Map Matching Algorithm: Empirical Review Based on Indian OpenStreetMap Road Network Data
Abstract: Locating devices on the road network is crucial for any location-based system. Accuracy of map matching algorithms may highly affect the accuracy of any location-based service. This paper includes an empirical review of five major map matching algorithms for locating a device on a digital road network. A standard dataset was used to simulate the working of map matching algorithms. After ascertaining the accuracy of map matching algorithms, it was tested on a real road network. Six different routes varying from 0.6 kilometers to 32 kilometers, covering a total distance of 82.2 kilometers were included in the experiment. Performance of map matching algorithms was evaluated on a total of 2094 road nodes with 1271070 Global Positioning System (GPS) points on the basis of matched, unmatched nodes with root mean square error. It was concluded that Hidden-Markov Model based map matching algorithms has reasonably good accuracy (96% using global data and 89% using Indian dataset) and execution time in comparison to geometric, topological, Kalman filter and Frechet distance based algorithms.
Keywords: Hidden markov model, kalman filter, frechet distance, OSM, GPS.
Received October 22, 2019; accepted June 8, 2021
Compact Tree Structures for Mining High Utility Itemsets
|
Anup Bhat Department of Computer Science and Engineering, Manipal Academy of Higher Education, India This email address is being protected from spambots. You need JavaScript enabled to view it. |
Harish Venkatarama Department of Computer Science and Engineering, Manipal Academy of Higher Education, India This email address is being protected from spambots. You need JavaScript enabled to view it. |
Geetha Maiya Department of Computer Science and Engineering, Manipal Academy of Higher Education, India This email address is being protected from spambots. You need JavaScript enabled to view it. |
Abstract: High Utility Item set Mining (HUIM) from large transaction databases has garnered significant attention as it accounts for the revenue of the items purchased in a transaction. Existing tree-based HUIM algorithms discard unpromising items and require at most two database scans for their construction. Hence, whenever utility threshold is changed, the trees have to be reconstructed from scratch. In this regard, the current study proposes to not only incorporate all the items in the tree structure but compactly represent transaction information. The proposed trees namely- Utility Prime Tree (UPT), Prime Cantor Function Tree (PCFT), and String based Utility Prime Tree (SUPT) store transaction-level information in a node unlike item-based prefix trees. Experiments conducted on both real and synthetic datasets compare the execution time and memory of these tree structures with a proposed Utility Count Tree (UCT) and existing IHUP, UP-Growth trees. Due to transaction-level encoding, these structures consume significantly less memory when compared to the tree structures in the literature.
Keywords: High utility itemset mining, tree based algorithms.
Received February 11, 2020; accepted July 13, 2021
An Architecture of IoT-Aware Healthcare Smart System by Leveraging Machine Learning
Abstract: In a healthcare environment, Internet of Things (IoT) sensors’ devices are integrated to help patients and Physicians remotely. Physicians interconnect with their patients to monitor their current health situation. However, a considerable number of real-time patient data produced by IoT devices makes healthcare data intensive. It is challenging to mine valuable features from real-time data traffic for efficient recommendations to patients. Thus, an intelligent healthcare system must analyze the real-time health conditions and predict suitable drugs based on the diseases’ symptoms. In this paper, an IoT architectural model for smart health care is proposed. This model utilizes clustering and Machine Learning (ML) techniques to predict suitable drugs for patients. First, Spark is used to manage the collected data on distributed servers. Second, the K-means clustering algorithm is used for disease-based categorization to make groups of the related features. Third, predictor techniques, i.e., Naïve Bayes and random forest, are used to classify suitable drugs for the patients. Two standard Unique Client Identifier (UCI) machine learning datasets have been conducted in the experiments. The first dataset consists of different types of thyroid diseases, while the second dataset contains drugs with recommended medicines. The experimental results depict that the performance, i.e., the accuracy of the proposed model, is superior in predicting the suitable drugs for patients, by which it provides a highly effective delivery healthcare service in IoT. Random Forest correctly classified 99.23% instances while Naive Bayes results are 95.52%.
Keywords: IoT, machine learning, big data, cloud computing, healthcare.
