As the density of wireless networks increases globally, the vulnerability of overlapped dense wireless communications to interference by hidden nodes and denial-of-service (DoS) attacks is becoming more apparent. Ther...As the density of wireless networks increases globally, the vulnerability of overlapped dense wireless communications to interference by hidden nodes and denial-of-service (DoS) attacks is becoming more apparent. There exists a gap in research on the detection and response to attacks on Medium Access Control (MAC) mechanisms themselves, which would lead to service outages between nodes. Classifying exploitation and deceptive jamming attacks on control mechanisms is particularly challengingdue to their resemblance to normal heavy communication patterns. Accordingly, this paper proposes a machine learning-based selective attack mitigation model that detects DoS attacks on wireless networks by monitoring packet log data. Based on the type of detected attack, it implements effective corresponding mitigation techniques to restore performance to nodes whose availability has been compromised. Experimental results reveal that the accuracy of the proposed model is 14% higher than that of a baseline anomaly detection model. Further, the appropriate mitigation techniques selected by the proposed system based on the attack type improve the average throughput by more than 440% compared to the case without a response.展开更多
A Luttinger liquid is a theoretical model describing interacting electrons in one-dimensional(1D)conductors.While individual 1D conductors have shown interesting Luttinger-liquid behaviors such as spin-charge separati...A Luttinger liquid is a theoretical model describing interacting electrons in one-dimensional(1D)conductors.While individual 1D conductors have shown interesting Luttinger-liquid behaviors such as spin-charge separation and power-law spectral density,the more interesting phenomena predicted in coupled Luttinger liquids of neighboring 1D conductors have been rarely observed due to the difficulty in creating such structures.Recently,we have successfully grown close-packed carbon nanotube(CNT)arrays with uniform chirality,providing an ideal material system for studying the coupled Luttinger liquids.Here,we report on the observation of tunable hyperbolic plasmons in the coupled Luttinger liquids of CNT arrays using scanning near-field optical microscopy.These hyperbolic plasmons,resulting from the conductivity anisotropy in the CNT array,exhibit strong spatial confinement,in situ tunability,and a wide spectral range.Despite their hyperbolic wavefronts,the plasmon propagation in the axial direction still adheres to the Luttinger-liquid theory.Our work not only demonstrates a fascinating phenomenon in coupled Luttinger liquids for fundamental physics exploration,but also provides a highly confined and in situ tunable hyperbolic plasmon in close-packed CNT arrays for future nanophotonic devices and circuits.展开更多
The Internet of Things(IoT)is extensively applied across various industrial domains,such as smart homes,factories,and intelligent transportation,becoming integral to daily life.Establishing robust policies for managin...The Internet of Things(IoT)is extensively applied across various industrial domains,such as smart homes,factories,and intelligent transportation,becoming integral to daily life.Establishing robust policies for managing and governing IoT devices is imperative.Secure authentication for IoT devices in resource-constrained environments remains challenging due to the limitations of conventional complex protocols.Prior methodologies enhanced mutual authentication through key exchange protocols or complex operations,which are impractical for lightweight devices.To address this,our study introduces the privacy-preserving software-defined range proof(SDRP)model,which achieves secure authentication with low complexity.SDRP minimizes the overhead of confidentiality and authentication processes by utilizing range proof to verify whether the attribute information of a user falls within a specific range.Since authentication is performed using a digital ID sequence generated from indirect personal data,it can avoid the disclosure of actual individual attributes.Experimental results demonstrate that SDRP significantly improves security efficiency,increasing it by an average of 93.02%compared to conventional methods.It mitigates the trade-off between security and efficiency by reducing leakage risk by an average of 98.7%.展开更多
