Pub Date : 2020-12-14DOI: 10.1109/IOTSMS52051.2020.9340165
Theodoros Nestoridis, C. Oikonomou, Anastasios Temperekidis, F. Gioulekas, P. Katsaros
Crowdsourcing aims to deliver services and content by aggregating contributions from a large user population. For mobile networks and IoT systems, crowdsourcing is used to gather and process sensor data from mobile devices (crowdsensing), in order to deliver real-time, context-aware services and possibly support user collaboration in extended geographic areas. In applications like geonsensitive navigation, location-based activity sharing and recommendations, the challenge of adequate service quality and user experience may be at stake, as the services are provided securely to an ever-growing user population. This happens due to the inherent trade-off between security and real-time performance that ultimately sets in doubt any scalability prospect beyond a certain user-interaction load. This work introduces a publish-subscribe architecture for mobile crowdsensing systems, which can be transparently scaled up to higher usage load, while retaining adequate performance and security by load balancing into multiple MQTT brokers. The security support combines a lightweight TLS implementation with an integrated mechanism for two-level access control: user-device interactions and message topics. We provide proof-of-concept measurements that show how our solution scales to increasing interaction loads through load-balancing the processing cost that includes the overhead of the security mechanisms applied. The system architecture was implemented in a vehicular crowdsensing navigation network that allows to exchange navigation information at real-time, for improved routing of vehicles to their destination.
{"title":"Scalable IoT architecture for balancing performance and security in mobile crowdsensing systems*","authors":"Theodoros Nestoridis, C. Oikonomou, Anastasios Temperekidis, F. Gioulekas, P. Katsaros","doi":"10.1109/IOTSMS52051.2020.9340165","DOIUrl":"https://doi.org/10.1109/IOTSMS52051.2020.9340165","url":null,"abstract":"Crowdsourcing aims to deliver services and content by aggregating contributions from a large user population. For mobile networks and IoT systems, crowdsourcing is used to gather and process sensor data from mobile devices (crowdsensing), in order to deliver real-time, context-aware services and possibly support user collaboration in extended geographic areas. In applications like geonsensitive navigation, location-based activity sharing and recommendations, the challenge of adequate service quality and user experience may be at stake, as the services are provided securely to an ever-growing user population. This happens due to the inherent trade-off between security and real-time performance that ultimately sets in doubt any scalability prospect beyond a certain user-interaction load. This work introduces a publish-subscribe architecture for mobile crowdsensing systems, which can be transparently scaled up to higher usage load, while retaining adequate performance and security by load balancing into multiple MQTT brokers. The security support combines a lightweight TLS implementation with an integrated mechanism for two-level access control: user-device interactions and message topics. We provide proof-of-concept measurements that show how our solution scales to increasing interaction loads through load-balancing the processing cost that includes the overhead of the security mechanisms applied. The system architecture was implemented in a vehicular crowdsensing navigation network that allows to exchange navigation information at real-time, for improved routing of vehicles to their destination.","PeriodicalId":147136,"journal":{"name":"2020 7th International Conference on Internet of Things: Systems, Management and Security (IOTSMS)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124470067","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-12-14DOI: 10.1109/IOTSMS52051.2020.9340224
Maria Bada, Ildiko Pete
Discussions on underground forums provide valuable insights to hackers’ practices, interests and motivations. Although Internet of Things (IoT) vulnerabilities have been extensively explored, the question remains how members of hacker communities perceive the IoT landscape. In this work, we present an analysis of IoT related discussions that are potentially cybercriminal in nature. In particular, we analyse forum threads that discuss the search engine Shodan. The source of these posts is the CrimeBB dataset provided by the Cambridge Cybercrime Centre (CCC)1. We analyse 1051 thread discussions from 19 forums between 2009 and 2020. The overall aim of our work is to explore the main use cases of Shodan and highlight hackers’ targets and motivations. We find that Shodan is versatile and is actively used by hackers as a tool for passive information gathering providing easier access to hackable targets. Our results suggest that Shodan plays a prominent role in various specific use cases including remote control of target devices, building botnets, Distributed Denial of Service attacks and identifying open databases.
