Caoxiao Li, Shuyin Xia, Jingcheng Fu, Zizhong Chen, Binggui Wang
The traditional genetic algorithm has the disadvantage of slow convergence speed and prematurity. In order to optimize the algorithm from the perspective of spatial analysis, a multi-granular genetic algorithm proposes a spatial partitioning method based on a completely random tree to improve the genetic algorithm. However, the accurate analysis of space by completely random trees is time-consuming. Therefore, an improved genetic algorithm based on k-mean is proposed in this paper. The individuals obtained by the genetic algorithm are clustered through k-means. Then, according to the clustering results, new individuals are generated in the subspace containing a small number of individuals and in the subspace to which the current optimal solution belongs, thus improving the performance of the genetic algorithm.
{"title":"An Improved Genetic Algorithm Based on k-means","authors":"Caoxiao Li, Shuyin Xia, Jingcheng Fu, Zizhong Chen, Binggui Wang","doi":"10.1145/3459104.3459164","DOIUrl":"https://doi.org/10.1145/3459104.3459164","url":null,"abstract":"The traditional genetic algorithm has the disadvantage of slow convergence speed and prematurity. In order to optimize the algorithm from the perspective of spatial analysis, a multi-granular genetic algorithm proposes a spatial partitioning method based on a completely random tree to improve the genetic algorithm. However, the accurate analysis of space by completely random trees is time-consuming. Therefore, an improved genetic algorithm based on k-mean is proposed in this paper. The individuals obtained by the genetic algorithm are clustered through k-means. Then, according to the clustering results, new individuals are generated in the subspace containing a small number of individuals and in the subspace to which the current optimal solution belongs, thus improving the performance of the genetic algorithm.","PeriodicalId":142284,"journal":{"name":"2021 International Symposium on Electrical, Electronics and Information Engineering","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133634742","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}
Hasan Akalp, Enes Furkan Cigdem, Seyma Yilmaz, Necva Bölücü, Burcu Can
There are various genres of music available in every period and field of human life. Every music genre represents a set of shared conventions. Today people have the opportunity to listen to any genre of music they want using various music platforms. However, with the increasing number of music genres, the management of these platforms becomes difficult. Language representation models such as BERT, DistilBERT have been proven to be useful in learning universal language representations. Such language representation models have achieved amazing results in many language understanding tasks. In this study, we apply language representation models for music genre classification using song lyrics. We examine whether language representation models are better than traditional deep learning models for music genre classification by comparing results and computation times. Experimental results show that BERT outperforms other models on one-label and multi-label classification with accuracy of 77.63% and 71.29% respectively. On the other hand, considering the time taken for one epoch, BERT runs 4 times faster than DistilBERT.
{"title":"Language Representation Models for Music Genre Classification Using Lyrics","authors":"Hasan Akalp, Enes Furkan Cigdem, Seyma Yilmaz, Necva Bölücü, Burcu Can","doi":"10.1145/3459104.3459171","DOIUrl":"https://doi.org/10.1145/3459104.3459171","url":null,"abstract":"There are various genres of music available in every period and field of human life. Every music genre represents a set of shared conventions. Today people have the opportunity to listen to any genre of music they want using various music platforms. However, with the increasing number of music genres, the management of these platforms becomes difficult. Language representation models such as BERT, DistilBERT have been proven to be useful in learning universal language representations. Such language representation models have achieved amazing results in many language understanding tasks. In this study, we apply language representation models for music genre classification using song lyrics. We examine whether language representation models are better than traditional deep learning models for music genre classification by comparing results and computation times. Experimental results show that BERT outperforms other models on one-label and multi-label classification with accuracy of 77.63% and 71.29% respectively. On the other hand, considering the time taken for one epoch, BERT runs 4 times faster than DistilBERT.","PeriodicalId":142284,"journal":{"name":"2021 International Symposium on Electrical, Electronics and Information Engineering","volume":"97 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117194307","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}
Vortex dynamics of wakes generated by two-dimensional rectangular pitching flat plates in free stream are examined with direct numerical simulation using Lagrangian vortex method. The developed method simulates external flow around complex geometry by tracking local velocities and vorticities of particles, introduced within the fluid domain. The viscous effect is modeled using a core spreading method coupled with the splitting and merging spatial adaptation scheme. The particle's velocity is calculated using Biot-Savart formulation. To accelerate computation, Fast Multipole Method (FMM) is employed. The solver is validated by performing an impulsively started cylinder at Reynolds number 550. The results of the computation have reasonable agreement with references listed in literature. For the vortex dynamics of pitching flat plate, the detaching LEV creates a remarkable peak in the lift force before the end of motion for the different pitching cases. For the low Reynolds number, force generated by the pitching flat plate is fairly independent of Reynolds numbers. The current studies also observed that TEV produced at higher Reynolds number has a stronger suction than that at smaller Reynolds numbers.