With the advancement of wireless network technology,vast amounts of traffic have been generated,and malicious traffic attacks that threaten the network environment are becoming increasingly sophisticated.While signatu...With the advancement of wireless network technology,vast amounts of traffic have been generated,and malicious traffic attacks that threaten the network environment are becoming increasingly sophisticated.While signature-based detection methods,static analysis,and dynamic analysis techniques have been previously explored for malicious traffic detection,they have limitations in identifying diversified malware traffic patterns.Recent research has been focused on the application of machine learning to detect these patterns.However,applying machine learning to lightweight devices like IoT devices is challenging because of the high computational demands and complexity involved in the learning process.In this study,we examined methods for effectively utilizing machine learning-based malicious traffic detection approaches for lightweight devices.We introduced the suboptimal feature selection model(SFSM),a feature selection technique designed to reduce complexity while maintaining the effectiveness of malicious traffic detection.Detection performance was evaluated on various malicious traffic,benign,exploits,and generic,using the UNSW-NB15 dataset and SFSM sub-optimized hyperparameters for feature selection and narrowed the search scope to encompass all features.SFSM improved learning performance while minimizing complexity by considering feature selection and exhaustive search as two steps,a problem not considered in conventional models.Our experimental results showed that the detection accuracy was improved by approximately 20%compared to the random model,and the reduction in accuracy compared to the greedy model,which performs an exhaustive search on all features,was kept within 6%.Additionally,latency and complexity were reduced by approximately 96%and 99.78%,respectively,compared to the greedy model.This study demonstrates that malicious traffic can be effectively detected even in lightweight device environments.SFSM verified the possibility of detecting various attack traffic on lightweight devices.展开更多
With the rise of remote work and the digital industry,advanced cyberattacks have become more diverse and complex in terms of attack types and characteristics,rendering them difficult to detect with conventional intrus...With the rise of remote work and the digital industry,advanced cyberattacks have become more diverse and complex in terms of attack types and characteristics,rendering them difficult to detect with conventional intrusion detection methods.Signature-based intrusion detection methods can be used to detect attacks;however,they cannot detect new malware.Endpoint detection and response(EDR)tools are attracting attention as a means of detecting attacks on endpoints in real-time to overcome the limitations of signature-based intrusion detection techniques.However,EDR tools are restricted by the continuous generation of unnecessary logs,resulting in poor detection performance and memory efficiency.Machine learning-based intrusion detection techniques for responding to advanced cyberattacks are memory intensive,using numerous features;they lack optimal feature selection for each attack type.To overcome these limitations,this study proposes a memory-efficient intrusion detection approach incorporating multi-binary classifiers using optimal feature selection.The proposed model detects multiple types of malicious attacks using parallel binary classifiers with optimal features for each attack type.The experimental results showed a 2.95%accuracy improvement and an 88.05%memory reduction using only six features compared to a model with 18 features.Furthermore,compared to a conventional multi-classification model with simple feature selection based on permutation importance,the accuracy improved by 11.67%and the memory usage decreased by 44.87%.The proposed scheme demonstrates that effective intrusion detection is achievable with minimal features,making it suitable for memory-limited mobile and Internet of Things devices.展开更多
As time and space constraints decrease due to the development of wireless communication network technology,the scale and scope of cyber-attacks targeting the Internet of Things(IoT)are increasing.However,it is difficu...As time and space constraints decrease due to the development of wireless communication network technology,the scale and scope of cyber-attacks targeting the Internet of Things(IoT)are increasing.However,it is difficult to apply high-performance security modules to the IoT owing to the limited battery,memory capacity,and data transmission performance depend-ing on the size of the device.Conventional research has mainly reduced power consumption by lightening encryption algorithms.However,it is difficult to defend large-scale information systems and networks against advanced and intelligent attacks because of the problem of deteriorating security perfor-mance.In this study,we propose wake-up security(WuS),a low-power security architecture that can utilize high-performance security algorithms in an IoT environment.By introducing a small logic that performs anomaly detection on the IoT platform and