{"title":"An exploration of the cybercrime ecosystem around Shodan","authors":"Maria Bada, Ildiko Pete","doi":"10.1109/IOTSMS52051.2020.9340224","DOIUrl":"https://doi.org/10.1109/IOTSMS52051.2020.9340224","url":null,"abstract":"Discussions on underground forums provide valuable insights to hackers’ practices, interests and motivations. Although Internet of Things (IoT) vulnerabilities have been extensively explored, the question remains how members of hacker communities perceive the IoT landscape. In this work, we present an analysis of IoT related discussions that are potentially cybercriminal in nature. In particular, we analyse forum threads that discuss the search engine Shodan. The source of these posts is the CrimeBB dataset provided by the Cambridge Cybercrime Centre (CCC)1. We analyse 1051 thread discussions from 19 forums between 2009 and 2020. The overall aim of our work is to explore the main use cases of Shodan and highlight hackers’ targets and motivations. We find that Shodan is versatile and is actively used by hackers as a tool for passive information gathering providing easier access to hackable targets. Our results suggest that Shodan plays a prominent role in various specific use cases including remote control of target devices, building botnets, Distributed Denial of Service attacks and identifying open databases.","PeriodicalId":147136,"journal":{"name":"2020 7th International Conference on Internet of Things: Systems, Management and Security (IOTSMS)","volume":"87 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116668625","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-12-14DOI: 10.1109/IOTSMS52051.2020.9340166
A. Sreeram, Udhaya S. Ravishankar, Narayana Rao Sripada, Baswaraj Mamidgi
In the literature so far, classification of respiratory diseases with cough signals has typically involved extracting standard spectral features such as Mel Frequency Cepstral Coefficients (MFCC), and other descriptive features such as Zero-Cross-Rates (ZCR), Entropy, Centroid, etc., from the cough signals, before developing classification models. However, with current trends in audio signal classification gearing towards deep learning, which typically make use of only the spectral features, investigating the potential of MFCCs alone in classifying respiratory diseases becomes quite imperative. MFCCs alone, are in fact theoretically quite powerful in providing all vital information about any audio signal, and therefore using them as the standalone set of features in classifying the respiratory diseases is worth investigating. Furthermore, the classification of respiratory diseases so far has only been limited to no more than two diseases. Hence, in order to make a break in this area, this paper investigates the potential of MFCC features alone in classifying respiratory diseases. This is done through the development of a new classification model that features deep learning model design. This method of investigation is similar to typical feature importance studies that fit models before identifying the contributing features. In this case, however, the features are already filtered, and so the model is optimized only by design to perform the study. Furthermore, in order to substantiate the results of the investigation, the model is made to classify more than just two respiratory diseases. For this we have selected five common respiratory diseases namely Asthma, COPD, ILD, Bronchitis and Pneumonia for the classification. Results show that the MFCC features alone do have the potential of classifying the respiratory diseases. This has been substantiated by achieving training accuracies on the model to fall between 85.86 to 97.83% and test accuracies between 87.02 to 88.50%.
{"title":"Investigating the potential of MFCC features in classifying respiratory diseases","authors":"A. Sreeram, Udhaya S. Ravishankar, Narayana Rao Sripada, Baswaraj Mamidgi","doi":"10.1109/IOTSMS52051.2020.9340166","DOIUrl":"https://doi.org/10.1109/IOTSMS52051.2020.9340166","url":null,"abstract":"In the literature so far, classification of respiratory diseases with cough signals has typically involved extracting standard spectral features such as Mel Frequency Cepstral Coefficients (MFCC), and other descriptive features such as Zero-Cross-Rates (ZCR), Entropy, Centroid, etc., from the cough signals, before developing classification models. However, with current trends in audio signal classification gearing towards deep learning, which typically make use of only the spectral features, investigating the potential of MFCCs alone in classifying respiratory diseases becomes quite imperative. MFCCs alone, are in fact theoretically quite powerful in providing all vital information about any audio signal, and therefore using them as the standalone set of features in classifying the respiratory diseases is worth investigating. Furthermore, the classification of respiratory diseases so far has only been limited to no more than two diseases. Hence, in order to make a break in this area, this paper investigates the potential of MFCC features alone in classifying respiratory diseases. This is done through the development of a new classification model that features deep learning model design. This method of investigation is similar to typical feature importance studies that fit models before