{"title":"Unsteady Vortex Dynamics of Two-Dimensional Pitching Flat Plate Using Lagrangian Vortex Method","authors":"Dung Viet Duong, L. Zuhal, H. Muhammad","doi":"10.1145/3459104.3459106","DOIUrl":"https://doi.org/10.1145/3459104.3459106","url":null,"abstract":"Vortex dynamics of wakes generated by two-dimensional rectangular pitching flat plates in free stream are examined with direct numerical simulation using Lagrangian vortex method. The developed method simulates external flow around complex geometry by tracking local velocities and vorticities of particles, introduced within the fluid domain. The viscous effect is modeled using a core spreading method coupled with the splitting and merging spatial adaptation scheme. The particle's velocity is calculated using Biot-Savart formulation. To accelerate computation, Fast Multipole Method (FMM) is employed. The solver is validated by performing an impulsively started cylinder at Reynolds number 550. The results of the computation have reasonable agreement with references listed in literature. For the vortex dynamics of pitching flat plate, the detaching LEV creates a remarkable peak in the lift force before the end of motion for the different pitching cases. For the low Reynolds number, force generated by the pitching flat plate is fairly independent of Reynolds numbers. The current studies also observed that TEV produced at higher Reynolds number has a stronger suction than that at smaller Reynolds numbers.","PeriodicalId":142284,"journal":{"name":"2021 International Symposium on Electrical, Electronics and Information Engineering","volume":"67 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115468164","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}
Efficient and accurate traffic prediction is the premise of the development of autonomous driving technology. In-depth research is made on the issue of short-term traffic speed prediction in autonomous driving systems. In view of the time-varying characteristics of the traffic main sentence, this paper designs and implements a traffic prediction system based on genetically improved wavelet neural networks. Through the training and learning of the historical average speed data of roads, it realizes the prediction of future road traffic conditions and helps the planning of travel routes. This algorithm circumvents the shortcomings of wavelet neural networks that easily fall into local minimums, and proposes to optimize the initial coefficients of wavelet neural networks by using the characteristics of global search of genetic algorithms to construct better neural networks. We have verified that the traffic speed prediction based on genetically improved wavelet neural network has a high degree of agreement with real data, and the effect is significantly better than the results of ordinary wavelet neural network, which has higher practical value.