executes the security module only when necessary according to the anomaly detection result,WuS improves security and power efficiency while using a relatively high-complexity security module in a low-power environment compared to the conventional method of periodically exe-cuting a high-performance security module.In this study,a Python simulator based on the UNSW-NB15 dataset is used to evaluate the power consumption,latency,and security of the proposed method.The evaluation results reveal that the power consumption of the proposed WuS mechanism is approxi-mately 51.8%and 27.2%lower than those of conventional high-performance security and lightweight security modules,respectively.Additionally,the laten-cies are approximately 74.8%and 65.9%lower,respectively.Furthermore,the WuS mechanism achieved a high detection accuracy of approximately 96.5%or greater,proving that the detection efficiency performance improved by approximately 33.5%compared to the conventional model.The performance evaluation results for the proposed model varied depending on the applied anomaly-detection model.Therefore,they can be used in various ways by selecting suitable models based on the performance levels required in each industry.展开更多
Detection and classification of the blurred and the non-blurred regions in images is a challenging task due to the limited available information about blur type,scenarios and level of blurriness.In this paper,we propo...Detection and classification of the blurred and the non-blurred regions in images is a challenging task due to the limited available information about blur type,scenarios and level of blurriness.In this paper,we propose an effective method for blur detection and segmentation based on transfer learning concept.The proposed method consists of two separate steps.In the first step,genetic programming(GP)model is developed that quantify the amount of blur for each pixel in the image.The GP model method uses the multiresolution features of the image and it provides an improved blur map.In the second phase,the blur map is segmented into blurred and non-blurred regions by using an adaptive threshold.A model based on support vector machine(SVM)is developed to compute adaptive threshold for the input blur map.The performance of the proposed method is evaluated using two different datasets and compared with various state-of-the-art methods.The comparative analysis reveals that the proposed method performs better against the state-of-the-art techniques.展开更多
Enormousmethods have been proposed for the detection and segmentation of blur and non-blur regions of the images.Due to the limited available information about blur type,scenario and the level of blurriness,detection ...Enormousmethods have been proposed for the detection and segmentation of blur and non-blur regions of the images.Due to the limited available information about blur type,scenario and the level of blurriness,detection and segmentation is a challenging task.Hence,the performance of the blur measure operator is an essential factor and needs improvement to attain perfection.In this paper,we propose an effective blur measure based on local binary pattern(LBP)with adaptive threshold for blur detection.The sharpness metric developed based on LBP used a fixed threshold irrespective of the type and level of blur,that may not be suitable for images with variations in imaging conditions,blur amount and type.Contrarily,the proposed measure uses an adaptive threshold for each input image based on the image and blur properties to generate improved sharpness metric.The adaptive threshold is computed based on the model learned through support vector machine(SVM).The performance of the proposed method is evaluated using two different datasets and is compared with five state-of-the-art methods.Comparative analysis reveals that the proposed method performs significantly better qualitatively and quantitatively against all of the compared methods.展开更多
Recently,with the normalization of non-face-to-face online environments in response to the COVID-19 pandemic,the possibility of cyberattacks through endpoints has increased.Numerous endpoint devices are managed meticu...Recently,with the normalization of non-face-to-face online environments in response to the COVID-19 pandemic,the possibility of cyberattacks through endpoints has increased.Numerous endpoint devices are managed meticulously to prevent cyberattacks and ensure timely responses to potential security threats.In particular,because telecommuting,telemedicine,and teleeducation are implemented in uncontrolled environments,attackers typically target vulnerable endpoints to acquire administrator rights or steal authentication information,and reports of endpoint attacks have been increasing considerably.Advanced persistent threats(APTs)using various novel variant malicious codes are a form of a sophisticated attack.However,conventional commercial