Received July 9, 2020; accepted January 19, 2021
An Improved Iris Localization Method
Abstract: Iris research has become an inevitable trend in the application of identity recognition due to its uniqueness, stability, non-aggression and other advantages. In this paper, an improved iris localization method is presented. When the iris inner boundary is located, a method for extracting the iris inner boundary based on morphology operations with multi-structural elements is proposed. Firstly, the iris image is pre-processed, and then the circular connected region in the pre-processed image is determined, the parameters of the circular connected region is extracted, finally the center and the radius of the circular connected region is obtained, i.e., the iris inner boundary is excavated. When the iris outer boundary is located, a method for locating iris outer boundary based on annular region and improved Hough transform is proposed. The iris image is first filtered, and then the filtered image is reduced and an annular region is intercepted, finally Hough transform is used to search the circle within the annular region, i.e., the center and the radius of the iris outer boundary is obtained. The experimental results show that the location accuracy rate of this proposed method is at least 95% and the average running time is increased by 46.2% even higher. Therefore, this proposed method has the advantages of high speed, high accuracy, strong robustness and practicability.
Keywords: Iris location, image pre-process, multi-structural elements, hough transform.
Received July 21, 2020; accepted July 4, 2021
Applying Deep Convolutional Neural Network (DCNN) Algorithm in the Cloud Autonomous Vehicles Traffic Model
Abstract: Connected and Automated Vehicles (CAVs) is an inspiring technology that has an immense prospect in minimizing road upsets and accidents, improving quality of life, and progressing the effectiveness of transportation systems. Owing to the advancements in the intelligent transportation system, CAV plays a vital role that can keeping life lively. CAV also offers to use to transportation care in producing societies protected more reasonable. The challenge over CAV applications is a new-fangled to enhance safety and efficiency. Cloud autonomous vehicles rely on a whole range of machine learning and data mining techniques to process all the sensor data. Supervised, Unsupervised, and even reinforcement learning are also being used in the process of creating cloud autonomous vehicles with the aim of error-free ones. At first, specialized algorithms have not been used directly in the cloud autonomous vehicles which need to be trained with various traffic environments. The creation of a traffic model environment to test the cloud autonomous vehicles is the prime motto of this paper. The deep Convolutional Neural Network (CNN) has been proposed under the traffic model to drive in a heavy traffic condition to evaluate the algorithm. This paper aims to research an insightful school of thought in the current challenges being faced in CAVs and the solutions by applying CNN. From the simulation results of the traffic model that has traffic and highway parameters, the CNN algorithm has come up with a 71.8% of error-free prediction.
Keywords: Cloud computing, neural network, prediction model, resource selection.
Received October 1, 2020; accepted April 7, 2021
Reinforcement Energy Efficient ant Colony Optimization of Mobile Ad Hoc Multipath Routing Performance Enhancement
Abstract: The Mobile Ad hoc Network (MANET) is a collection of mobile nodes that operates without infrastructure and is self-administrative. It often changes the topology called dynamic network. The nodes are in the dynamic network spending more energy on path finding that resultant the node quickly draining and being inactive in the network. To overcome this issue, propose an Improve Energy Capability Backup Route ant colony optimization (IEC-BR) algorithm for reducing energy usage, enhancing network lifetime, improving packet delivery, and decreasing end-to-end delay. The algorithm uses a backup route for minimizing the route researching and node energy preservation. Stability factors and Quality of Services (QoS) parameters help to enhance the network lifetime and QoS. The stability factor uses the node residual energy and the link lifetime. Hop count and bandwidth are used for the QoS parameters. The simulation result has proved the proposed IEC-BR technique improves packet delivery, node lifetime by 31% and reduces the network delay compared to the traditional Ad hoc On-demand Distance Vector (AODV) and Multi Objective Ad hoc On-demand Distance Vector (MOAODV) routing protocols.
Keywords: QoS, MANET, energy efficient, routing, ACO.