{"title":"Short-term Traffic Prediction Based on Genetic Improved Wavelet Neural Network","authors":"Tianzi Ma, Hao Chen","doi":"10.1145/3459104.3459183","DOIUrl":"https://doi.org/10.1145/3459104.3459183","url":null,"abstract":"Efficient and accurate traffic prediction is the premise of the development of autonomous driving technology. In-depth research is made on the issue of short-term traffic speed prediction in autonomous driving systems. In view of the time-varying characteristics of the traffic main sentence, this paper designs and implements a traffic prediction system based on genetically improved wavelet neural networks. Through the training and learning of the historical average speed data of roads, it realizes the prediction of future road traffic conditions and helps the planning of travel routes. This algorithm circumvents the shortcomings of wavelet neural networks that easily fall into local minimums, and proposes to optimize the initial coefficients of wavelet neural networks by using the characteristics of global search of genetic algorithms to construct better neural networks. We have verified that the traffic speed prediction based on genetically improved wavelet neural network has a high degree of agreement with real data, and the effect is significantly better than the results of ordinary wavelet neural network, which has higher practical value.","PeriodicalId":142284,"journal":{"name":"2021 International Symposium on Electrical, Electronics and Information Engineering","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127713039","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}
In this paper, the problems, such as engineering model, semantic description, basic algorithm, heuristic resolving and platform of semantic engineering are discussed. Taking the humanoid resolving of mathematical application problem as an example, the semantic instances’ aggregation, clause as semantic unit, formula variable's labels matching and space search methods are proposed; the scene's semantic description frame is used to describe context semantics, and the common data pool which includes the word segmentation chain, scene frame database, formula knowledge database and resolving rule database are designed; The basic algorithms such as scene frame matching, correspondence between data elements and formula variables, formula variable constraint algorithm and resolving operation mechanism are established; The common semantic description and selection of problems are realized, and the accelerated operation mechanism of heuristic resolving is also developed; The semantic data platform is built. Finally, the paper summarizes the general semantic ideas of humanoid resolving the primary mathematical application problems, and puts forward the next improvement plan.
{"title":"Several Problems of Semantic Engineering A Case Study of Humanoid Resolving the Primary Mathematics Application Problems","authors":"Ping Zhu","doi":"10.1145/3459104.3459113","DOIUrl":"https://doi.org/10.1145/3459104.3459113","url":null,"abstract":"In this paper, the problems, such as engineering model, semantic description, basic algorithm, heuristic resolving and platform of semantic engineering are discussed. Taking the humanoid resolving of mathematical application problem as an example, the semantic instances’ aggregation, clause as semantic unit, formula variable's labels matching and space search methods are proposed; the scene's semantic description frame is used to describe context semantics, and the common data pool which includes the word segmentation chain, scene frame database, formula knowledge database and resolving rule database are designed; The basic algorithms such as scene frame matching, correspondence between data elements and formula variables, formula variable constraint algorithm and resolving operation mechanism are established; The common semantic description and selection of problems are realized, and the accelerated operation mechanism of heuristic resolving is also developed; The semantic data platform is built. Finally, the paper summarizes the general semantic ideas of humanoid resolving the primary mathematical application problems, and puts forward the next improvement plan.","PeriodicalId":142284,"journal":{"name":"2021 International Symposium on Electrical, Electronics and Information Engineering","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129041929","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}
Hand gestures are a symbolic and non-vocal language and are used by an individual to communicate. With computer vision, hand gestures can be detected and be used to talk with a capable computer, leading to the field of Human-Computer interconnection. The field of computer vision has been achieving cutting edge results with the advent of deep learning models. The work implements the Inception v3 architecture [1], which is a convolutional neural network. The model is retrained on our data set using Transfer learning, with which we reduce the requirements on computational resources, data and time. In this project, a hand gesture is performed in front of a web camera of a system. The gestures are predicted as one among six gestures with a corresponding probability. This project deals with the applications of the detected hand gestures in home automation and control of a robotic arm. Hand gestures are simple to perform, and it makes managing home effortless compared to manually intervening and providing instructions to a machine. In the home automation model, the gesture classification results from the system are transmitted to the microcontroller which switches on or off a home device. The robotic arm is a mechanical system which is used in manipulating the movement of lifting, moving, and placing the workpiece to lighten the work of man. It is equipped with servo motors and is controlled by our hand gestures to perform lifting and dropping of objects and rotation of the robotic arm.