antivirus and anti-malware systems that use signature-based attack detectionmethods cannot satisfactorily respond to such attacks.In this paper,we propose a method that expands the detection coverage inAPT attack environments.In this model,an open-source threat detector and log collector are used synergistically to improve threat detection performance.Extending the scope of attack log collection through interworking between highly accessible open-source tools can efficiently increase the detection coverage of tactics and techniques used to deal with APT attacks,as defined by MITRE Adversarial Tactics,Techniques,and Common Knowledge(ATT&CK).We implemented an attack environment using an APT attack scenario emulator called Carbanak and analyzed the detection coverage of Google Rapid Response(GRR),an open-source threat detection tool,and Graylog,an open-source log collector.The proposed method expanded the detection coverage against MITRE ATT&CK by approximately 11%compared with that conventional methods.展开更多
With the introduction of 5G technology,the application of Internet of Things(IoT)devices is expanding to various industrial fields.However,introducing a robust,lightweight,low-cost,and low-power security solution to t...With the introduction of 5G technology,the application of Internet of Things(IoT)devices is expanding to various industrial fields.However,introducing a robust,lightweight,low-cost,and low-power security solution to the IoT environment is challenging.Therefore,this study proposes two methods using a data compression technique to detect malicious traffic efficiently and accurately for a secure IoT environment.The first method,compressed sensing and learning(CSL),compresses an event log in a bitmap format to quickly detect attacks.Then,the attack log is detected using a machine-learning classification model.The second method,precise re-learning after CSL(Ra-CSL),comprises a two-step training.It uses CSL as the 1st step analyzer,and the 2nd step analyzer is applied using the original dataset for a log that is detected as an attack in the 1st step analyzer.In the experiment,the bitmap rule was set based on the boundary value,which was 99.6%true positive on average for the attack and benign data found by analyzing the training data.Experimental results showed that the CSL was effective in reducing the training and detection time,and Ra-CSL was effective in increasing the detection rate.According to the experimental results,the data compression technique reduced the memory size by up to 20%and the training and detection times by 67%when compared with the conventional technique.In addition,the proposed technique improves the detection accuracy;the Naive Bayes model with the highest performance showed a detection rate of approximately 99%.展开更多
With the development of the 5th generation of mobile communi-cation(5G)networks and artificial intelligence(AI)technologies,the use of the Internet of Things(IoT)has expanded throughout industry.Although IoT networks ...With the development of the 5th generation of mobile communi-cation(5G)networks and artificial intelligence(AI)technologies,the use of the Internet of Things(IoT)has expanded throughout industry.Although IoT networks have improved industrial productivity and convenience,they are highly dependent on nonstandard protocol stacks and open-source-based,poorly validated software,resulting in several security vulnerabilities.How-ever,conventional AI-based software vulnerability discovery technologies cannot be applied to IoT because they require excessive memory and com-puting power.This study developed a technique for optimizing training data size to detect software vulnerabilities rapidly while maintaining learning accuracy.Experimental results using a software vulnerability classification dataset showed that different optimal data sizes did not affect the learning performance of the learning models.Moreover,the minimal data size required to train a model without performance degradation could be determined in advance.For example,the random forest model saved 85.18%of memory and improved latency by 97.82%while maintaining a learning accuracy similar to that achieved when using 100%of data,despite using only 1%.展开更多