identifying the contributing features. In this case, however, the features are already filtered, and so the model is optimized only by design to perform the study. Furthermore, in order to substantiate the results of the investigation, the model is made to classify more than just two respiratory diseases. For this we have selected five common respiratory diseases namely Asthma, COPD, ILD, Bronchitis and Pneumonia for the classification. Results show that the MFCC features alone do have the potential of classifying the respiratory diseases. This has been substantiated by achieving training accuracies on the model to fall between 85.86 to 97.83% and test accuracies between 87.02 to 88.50%.","PeriodicalId":147136,"journal":{"name":"2020 7th International Conference on Internet of Things: Systems, Management and Security (IOTSMS)","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127501042","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-12-14DOI: 10.1109/IOTSMS52051.2020.9340192
G. Khalimov, Y. Kotukh, Svitlana Khalimova
Internet of things (IoT) is a growing technology with a big market and impact to our lives. It can ease various different tasks for us. Meanwhile, IoT has many serious security threats, like data breaches, side-channel attacks, and virus and data authentication. Our present classical cryptography, like the Rivest-Shamir-Adleman (RSA) algorithm, work well under the classical computers. However, the technology is slowly shifting towards quantum computing, which has immense processing power and is more than enough to break the current cryptographic primitives in affordable time. So, it is required to design quantum cryptographic algorithms to prevent our systems from security breaches even before quantum computers will be available for commercial purposes on the market. In this paper, we describe a new implementation of MST3 cryptosystems based on the group of automorphisms of the field of the Pu function. The main difference of the presented implementation is the extension of the logarithmic signature and, as a consequence, the presence of multi-stage recovery of message parts from the ciphertext. The presented implementation of the cryptosystem is more reliable. The cryptanalysis complexity and message size for encryption are square times larger than the MST3 cryptosystem in the Suzuki group. This approach shows advantages and it is a quantum safe for the IoT use.
{"title":"Encryption scheme based on the automorphism group of the Ree function field","authors":"G. Khalimov, Y. Kotukh, Svitlana Khalimova","doi":"10.1109/IOTSMS52051.2020.9340192","DOIUrl":"https://doi.org/10.1109/IOTSMS52051.2020.9340192","url":null,"abstract":"Internet of things (IoT) is a growing technology with a big market and impact to our lives. It can ease various different tasks for us. Meanwhile, IoT has many serious security threats, like data breaches, side-channel attacks, and virus and data authentication. Our present classical cryptography, like the Rivest-Shamir-Adleman (RSA) algorithm, work well under the classical computers. However, the technology is slowly shifting towards quantum computing, which has immense processing power and is more than enough to break the current cryptographic primitives in affordable time. So, it is required to design quantum cryptographic algorithms to prevent our systems from security breaches even before quantum computers will be available for commercial purposes on the market. In this paper, we describe a new implementation of MST3 cryptosystems based on the group of automorphisms of the field of the Pu function. The main difference of the presented implementation is the extension of the logarithmic signature and, as a consequence, the presence of multi-stage recovery of message parts from the ciphertext. The presented implementation of the cryptosystem is more reliable. The cryptanalysis complexity and message size for encryption are square times larger than the MST3 cryptosystem in the Suzuki group. This approach shows advantages and it is a quantum safe for the IoT use.","PeriodicalId":147136,"journal":{"name":"2020 7th International Conference on Internet of Things: Systems, Management and Security (IOTSMS)","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123487694","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-12-14DOI: 10.1109/IOTSMS52051.2020.9340204
Erin Beckwith, Geethapriya Thamilarasu
Traditional security solutions that rely on public key infrastructure present scalability and transparency challenges when deployed in Internet of Things (IoT). In this paper, we develop a blockchain based authentication mechanism for IoT that can be integrated into the traditional transport layer security protocols such as Transport Layer Security (TLS) and Datagram Transport Layer Security (DTLS). Our proposed mechanism is an alternative to the traditional Certificate Authority (CA)-based Public Key Infrastructure (PKI) that relies on x.509 certificates. Specifically, the proposed solution enables the modified TLS/DTLS a viable option for resource constrained IoT devices where minimizing memory utilization is critical. Experiments show that blockchain based authentication can reduce dynamic memory usage by up to 20%, while only minimally increasing application image size and time of execution of the TLS/DTLS handshake.