{"title":"Real-time Hand Gesture Recognition for Robotic Arm and Home Automation","authors":"A. Varshini, G. Bhavani, Vithya, R. Thilagavathy","doi":"10.1145/3459104.3459142","DOIUrl":"https://doi.org/10.1145/3459104.3459142","url":null,"abstract":"Hand gestures are a symbolic and non-vocal language and are used by an individual to communicate. With computer vision, hand gestures can be detected and be used to talk with a capable computer, leading to the field of Human-Computer interconnection. The field of computer vision has been achieving cutting edge results with the advent of deep learning models. The work implements the Inception v3 architecture [1], which is a convolutional neural network. The model is retrained on our data set using Transfer learning, with which we reduce the requirements on computational resources, data and time. In this project, a hand gesture is performed in front of a web camera of a system. The gestures are predicted as one among six gestures with a corresponding probability. This project deals with the applications of the detected hand gestures in home automation and control of a robotic arm. Hand gestures are simple to perform, and it makes managing home effortless compared to manually intervening and providing instructions to a machine. In the home automation model, the gesture classification results from the system are transmitted to the microcontroller which switches on or off a home device. The robotic arm is a mechanical system which is used in manipulating the movement of lifting, moving, and placing the workpiece to lighten the work of man. It is equipped with servo motors and is controlled by our hand gestures to perform lifting and dropping of objects and rotation of the robotic arm.","PeriodicalId":142284,"journal":{"name":"2021 International Symposium on Electrical, Electronics and Information Engineering","volume":"120 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133599438","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}
Distributed privacy-preserving data mining (DPPDM) has been attracting enormous attention. It allows multiple participants to jointly use their datasets as a whole to train a model while preserving data privacy. Many works have been looking into the semi-supervised learning in DPPDM, to combine both labeled and unlabeled data for better performance. However, these works only provide transductive solutions, which means they can only give predictions for instances in the training set, and not for any new data sample beyond the set. Meanwhile, these methods are constructed with approximate calculations for security concerns, leading to sub-optimal results and limited effectiveness. In this paper, a mixture-model-based solution is proposed for inductive and effective semi-supervised learning in DPPDM. Our motivation lies in combining mixture models and graph-based methods to construct an anchor mixture with the ability of label prediction. We also propose an optimization process, which is accurately calculated through secure computation protocols, to achieve effectiveness. Experiments on synthetic and real-world datasets demonstrate that our proposal outperforms state-of-the-art methods in both transductive and inductive tasks.
{"title":"Inductive and Effective Privacy-preserving Semi-supervised Learning with Harmonic Anchor Mixture","authors":"Zhi Li, Zhoujun Li","doi":"10.1145/3459104.3459187","DOIUrl":"https://doi.org/10.1145/3459104.3459187","url":null,"abstract":"Distributed privacy-preserving data mining (DPPDM) has been attracting enormous attention. It allows multiple participants to jointly use their datasets as a whole to train a model while preserving data privacy. Many works have been looking into the semi-supervised learning in DPPDM, to combine both labeled and unlabeled data for better performance. However, these works only provide transductive solutions, which means they can only give predictions for instances in the training set, and not for any new data sample beyond the set. Meanwhile, these methods are constructed with approximate calculations for security concerns, leading to sub-optimal results and limited effectiveness. In this paper, a mixture-model-based solution is proposed for inductive and effective semi-supervised learning in DPPDM. Our motivation lies in combining mixture models and graph-based methods to construct an anchor mixture with the ability of label prediction. We also propose an optimization process, which is accurately calculated through secure computation protocols, to achieve effectiveness. Experiments on synthetic and real-world datasets demonstrate that our proposal outperforms state-of-the-art methods in both transductive and inductive tasks.","PeriodicalId":142284,"journal":{"name":"2021 International Symposium on Electrical, Electronics and Information Engineering","volume":"124 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128116650","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}
This paper introduce the flight test method of civil airplane taking-off and landing in strong crosswind condition. Observe and analyse the relationships between pilot's inputs and airplane behaviour. Thus, establish control objectives and precautionary measurements in strong crosswind flight test. In strong crosswind flight test, while airplane is accelerating during take-off, pilots’ objectives are to align the airplane to the center-line of the runway and balance the load of the main wheels with increasing velocity. While approaching and landing, excessive sideslip angle or roll angle should be avoided to prevent airplane damage upon touchdown, during deceleration after touchdown, thrust-reverser may be activated when the airplane is steady aligned with the runway. On this basis, the hazards of losing control, drift off the runway and powerplant failure may be avoided. From the hazards identified, we may deduce that the hazard level of the flight test is high in nature, however, preparations and pre-planning in flight test methods, flight training in advance, and even finding the appropriate test environment may significantly reduce the hazard level.
