inference across various object detection models. In our previous work we developed an Ensemble of Classification Mechanisms (ECM), which suppresses false positives by rechecking objects detected by an object detector with a different image classification model. However, ECM cannot achieve both fast inference speed and high detection performance at the same time. It also struggles in environments that have a mixture of roads suitable for false positive suppression and unsuited for false positive suppression. To solve these problems, we propose “Dynamic Ensemble of Classification Mechanisms”. Since this method uses ECM only when deemed necessary, stable inference can be achieved efficiently without excessive suppression of false positives. In order to evaluate our new method, we constructed an evaluation dataset that includes objects that cause false positives in pothole detection. Our evaluation experiments show that ECM achieves higher precision, AP, and F1 compared to existing methods. In addition, DynamicECM improves the trade-off between speed and detection performance better than ECM, and achieves stable inference on datasets that would ECM would struggle on. Our method is highly scalable and expected to contribute to the stability and efficiency of inference for various object detection models.

Mobile Crowd Sensing (MCS) technology, as an emerging data collection paradigm, offers distinct advantages, particularly in applications like smart city management. However, existing researches inadequately address the comprehensive solution to the problem of reliable task allocation according to the requirements such as task budget, sensory data quality, and real-time data collection, especially under varying participant engagement in MCS systems. To bridge this gap, we propose the Grid-based Two-stage Dynamic Incentive Mechanism (GTDIM). In the first stage, the Candidate Participant Set (CPS) establishment phase, participants receive compensation for collecting sensory data when a sufficient number are available. When participants are insufficient, additional rewards inspired by the grid division of sensing areas are progressively offered to attract more participants. In the subsequent stage, utilizing the established CPS, participants are selected through a greedy algorithm based on the newly devised Participant Matching Index (PMI), which integrates various participant features. Extensive simulation results reveal the impact of PMI on participant selection. Numerical findings conclusively demonstrate GTDIM’s superior performance over baseline incentive mechanisms in terms of task assignment ratio, participant payment, and especially when dealing with larger sensing tasks.

The proposal of Integrated Sensing and Communications has once again drawn researchers’ attention to WiFi sensing, propelling applications based on WiFi sensing into an advanced stage. However, the current field of activity recognition only identifies fixed categories of activities, neglecting the growing demand for perceiving activity types in real applications over time. In response to the issue, we present WiCAR, a WiFi activity recognition system designed for class incremental scenarios. WiCAR takes antenna array-fused image data as input, employing the Wi-RA model with parallel stacked activation functions as its backbone network. To alleviate the typical catastrophic forgetting issue in class-incremental learning, WiCAR employs a strategy of replaying known data. Additionally, we adopts knowledge distillation to improve accuracy among old samples during the incremental process. To tackle the imbalance in the number of samples between old and new classes, the model is updated through weight alignment. This serious of strategies endows the system with the capability to progressively learn and handle new classes. We conducted extensive experiments to evaluate the system performance. The experimental results demonstrate that our system exhibits excellent performance regardless of the number of tasks, whether tasks are uniform or non-uniform, and the order of task arrivals. The highest average accuracy reaches 96.429%, and even in the presence of six incremental stages, the average accuracy remains at 92.867%.

Vehicular Ad Hoc Network (VANET) facilitates the exchange of vehicular information through Vehicle-to-Vehicle (V2V) communication, contributing to Cooperative Intelligent Transportation Systems (C-ITS). The transmitted messages among vehicles are vulnerable to various security threats executed by malicious insider nodes. The dynamic VANET necessitates context-aware solutions for detecting various security attacks. Existing learning and deterministic mechanisms showed high detection accuracy for attacks on which they were trained explicitly for large datasets. Therefore, we propose an intelligent framework utilizing Deep Reinforcement Learning (DRL) for attack detection in evolving scenarios and mitigate the need for extensive training datasets. Our approach employs a Deep Q Network (DQN) trained on a compact dataset encompassing multiple attacks. The trained model is then applied to an unknown and extensive dataset, detecting various attacks with high accuracy. Notably, the model autonomously updates itself upon observing changes in the network context. This framework represents a promising security solution that is effective and adaptable for V2V communication in VANET.

Great interest is directed toward real-time applications to determine the exact location of sensor nodes deployed in an area of interest. In this paper, we present a novel approach using a combination of the Kalman filter and regularized bounding box method for localizing unknown nodes in an area using an FPGA-enhanced edge computing UAV whose trajectory is known and is represented as the position of many anchors. The UAV is equipped with a GPS system that allows it to gather location data of sensor nodes as it moves around its environment. We employ a regularized bounding box to predict the positions of the unknown nodes using regularization factors and we use the Kalman filter algorithm to smooth and improve the accuracy of the sensor nodes to be localized. In order to localize the unknown nodes, the UAV receives the number of hops from each node and uses this information as input to the localization algorithm. Furthermore, the use of an FPGA board allows for real-time processing of sensory data, enabling the UAV to make fast and accurate decisions in dynamic environments. The localization algorithm was implemented on the FPGA board “Zynq MiniZed 7007s evaluation board” using Xilinx blocks in Simulink, and the generated code was converted into VHDL using Xilinx System Generator. The algorithm was simulated and synthesized using “Vivado” software. In fact, the proposed system was evaluated by comparing the performances achieved through two different implementations: Hardware and Software implementation. In effect, the performance of FPGA hardware implementation presents a new achievement in localization due to its easy testing and fast implementation. Our results show that this approach can efficiently locate unknown nodes with good latency and high accuracy. In fact, the execution time of the FPGA-integrated algorithm is reduced by about 60 times compared to the software implementation and the power consumption is about 100 mW, which proves the suitability of FPGA for localization in WSNs, offering a promising solution for various mobile WSN applications.