{"title":"BA-TLS: Blockchain Authentication for Transport Layer Security in Internet of Things","authors":"Erin Beckwith, Geethapriya Thamilarasu","doi":"10.1109/IOTSMS52051.2020.9340204","DOIUrl":"https://doi.org/10.1109/IOTSMS52051.2020.9340204","url":null,"abstract":"Traditional security solutions that rely on public key infrastructure present scalability and transparency challenges when deployed in Internet of Things (IoT). In this paper, we develop a blockchain based authentication mechanism for IoT that can be integrated into the traditional transport layer security protocols such as Transport Layer Security (TLS) and Datagram Transport Layer Security (DTLS). Our proposed mechanism is an alternative to the traditional Certificate Authority (CA)-based Public Key Infrastructure (PKI) that relies on x.509 certificates. Specifically, the proposed solution enables the modified TLS/DTLS a viable option for resource constrained IoT devices where minimizing memory utilization is critical. Experiments show that blockchain based authentication can reduce dynamic memory usage by up to 20%, while only minimally increasing application image size and time of execution of the TLS/DTLS handshake.","PeriodicalId":147136,"journal":{"name":"2020 7th International Conference on Internet of Things: Systems, Management and Security (IOTSMS)","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116203716","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-12-14DOI: 10.1109/IOTSMS52051.2020.9340163
Marco Pomalo, V. T. Le, Nabil El Ioini, C. Pahl, H. Barzegar
The rapid development of telecommunications and cellular network technologies gave birth to a range of services and scenarios that were considered impossible a decade ago. Various architectures, scenarios, and use-cases can be deployed on top of the different generations of cellular networks to solve different business cases. Some scenarios require a high level of reliability due to their critical usage e.g., Vehicular Edge computing, medical IoT and so on. When offering services at the edge of the network, the information exchanged needs to be current and valid for systematic performance assessment and modeling. However, in order to run experiments, access to valid and reliable telecommunication data e.g., eNodeB (Base Station) properties, and configurations is not easy, since in most cases data is either confidential or at least difficult to obtain, especially when dealing with cross organizational boundaries (e.g., data coming from multiple telecom operators). To bridge this gap and allow researchers to build, test and analyze new protocols and algorithms with telecommunication data, we designed a mobile data generator (DG) for multi-domain cellular networks. Our generator provides a range of possible configurations and handles scenarios that include multiple participants, authorities and organizations. In this paper, we present the design and implementation of our generator. We evaluated the data generator by considering different scenarios, specifically, we have tested service interruptions and mobile network migration since these scenarios require a considerable amount of data.
{"title":"A Data Generator for Cloud-Edge Vehicle Communication in Multi Domain Cellular Networks","authors":"Marco Pomalo, V. T. Le, Nabil El Ioini, C. Pahl, H. Barzegar","doi":"10.1109/IOTSMS52051.2020.9340163","DOIUrl":"https://doi.org/10.1109/IOTSMS52051.2020.9340163","url":null,"abstract":"The rapid development of telecommunications and cellular network technologies gave birth to a range of services and scenarios that were considered impossible a decade ago. Various architectures, scenarios, and use-cases can be deployed on top of the different generations of cellular networks to solve different business cases. Some scenarios require a high level of reliability due to their critical usage e.g., Vehicular Edge computing, medical IoT and so on. When offering services at the edge of the network, the information exchanged needs to be current and valid for systematic performance assessment and modeling. However, in order to run experiments, access to valid and reliable telecommunication data e.g., eNodeB (Base Station) properties, and configurations is not easy, since in most cases data is either confidential or at least difficult to obtain, especially when dealing with cross organizational boundaries (e.g., data coming from multiple telecom operators). To bridge this gap and allow researchers to build, test and analyze new protocols and algorithms with telecommunication data, we designed a mobile data generator (DG) for multi-domain cellular networks. Our generator provides a range of possible configurations and handles scenarios that include multiple participants, authorities and organizations. In this paper, we present the design and implementation of our generator. We evaluated the data generator by considering different scenarios, specifically, we have tested service interruptions and mobile network migration since these scenarios require a considerable amount of data.","PeriodicalId":147136,"journal":{"name":"2020 7th International Conference on Internet of Things: Systems, Management and Security (IOTSMS)","volume":"363 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115965614","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-12-14DOI: 10.1109/iotsms52051.2020.9340191