{"title":"Research on Flight Technique and Hazard Control for Civil Airplane Crosswind Flight Test","authors":"Yunpeng Wu, Yang Liu","doi":"10.1145/3459104.3459109","DOIUrl":"https://doi.org/10.1145/3459104.3459109","url":null,"abstract":"This paper introduce the flight test method of civil airplane taking-off and landing in strong crosswind condition. Observe and analyse the relationships between pilot's inputs and airplane behaviour. Thus, establish control objectives and precautionary measurements in strong crosswind flight test. In strong crosswind flight test, while airplane is accelerating during take-off, pilots’ objectives are to align the airplane to the center-line of the runway and balance the load of the main wheels with increasing velocity. While approaching and landing, excessive sideslip angle or roll angle should be avoided to prevent airplane damage upon touchdown, during deceleration after touchdown, thrust-reverser may be activated when the airplane is steady aligned with the runway. On this basis, the hazards of losing control, drift off the runway and powerplant failure may be avoided. From the hazards identified, we may deduce that the hazard level of the flight test is high in nature, however, preparations and pre-planning in flight test methods, flight training in advance, and even finding the appropriate test environment may significantly reduce the hazard level.","PeriodicalId":142284,"journal":{"name":"2021 International Symposium on Electrical, Electronics and Information Engineering","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132139233","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}
Roman Jakubícek, Tomáš Vičar, Jiří Chmelík, P. Ourednicek, J. Jan
In this paper, we present a method, based on deep learning, for prediction of non-contrast CT image from a single contrast image. For training of this image-to-image translation task, virtual contrast and virtual non-contrast (VNC) images were created from spectral CT data by Philips IntelliSpace Portal (ISP) software. Virtual version of conventional CT (cCT) images and VNC images allows to train paired supervised image-to-image translation models. Two different schemes were tested to train the Convolutional Neural Network (CNN) with U-Net architecture, using standard training with L1/L2 loss as well as training via conditional Generative Adversarial Network (cGAN) with an additional Wasserstein modification (WcGAN). Qualitatively, the proposed method achieves similar results to the original VNC images. However, quantitatively, the trained CNN provides a slightly smaller density reduction in some tissues. Non-contrast image can be predicted from a single conventional CT image, without the need for pre- and post-contrast scan or without a spectral CT scan.