Fly Ormia ochracea ears have been well-studied and mimicked to achieve subwavelength directional sensing,but their efficacy in sound source localization in three dimensions,utilizing sound from the X-,Y-,and Z-axes,ha...Fly Ormia ochracea ears have been well-studied and mimicked to achieve subwavelength directional sensing,but their efficacy in sound source localization in three dimensions,utilizing sound from the X-,Y-,and Z-axes,has been less explored.This paper focuses on a mm-sized array of three Ormia ochracea ear-inspired piezoelectric MEMS directional microphones,where their in-plane directionality is considered a cue to demonstrate sound source localization in three dimensions.In the array,biomimetic MEMS directional microphones are positioned in a 120°angular rotation;as a result,six diaphragms out of three directional microphones keep a normal-axis relative to the sound source at six diferent angles in the azimuth plane starting from 0°to 360°in intervals of±30°.In addition,the cosine-dependent horizontal component of the applied sound gives cues for Z-axis directional sensing.The whole array is frst analytically simulated and then experimentally measured in an anechoic chamber.Both results are found to be compliant,and the angular resolution of sound source localization in three dimensions is found to be±2°at the normal axis.The resolution at the azimuth plane is found to be±1.28,and the same array shows a±4.28°resolution when sound is varied from the elevation plane.Looking at the scope within this area combined with the presented results,this work provides a clear understanding of sound source localization in three dimensions.展开更多
Heavy metal pollution is a key environmental problem.Selectively extracting heavy metals could accomplish water purification and resource recycling simultaneously.Adsorption is a promising approach with a facile proce...Heavy metal pollution is a key environmental problem.Selectively extracting heavy metals could accomplish water purification and resource recycling simultaneously.Adsorption is a promising approach with a facile process,adaptability for the broad concentration of feed water,and high selectivity.However,the adsorption method faces challenges in synthesizing highperformance sorbents and regenerating adsorbents effectively.FeOOH is an environmentally friendly sorbent with low-cost production on a large scale.Nevertheless,the selectivity behavior and regeneration of FeOOH are seldom studied.Therefore,we investigated the selectivity of FeOOH in a mixed solution of Co^(2+),Ni^(2+),and Pb^(2+)and proposed to enhance the capacity of FeOOH and regenerate it by using external charges.Without charge,the FeOOH electrode shows a Pb^(2+)uptake capacity of 20 mg/g.After applying a voltage of-0.2/+0.8 V,the uptake capacity increases to a maximum of 42 mg/g and the desorption ratio is 70%-80%.In 35 cycles,FeOOH shows a superior selectivity towards Pb^(2+)compared with Co^(2+)and Ni^(2+),with a purity of 97%±3%in the extracts.The high selectivity is attributed to the lower activation energy for Pb^(2+)sorption.The capacity retentions at the 5^(th)and the 35^(th)cycles are ca.80%and ca.50%,respectively,comparable to the chemical regeneration method.With industrially exhausted granular ferric hydroxide as the electrode material,the system exhibits a Pb^(2+)uptake capacity of 37.4 mg/g with high selectivity.Our work demonstrates the feasibility of regenerating FeOOH by charge and provides a new approach for recycling and upcycling FeOOH sorbent.展开更多
基金supported by the Ministry of Trade,Industry and Energy(MOTIE)under Training Industrial Security Specialist for High-Tech Industry(RS-2024-00415520)supervised by the Korea Institute for Advancement of Technology(KIAT)the Ministry of Science and ICT(MSIT)under the ICT Challenge and Advanced Network of HRD(ICAN)Program(No.IITP-2022-RS-2022-00156310)supervised by the Institute of Information&Communication Technology Planning&Evaluation(IITP).
文摘As the density of wireless networks increases globally, the vulnerability of overlapped dense wireless communications to interference by hidden nodes and denial-of-service (DoS) attacks is becoming more apparent. There exists a gap in research on the detection and response to attacks on Medium Access Control (MAC) mechanisms themselves, which would lead to service outages between nodes. Classifying exploitation and deceptive jamming attacks on control mechanisms is particularly challengingdue to their resemblance to normal heavy communication patterns. Accordingly, this paper proposes a machine learning-based selective attack mitigation model that detects DoS attacks on wireless networks by monitoring packet log data. Based on the type of detected attack, it implements effective corresponding mitigation techniques to restore performance to nodes whose availability has been compromised. Experimental results reveal that the accuracy of the proposed model is 14% higher than that of a baseline anomaly detection model. Further, the appropriate mitigation techniques selected by the proposed system based on the attack type improve the average throughput by more than 440% compared to the case without a response.
基金supported by the National Key R&D Program of China(Grant No.2021YFA1202902)the National Natural Science Foundation of China(Grant Nos.12374292 and 12074244)B.L.acknowledges support from the Development Scholarship for Outstanding Ph.D.of Shanghai Jiao Tong University.J.K.acknowledges support from the National Research Foundation of Korea(NRF)grant funded by the Korean government(MSIT)(Grant No.NRF-RS-2024-00454528).