{"title":"IOTSMS 2020 Cover Page","authors":"","doi":"10.1109/iotsms52051.2020.9340191","DOIUrl":"https://doi.org/10.1109/iotsms52051.2020.9340191","url":null,"abstract":"","PeriodicalId":147136,"journal":{"name":"2020 7th International Conference on Internet of Things: Systems, Management and Security (IOTSMS)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131148592","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-12-14DOI: 10.1109/IOTSMS52051.2020.9340222
E. Elbasi
Nowadays, multimedia elements such as image, video, animation, audio, and software can be distributed in a short time anywhere in the world using internet technology. Content owners are concerned about copyright protection and authentication. Watermarking is one of the method to protect patient information in medical images such as magnetic resonance and ultrasound imaging. In the literature, there are several methods proposed using both frequency-domain (etc. digital cosine transforms (DCT), digital wavelet transforms (DWT), digital Fourier transforms (DFT), digital radon transform (DRT)) and spatial domain (least significant bits (LSB)) methods. Mostly digital wavelet transform based watermarking methods give very promising results after common attacks. In the DWT algorithm, scaling factor is used in embedding and extraction. A scaling factor is a number between 0 and 1 which has been determined by the user which is not efficient. We can use the brute force method which solves a problem by checking all the possible cases, but it is slow. In this work, we use simulated annealing heuristic search algorithm to find out the best scaling factor to reach more robust, transparent, high data capacity and resistant watermarking method in medical images. Experimental results show that simulated annealing-based scaling factor determination in frequency domain watermarking gives more robust, resistant, and transparent watermarked images.
{"title":"Enhancing the robustness of watermarked medical images using heuristic search algorithm","authors":"E. Elbasi","doi":"10.1109/IOTSMS52051.2020.9340222","DOIUrl":"https://doi.org/10.1109/IOTSMS52051.2020.9340222","url":null,"abstract":"Nowadays, multimedia elements such as image, video, animation, audio, and software can be distributed in a short time anywhere in the world using internet technology. Content owners are concerned about copyright protection and authentication. Watermarking is one of the method to protect patient information in medical images such as magnetic resonance and ultrasound imaging. In the literature, there are several methods proposed using both frequency-domain (etc. digital cosine transforms (DCT), digital wavelet transforms (DWT), digital Fourier transforms (DFT), digital radon transform (DRT)) and spatial domain (least significant bits (LSB)) methods. Mostly digital wavelet transform based watermarking methods give very promising results after common attacks. In the DWT algorithm, scaling factor is used in embedding and extraction. A scaling factor is a number between 0 and 1 which has been determined by the user which is not efficient. We can use the brute force method which solves a problem by checking all the possible cases, but it is slow. In this work, we use simulated annealing heuristic search algorithm to find out the best scaling factor to reach more robust, transparent, high data capacity and resistant watermarking method in medical images. Experimental results show that simulated annealing-based scaling factor determination in frequency domain watermarking gives more robust, resistant, and transparent watermarked images.","PeriodicalId":147136,"journal":{"name":"2020 7th International Conference on Internet of Things: Systems, Management and Security (IOTSMS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132190532","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-12-14DOI: 10.1109/IOTSMS52051.2020.9340162
E. Elbasi
Surveillance systems are widely used in airports, streets, banks, military areas, borders, hospitals, and schools. There are two types of surveillance systems which are real-time systems and offline surveillance systems. Usually, security people track videos on time in monitoring rooms to find out abnormal human activities. Real-time human tracking from videos is very expensive especially in airports, borders, and streets due to the huge number of surveillance cameras. There are a lot of research works have been done for automated surveillance systems. In this paper, we presented a new surveillance system to recognize human activities from several cameras using machine learning algorithms. Sequences of images are collected from cameras using the internet of things technology from indoor or outdoor areas. A feature vector is created for each recognized moving object, then machine learning algorithms are applied to extract moving object activities. The proposed abnormal event detection system gives very promising results which are more than 96% accuracy in Multilayer Perceptron, Iterative Classifier Optimizer, and Random Forest algorithms.