Due to enormous growth of Internet of Things (IoT) in the last decade, the amount of data generated through smart devices is increasing exponentially. Fog computing has emerged as a potential technology to deal such a huge volume of data in which task offloading is the most important aspect which has attracted significant attention. Many research works have been carried out, however, task offloading with latency sensitivity, reliability and result migration over a mobile user environment is still not widely addressed. In this paper, we propose a method for delay-sensitive and fault minimized task offloading for service requests made through a mobile/vehicular end user environment implemented via Software Defined Network (SDN) controllers integrated with the fog layer. This is a novel multi-phased model involving determining the optimal number of SDN controllers, clustering of the fog nodes (FNs) on the basis of SDN proximities, task prioritization and Gravitational Search Algorithm (GSA) based target FN selection. The simulation outcomes of our proposed approach show that there is a reduction in delay by around 23%–30% and around 60%–80% lesser number of tasks unassigned in each round as compared to two base algorithms.

In this study, the Efficient Multikeyword Fully Homomorphic Search Encryption (EMK-FHSE) model is proposed to improve cloud storage security for sensitive data. When fully homomorphic encryption (FHE) and search encryption (SE) technologies are coupled, Fully Homomorphic Search Encryption (FHSE) is a strategy that realizes the shared information's controlled privacy and search security. As more and more encrypted data is kept on cloud servers (CSs), a single-keyword SE approach may cause multiple keyword index duplication concerns, making it challenging for CSs to search for the encrypted information. To reduce these problems, a novel efficiency bottleneck has been developed. An Adaptive Privacy-Preserving Fuzzy Multi-Keyword Search (APPFMK) approach is presented to address the difficulties of low search effectiveness in a single-keyword searching strategy and the high processing cost of the existing multi-keyword schemes. Cloud servers (CS) hold enormous volumes of encrypted data, and the necessary encrypted index is transmitted to the closest edge node (EN) to enable multi-keyword searches and supported decryption. According to security research, the EMK-FHSE multi-keyword index is safe in distinguishability under chosen keyword attacks. The results section compares the proposed model's search, storage, trapdoor, calculation, storage and validation times to those of several other models. The proposed model could achieve the following values: 60.81 kb for storage, 10.92 for the trapdoor, 6.85 ms for search, 0.44 ms for computation cost by changing the keyword in a trapdoor, 156.31 ms for computation cost by changing the keyword in a dictionary, 0.44 kb for storage cost by changing the keyword in a trapdoor, 1.81 kb for storage cost by changing the keyword in a dictionary and 0.016seconds for verification time, respectively.

While UbiComp research has steadily improved the performance of localisation systems, the analysis of such datasets remains largely unaddressed. In this paper, we present a tool to facilitate querying and analysis of localisation time-series with a focus on semantic localisation. Drawing on well-established models to represent movement and mobility, we first develop a query language for localisation datasets. We then develop a software library in R that implements this querying. We use case studies to demonstrate how our programming tool can be used to query localisation datasets. Our work addresses an important gap in localisation research, by providing a flexible tool that can model and analyse localisation data programmatically and in real time.

Passive Displacement Cooling (PDC) is a relatively recent technology gaining attention as a means of significantly reducing building energy consumption overheads, especially in tropical climates. PDC eliminates the use of mechanical fans, instead using chilled-water heat exchangers to perform convective cooling. In this paper, we identify and characterize the impact of several key parameters affecting occupant comfort in a $1000m^2$ open-floor area (consisting of multiple zones) of a ZEB (Zero Energy Building) deployed with PDC units and tackle the problem of setting the temperature setpoint of the PDC units to assure occupant thermal comfort and yet conserve energy. We tackle two key practical challenges: (a) the zone-level (i.e., occupant-experienced) temperature differs significantly, depending on occupancy levels, from that measured by the ceiling-mounted thermal sensors that drive the PDC control loop, (b) sparsely deployed sensors are unable to capture the often-significant differences in ambient temperature across neighboring zones. Using extensive real-world coarser-grained measurement data (collected over 60 days under varying occupancy conditions), (a) we first uncover the various parameters that affect the occupant-level ambient temperature, and then (b) devise a trace-based model that helps identify the optimum combination of PDC setpoints, collectively across multiple zones, while accommodating variations in the occupancy levels and weather conditions. Using this trace-based model, our OcAPO system can assure ambient temperature experienced by occupants within a tolerance of $0.3°C$. In contrast, the existing approach of occupancy-agnostic, rule-based setpoint control violates this tolerance interval more than 80% of the time. However, this initial model requires unnecessary and continual database lookups and is unable to derive finer-grained setpoints, thereby potentially missing opportunities for additional energy savings. We thus collected data for another 15 days, with finer-grained setpoint control in increments of 0.2${}^{\circ }$ under varying occupancy conditions in the second phase. To determine PDC setpoints efficiently, we subsequently used the empirical data to train a KNN-based regression model. Additional studies on our real-world testbed demonstrate the regressor-based OcAPO approach is able to assure occupant-level ambient temperature within a narrow $0.2°C$ tolerance. We also demonstrate that the regression version of OcAPO can reduce the opening percentage of PDC valves (an in