Received January 26, 2021; accepted September 16, 2021
A Multi-Group Structural Equation Modeling For Assessing Behavioral Intention of Using Mobile Cloud Computing-The Case of Jordanian Universities During The Covid19 Pandemic
Abstract: The adoption of new technologies in Jordanian Universities related to cloud services, shows differences in practices between faculty and staff members. Resistance to adoption may accrue by faculty and staff members who are accustomed and favoring old practices. A questionnaire was developed based on the Unified Theory of Acceptance and Use of Technology (UTAUT) model to identify factors that affect behavioral intentions that lead to the use of mobile cloud computing during the covid-19 pandemic, taking into consideration Work-type as the mediating factor. Five Jordanian Universities participated in this study, with a total response of 153 faculty and staff members. The conceptual proposed model was tested to ensure the fitness of the structural model for providing correct estimations. The collected sample was subjected to confirmatory factor analysis to ensure construct, convergent and discriminant validity. The results came positive in terms of composite reliability as they were above 0.70, for Average Variance Extracted (AVE) it came more than 0.05and Cronbach alpha exceeded 0.70. The results revealed the fitness of the proposed model to measure differences in behavioral intentions towards adopting mobile cloud services between faculty members and employees. Moreover, the results showed that work type had some interesting moderating impact on the tested relationships. Moreover, the results showed that there is a high Behavioral Intention (BI) between faculty and staff to use mobile cloud services and solutions within their workplace. In addition, the results showed some inequalities of the behavioral intention toward the adoption of mobile cloud services in Jordanian Universities between the two groups. These results call the university administration to clarify these factors for user groups to obtain a better judgment on investment and future practices for using new technologies.
Keywords: Mobile cloud computing, structural equation modeling, UTAUT model, Jordanian universities.
Received January 31, 2021; accepted October18, 2021
IoT Security Using AES Encryption Technology based ESP32 Platform
Abstract: The Internet of Things (IoT) is one of the most important modern technologies that have attracted the most interesting areas of life, whether industrial, academic, or other, in recent years. The main goal is to integrate the physical world with the digital world through a seamless ecosystem, and this constitutes a new era for the Internet. This technology provides high commercial value to enterprises as it provides many opportunities in many applications such as energy, health, and other sectors. However, this technology suffers from many security problems, as it is considered the biggest challenge due to its complex environment and the limited resources of its devices. There is a lot of research to find successful security solutions in IoT, in this research, a proposed solution to secure IoT systems using Advanced Encryption Standard (AES) technology is achieved. Some sensors were linked as an example of the Internet of Things. The data is received by the card created and developed by Espressif Systems (ESP32) module, where its encrypted then sends to the internet site through an authorized person to be received from anywhere, and it is also possible to receive it via a published IP which is announced within the internal network of the ESP32 device module. The decryption part is proposed at last to find out the true values of the sensors. The proposed approach shows good secured and balanced results at the end.
Keywords: ESP32, IoT Security, secure boot, AES.
Received May 27, 2021; accepted September 9, 2021
Performance Analysis of Efficient Spectrum Utilization in Cognitive Radio Networks by Dynamic Spectrum Access and Artificial Neuron Network Algorithms
Abstract: Efficient spectrum utilization is a prominent issue in cognitive radio networks. Owing to this, power allocation policies are proposed which underlay cognitive radio networks together among all prime nodes, secondary nodes, eavesdropper and secondary sender powered by renewable energy that is harvested from primary sender to acquire improved energy efficiency to enhance transmission rate, throughput, and Spectrum Utilization (SU). As a result, there is a need for combination of Dynamic Spectrum Access (DSA) algorithm, Artificial Neuron Network (ANN) algorithm which will make an allotment of obtainable network assets for various elements challenging for their resources. The prime objective of this paper is to intend a route control based multi-path Quality of Service (QoS) and to find substitute paths between Secondary User (SU) source and SU destination fulfilling QoS metrics, specifically providing maximal throughput and minimal delay. In order that primary substitute channels along the paths are used completely to reduce data packets loss by using Network Simulator 2 (NS2) software tool.
Keywords: Cognitive radio networks, DSA, ANN, SU, radio frequency, energy harvesting.