{"title":"Reliable abnormal event detection from IoT surveillance systems","authors":"E. Elbasi","doi":"10.1109/IOTSMS52051.2020.9340162","DOIUrl":"https://doi.org/10.1109/IOTSMS52051.2020.9340162","url":null,"abstract":"Surveillance systems are widely used in airports, streets, banks, military areas, borders, hospitals, and schools. There are two types of surveillance systems which are real-time systems and offline surveillance systems. Usually, security people track videos on time in monitoring rooms to find out abnormal human activities. Real-time human tracking from videos is very expensive especially in airports, borders, and streets due to the huge number of surveillance cameras. There are a lot of research works have been done for automated surveillance systems. In this paper, we presented a new surveillance system to recognize human activities from several cameras using machine learning algorithms. Sequences of images are collected from cameras using the internet of things technology from indoor or outdoor areas. A feature vector is created for each recognized moving object, then machine learning algorithms are applied to extract moving object activities. The proposed abnormal event detection system gives very promising results which are more than 96% accuracy in Multilayer Perceptron, Iterative Classifier Optimizer, and Random Forest algorithms.","PeriodicalId":147136,"journal":{"name":"2020 7th International Conference on Internet of Things: Systems, Management and Security (IOTSMS)","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124435137","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-12-14DOI: 10.1109/IOTSMS52051.2020.9340194
Javad Sohankar, Madhurima Pore, Ayan Banerjee, Koosha Sadeghi, S. Gupta
Physiological signal based interactive systems communicate with human users in real time manner. However, the large size of data generated by sensors, complex computations necessary for processing physiological signals (e.g. machine learning algorithms) hamper the real-time performance of such systems. The main challenges to overcome these issues are limited computational capability of mobile platform and also the latency of offloading computation to servers. A solution is to use predictive models to access future data in order to improve the response time of the system. However, these predictive models have complex computation which result in high execution times on mobile phone that interferes with real time performance. With the advent of OpenCL enabled GPUs in mobile platform, there is a potential of developing general purpose applications (e.g. predictive models) which offload complex computation to GPUs. Although the use of GPUs will reduce the computation time in physiological signal based mobile systems, satisfying the time constraints of these systems can be challenging. That is due to the dynamically changing nature of physiological data which requires frequent updating of physiological models in the system. In this work, computations of a predictive model for brain signals is offloaded to mobile phone GPU. The evaluation of the performance shows that GPU can outperform CPU in mobile platform for general purpose computing.
{"title":"Machine Learning Based Predictive Models in Mobile Platforms Using CPU-GPU","authors":"Javad Sohankar, Madhurima Pore, Ayan Banerjee, Koosha Sadeghi, S. Gupta","doi":"10.1109/IOTSMS52051.2020.9340194","DOIUrl":"https://doi.org/10.1109/IOTSMS52051.2020.9340194","url":null,"abstract":"Physiological signal based interactive systems communicate with human users in real time manner. However, the large size of data generated by sensors, complex computations necessary for processing physiological signals (e.g. machine learning algorithms) hamper the real-time performance of such systems. The main challenges to overcome these issues are limited computational capability of mobile platform and also the latency of offloading computation to servers. A solution is to use predictive models to access future data in order to improve the response time of the system. However, these predictive models have complex computation which result in high execution times on mobile phone that interferes with real time performance. With the advent of OpenCL enabled GPUs in mobile platform, there is a potential of developing general purpose applications (e.g. predictive models) which offload complex computation to GPUs. Although the use of GPUs will reduce the computation time in physiological signal based mobile systems, satisfying the time constraints of these systems can be challenging. That is due to the dynamically changing nature of physiological data which requires frequent updating of physiological models in the system. In this work, computations of a predictive model for brain signals is offloaded to mobile phone GPU. The evaluation of the performance shows that GPU can outperform CPU in mobile platform for general purpose computing.","PeriodicalId":147136,"journal":{"name":"2020 7th International Conference on Internet of Things: Systems, Management and Security (IOTSMS)","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-12-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130056543","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：