文摘A Luttinger liquid is a theoretical model describing interacting electrons in one-dimensional(1D)conductors.While individual 1D conductors have shown interesting Luttinger-liquid behaviors such as spin-charge separation and power-law spectral density,the more interesting phenomena predicted in coupled Luttinger liquids of neighboring 1D conductors have been rarely observed due to the difficulty in creating such structures.Recently,we have successfully grown close-packed carbon nanotube(CNT)arrays with uniform chirality,providing an ideal material system for studying the coupled Luttinger liquids.Here,we report on the observation of tunable hyperbolic plasmons in the coupled Luttinger liquids of CNT arrays using scanning near-field optical microscopy.These hyperbolic plasmons,resulting from the conductivity anisotropy in the CNT array,exhibit strong spatial confinement,in situ tunability,and a wide spectral range.Despite their hyperbolic wavefronts,the plasmon propagation in the axial direction still adheres to the Luttinger-liquid theory.Our work not only demonstrates a fascinating phenomenon in coupled Luttinger liquids for fundamental physics exploration,but also provides a highly confined and in situ tunable hyperbolic plasmon in close-packed CNT arrays for future nanophotonic devices and circuits.
基金funding from the Korea Institute for Advancement of Technology(KIAT)through a grant provided by the Korean Government Ministry of Trade,Industry,and Energy(MOTIE)(RS-2024-00415520,Training Industrial Security Specialist for High-Tech Industry)Additional support was received from the Ministry of Science and ICT(MSIT)under the ICAN(ICT Challenge and Advanced Network of HRD)program(No.IITP-2022-RS-2022-00156310)overseen by the Institute of Information&Communication Technology Planning and Evaluation(IITP).
文摘The Internet of Things(IoT)is extensively applied across various industrial domains,such as smart homes,factories,and intelligent transportation,becoming integral to daily life.Establishing robust policies for managing and governing IoT devices is imperative.Secure authentication for IoT devices in resource-constrained environments remains challenging due to the limitations of conventional complex protocols.Prior methodologies enhanced mutual authentication through key exchange protocols or complex operations,which are impractical for lightweight devices.To address this,our study introduces the privacy-preserving software-defined range proof(SDRP)model,which achieves secure authentication with low complexity.SDRP minimizes the overhead of confidentiality and authentication processes by utilizing range proof to verify whether the attribute information of a user falls within a specific range.Since authentication is performed using a digital ID sequence generated from indirect personal data,it can avoid the disclosure of actual individual attributes.Experimental results demonstrate that SDRP significantly improves security efficiency,increasing it by an average of 93.02%compared to conventional methods.It mitigates the trade-off between security and efficiency by reducing leakage risk by an average of 98.7%.
基金supported by the Korea Institute for Advancement of Technology(KIAT)Grant funded by the Korean Government(MOTIE)(P0008703,The Competency Development Program for Industry Specialists)MSIT under the ICAN(ICT Challenge and Advanced Network of HRD)Program(No.IITP-2022-RS-2022-00156310)supervised by the Institute of Information&Communication Technology Planning and Evaluation(IITP).
文摘With the advancement of wireless network technology,vast amounts of traffic have been generated,and malicious traffic attacks that threaten the network environment are becoming increasingly sophisticated.While signature-based detection methods,static analysis,and dynamic analysis techniques have been previously explored for malicious traffic detection,they have limitations in identifying diversified malware traffic patterns.Recent research has been focused on the application of machine learning to detect these patterns.However,applying machine learning to lightweight devices like IoT devices is challenging because of the high computational demands and complexity involved in the learning process.In this study,we examined methods for effectively utilizing machine learning-based malicious traffic detection approaches for lightweight devices.We introduced the suboptimal feature selection model(SFSM),a feature selection technique designed to reduce complexity while maintaining the effectiveness of malicious traffic detection.Detection performance was evaluated on various malicious traffic,benign,exploits,and generic,using the UNSW-NB15 dataset and SFSM sub-optimized hyperparameters for feature selection and narrowed the search scope to encompass all features.SFSM improved learning performance while minimizing complexity by considering feature selection and exhaustive search as two steps,a problem not considered in conventional models.Our experimental results showed that the detection accuracy was improved by approximately 20%compared to the random model,and the reduction in accuracy compared to the greedy model,which performs an exhaustive search on all features,was kept within 6%.Additionally,latency and complexity were reduced by approximately 96%and 99.78%,respectively,compared to the greedy model.This study demonstrates that malicious traffic can be effectively detected even in lightweight device environments.SFSM verified the possibility of detecting various attack traffic on lightweight devices.