Received November 24, 2019; accepted February 17, 2021
An Effective Fault-Tolerance Technique in Web Services: An Approach based on Hybrid Optimization Algorithm of PSO and Cuckoo Search
Abstract: Software rejuvenation is an effective technique to counteract software aging in continuously-running application such as web service based systems. In client-server applications, where the server is intended to run perpetually, rejuvenation of the server process periodically during the server idle times increases the availability of that service. In these systems, web services are allocated based on the user’s requirements and server’s facilities. Since the selection of a service among candidates while maintaining the optimal quality of service is an Non-Deterministic Polynomial (NP)-hard problem, Meta-heuristics seems to be suitable. In this paper, we proposed dynamic software rejuvenation as a proactive fault-tolerance technique based on a combination of Cuckoo Search (CS) and Particle Swarm Optimization (PSO) algorithms called Computer Program Deviation Request (CPDR). Simulation results on Web Site Dream (WS-DREAM) dataset revealed that our strategy can decrease the failure rate of web services on average 38.6 percent in comparison with Genetic Algorithm (GA), Decision-Tree (DT) and Whale Optimization Algorithm (WOA) strategies.
Keywords: Software aging, software rejuvenation, web service, cuckoo search, particle swarm optimization.
Received January 2, 2020; accepted March 23, 2021
An Efficient Intrusion Detection Framework Based on Embedding Feature Selection and Ensemble Learning Technique
Abstract: Network security has emerged as a crucial universal issue that affects enterprises, governments, and individuals. The strategies utilized by the attackers are continuing to evolve, and therefore the rate of attacks targeting the network system has expanded dramatically. An Intrusion Detection System (IDS) is one of the significant defense solutions against sophisticated cyberattacks. However, the challenge of improving the accuracy, detection rate, and minimal false alarms of the IDS continues. This paper proposes a robust and effective intrusion detection framework based on the ensemble learning technique using eXtreme Gradient Boosting (XGBoost) and an embedded feature selection method. Further, the best uniform feature subset is extracted using the up-to-date real-world intrusion dataset Canadian Institute for Cybersecurity Intrusion Detection (CICIDS2017) for all attacks. The proposed IDS framework has successfully exceeded several evaluations on a big test dataset over both multi and binary classification. The achieved results are promising on various measurements with an accuracy overall, precision, detection rate, specificity, F-score, false-negative rate, false-positive rate, error rate, and The Area Under the Curve (AUC) scores of 99.86%, 99.69%, 99.75%, 99.69%, 99.72%, 0.17%, 0.2%, 0.14%, and 99.72 respectively for abnormal class. Moreover, the achieved results of multi-classification are also remarkable and impressively great on all performance metrics.
Keywords: Network security, intrusion detection, ensemble learning, xgboost algorithm, features selection.
Received February 18, 2020; accepted August 29, 2021
CTL Model Checking Based on Binary Classification of Machine Learning
|
Weijun Zhu School of Computer and Artificial Intelligence Zhengzhou University, China This email address is being protected from spambots. You need JavaScript enabled to view it. |
Huanmei Wu College of Public Health Temple University, USA This email address is being protected from spambots. You need JavaScript enabled to view it. |
Abstract: In this study, we establish and pioneer an approximate Computational Tree Logic (CTL) Model Checking (MC) technique, in order to avoid the famous State Explosion (SE) problem in the Computational Tree Logic Model Checking (CTLMC). To this end, some Machine Learning (ML) algorithms are introduced and employed. On this basis, CTL model checking is induced to binary classification of machine learning, by mapping all the two different results of CTL model checking into all the two different results of binary classification of machine learning, respectively. The experimental results indicate that the newly proposed approach has a maximal accuracy of 100% on our randomly generated data set, compared with the latest algorithm in the classical CTL model checking. Furthermore, the average speed of the new approach is at most 120 thousand times higher than that of the latest algorithm, which appears in the current version of a popular model checker called NuXMV, in the classical CTL model checking. These observations prompt that the new method can get CTL model checking results quickly and accurately, since the SE problem is avoided completely.
Keywords: Model checking, computational tree logic, machine learning, binary classification.