{"title":"Deep-learning Based Prediction of Virtual Non-contrast CT Images","authors":"Roman Jakubícek, Tomáš Vičar, Jiří Chmelík, P. Ourednicek, J. Jan","doi":"10.1145/3459104.3460237","DOIUrl":"https://doi.org/10.1145/3459104.3460237","url":null,"abstract":"In this paper, we present a method, based on deep learning, for prediction of non-contrast CT image from a single contrast image. For training of this image-to-image translation task, virtual contrast and virtual non-contrast (VNC) images were created from spectral CT data by Philips IntelliSpace Portal (ISP) software. Virtual version of conventional CT (cCT) images and VNC images allows to train paired supervised image-to-image translation models. Two different schemes were tested to train the Convolutional Neural Network (CNN) with U-Net architecture, using standard training with L1/L2 loss as well as training via conditional Generative Adversarial Network (cGAN) with an additional Wasserstein modification (WcGAN). Qualitatively, the proposed method achieves similar results to the original VNC images. However, quantitatively, the trained CNN provides a slightly smaller density reduction in some tissues. Non-contrast image can be predicted from a single conventional CT image, without the need for pre- and post-contrast scan or without a spectral CT scan.","PeriodicalId":142284,"journal":{"name":"2021 International Symposium on Electrical, Electronics and Information Engineering","volume":"208 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134110808","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}
The popularity of wireless mesh network (WMN) systems grows rapidly with its main usefulness to expand the coverage of wireless communication without wired-based infrastructures. Together with software-defined networking, WMN can be programmed and adapted to dynamic wireless environments. Software-defined wireless mesh network (SDWMN) gets therefore increasing attentions in networking research as well as network-centric application communities. In this paper, SDWMN has been designed as the main underlying platform that allows sensor nodes installed on the road to relay their sensed data. Particularly, this research is concerned with the development of vehicular traffic monitoring technique that can sense the presence of vehicles passing by the SDWMN sensor nodes. Since the penetration of vehicles equipped with WIFI devices is significantly increased, a WIFI packet measurement application has been developed for each SDWMN node to detect the service set identifier (SSID) of the wireless communication for monitoring the vehicle traffic. Each traveling vehicle that provides SSID can be sensed by each SDWMN node along with the corresponding time stamp. A functionality has then been developed to map the raw sensor data to obtain the travel time of vehicles. This technique has been developed on a real SDWMN system testbed and its functionality is tested on Phayathai road in Bangkok, Thailand. The obtained experimental results suggest the practicality of this vehicular traffic monitoring technique with up to 5,000 data records obtainable per sensor node per day. With continuously growing number of WIFI-equipped vehicles, it is believed that the proposed technique can be used cost-effectively to provide real-time vehicular traffic conditions in the future without cost burdens from other conventional vehicular traffic sensors requiring highly costed communication infrastructures.
{"title":"Testing of Vehicular Traffic Monitoring Technique by Using WIFI Packet Measurement in Software-defined Wireless Mesh Network","authors":"Meechai Homchan, C. Aswakul","doi":"10.1145/3459104.3459138","DOIUrl":"https://doi.org/10.1145/3459104.3459138","url":null,"abstract":"The popularity of wireless mesh network (WMN) systems grows rapidly with its main usefulness to expand the coverage of wireless communication without wired-based infrastructures. Together with software-defined networking, WMN can be programmed and adapted to dynamic wireless environments. Software-defined wireless mesh network (SDWMN) gets therefore increasing attentions in networking research as well as network-centric application communities. In this paper, SDWMN has been designed as the main underlying platform that allows sensor nodes installed on the road to relay their sensed data. Particularly, this research is concerned with the development of vehicular traffic monitoring technique that can sense the presence of vehicles passing by the SDWMN sensor nodes. Since the penetration of vehicles equipped with WIFI devices is significantly increased, a WIFI packet measurement application has been developed for each SDWMN node to detect the service set identifier (SSID) of the wireless communication for monitoring the vehicle traffic. Each traveling vehicle that provides SSID can be sensed by each SDWMN node along with the corresponding time stamp. A functionality has then been developed to map the raw sensor data to obtain the travel time of vehicles. This technique has been developed on a real SDWMN system testbed and its functionality is tested on Phayathai road in Bangkok, Thailand. The obtained experimental results suggest the practicality of this vehicular traffic monitoring technique with up to 5,000 data records obtainable per sensor node per day. With continuously growing number of WIFI-equipped vehicles, it is believed that the proposed technique can be used cost-effectively to provide real-time vehicular traffic conditions in the future without cost burdens from other conventional vehicular traffic sensors requiring highly costed communication infrastructures.","PeriodicalId":142284,"journal":{"name":"2021 International Symposium on Electrical, Electronics and Information Engineering","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126987391","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：