Received October 10, 2020; accepted October 14, 2021
Multichannel Based IoT Malware Detection System Using System Calls and Opcode Sequences
Abstract: The rapid development in the field of the Internet of things gives rise to many malicious attacks, since it holds many smart objects whose lack of an efficient security framework. These kinds of security issues bring the entire halt-down situation to all smart objects that are connected to the network. In this work, multichannel Convolutional Neural Network (CNN) is proposed whereas each channel’s CNN works on each type of input parameter. This model has two channels connected in a parallel manner, with one CNN taking an opcode sequence as input and the other CNN running with system calls. These extracted system calls and opcode sequences of elf files were discriminated against using two more deep learning algorithms along with multichannel CNN, namely Recurrent Neural Network (RNN) and CNN, and a few recent existing solutions. The performance analysis of the aforementioned algorithms has been carried out and evaluated using accuracy, precision, recall, F1-measure, and time. The experimental results show that multichannel CNN outperforms the remaining considered techniques by achieving a high accuracy of 99.8% for classifying malicious samples from benign ones. The real-time Internet of Things (IoT) malware samples were collected from the IoT honeyPot (IOTPOT), which emulates different CPU architectures of IoT devices.
Keywords: System calls, IoT malwares, fog computing, RNN, CNN, multichannel CNN.
Received November 27, 2020; accepted July 29, 2021
An Efficient Ensemble Architecture for Privacy and Security of Electronic Medical Records
|
Ömer Kasım Department of Electrical and Electronics Engineering Kutahya Dumlupinar University, Turkey This email address is being protected from spambots. You need JavaScript enabled to view it. |
Abstract: Electronic medical records, one of the sensitive data, are stored in public or private cloud service providers. Cloud systems provide security with firewall and intrusion detection systems, and these systems ensure privacy with access control and end-to-end encryption. However, while sending data to the cloud system, attackers can capture the data with the help of Man in the Middle attacks and vulnerabilities of the storage systems. In the middleware architecture proposed in this study, access control protocol, key distributor and end-to-end hybrid encryption which are based on user roles were innovatively used to overcome security issues in data transmission. In this system, writing and updating requests are encrypted asymmetrically, and reading requests were encrypted symmetrically. This solution distinguishes the proposed method from previous studies. According to this solution the operating performance of the system is increased. In addition, the attacker cannot see the actual data in a cyber-attacks because the sensitive data is distributed to the users with their private keys. This result shows that the access, write and update of electronic medical records are performed with the principles of security and privacy.
Keywords: Electronic medical records, sensitive data security, hybrid encryption and decryption, access control.
Received December 15, 2020; accepted August 17, 2021
Multi-Pose Facial Expression Recognition Using Hybrid Deep Learning Model with Improved Variant of Gravitational Search Algorithm
Abstract: The recognition of human facial expressions with the variation of poses is one of the challenging tasks in real-time applications such as human physiological interaction detection, intention analysis, marketing interest evaluation, mental disease diagnosis, etc. This research work addresses the problem of expression recognition from different facial poses at the yaw angle. The major contribution of the paper is the proposal of an autonomous pose variant facial expression recognition framework using the amalgamation of a hybrid deep learning model with an improved quantum inspired gravitational search algorithm. The hybrid deep learning model is the integration of the convolutional neural network and recurrent neural network. The applicability of the hybrid deep learning model can be considered as significant if the feature set is efficiently optimized to have the discriminative features respective to each expression class. Here, the Improved Quantum Inspired Gravitational Search Algorithm (IQI-GSA) is utilized for the selection and optimization of features. The IQI-GSA method is significant for optimizing the features compared to quantum-behaved binary gravitation search algorithm for handing the local optima and stochastic characteristics. Comparing with state-of-art techniques, the proposed framework exhibits the outperformed recognition rate for experimentation on Karolinska Directed Emotional Faces (KDEF) and Japanese Female Facial Expression (JAFFE) datasets.
Keywords: Deep learning, emotion recognition, quantum computing, swarm intelligence, convolutional neural networks, recurrent neural networks.
Received January 9, 2021; accepted August 15, 2021