基金supported by MOTIE under Training Industrial Security Specialist for High-Tech Industry(RS-2024-00415520)supervised by the Korea Institute for Advancement of Technology(KIAT),and by MSIT under the ICT Challenge and Advanced Network of HRD(ICAN)Program(No.IITP-2022-RS-2022-00156310)supervised by the Institute of Information&Communication Technology Planning&Evaluation(IITP)。
文摘With the rise of remote work and the digital industry,advanced cyberattacks have become more diverse and complex in terms of attack types and characteristics,rendering them difficult to detect with conventional intrusion detection methods.Signature-based intrusion detection methods can be used to detect attacks;however,they cannot detect new malware.Endpoint detection and response(EDR)tools are attracting attention as a means of detecting attacks on endpoints in real-time to overcome the limitations of signature-based intrusion detection techniques.However,EDR tools are restricted by the continuous generation of unnecessary logs,resulting in poor detection performance and memory efficiency.Machine learning-based intrusion detection techniques for responding to advanced cyberattacks are memory intensive,using numerous features;they lack optimal feature selection for each attack type.To overcome these limitations,this study proposes a memory-efficient intrusion detection approach incorporating multi-binary classifiers using optimal feature selection.The proposed model detects multiple types of malicious attacks using parallel binary classifiers with optimal features for each attack type.The experimental results showed a 2.95%accuracy improvement and an 88.05%memory reduction using only six features compared to a model with 18 features.Furthermore,compared to a conventional multi-classification model with simple feature selection based on permutation importance,the accuracy improved by 11.67%and the memory usage decreased by 44.87%.The proposed scheme demonstrates that effective intrusion detection is achievable with minimal features,making it suitable for memory-limited mobile and Internet of Things devices.
基金supplemented by a paper presented at the 6th International Symposium on Mobile Internet Security(MobiSec 2022).
文摘As time and space constraints decrease due to the development of wireless communication network technology,the scale and scope of cyber-attacks targeting the Internet of Things(IoT)are increasing.However,it is difficult to apply high-performance security modules to the IoT owing to the limited battery,memory capacity,and data transmission performance depend-ing on the size of the device.Conventional research has mainly reduced power consumption by lightening encryption algorithms.However,it is difficult to defend large-scale information systems and networks against advanced and intelligent attacks because of the problem of deteriorating security perfor-mance.In this study,we propose wake-up security(WuS),a low-power security architecture that can utilize high-performance security algorithms in an IoT environment.By introducing a small logic that performs anomaly detection on the IoT platform and executes the security module only when necessary according to the anomaly detection result,WuS improves security and power efficiency while using a relatively high-complexity security module in a low-power environment compared to the conventional method of periodically exe-cuting a high-performance security module.In this study,a Python simulator based on the UNSW-NB15 dataset is used to evaluate the power consumption,latency,and security of the proposed method.The evaluation results reveal that the power consumption of the proposed WuS mechanism is approxi-mately 51.8%and 27.2%lower than those of conventional high-performance security and lightweight security modules,respectively.Additionally,the laten-cies are approximately 74.8%and 65.9%lower,respectively.Furthermore,the WuS mechanism achieved a high detection accuracy of approximately 96.5%or greater,proving that the detection efficiency performance improved by approximately 33.5%compared to the conventional model.The performance evaluation results for the proposed model varied depending on the applied anomaly-detection model.Therefore,they can be used in various ways by selecting suitable models based on the performance levels required in each industry.
基金This work was supported by the BK-21 FOUR program through National Research Foundation of Korea(NRF)under Ministry of Education.
文摘Detection and classification of the blurred and the non-blurred regions in images is a challenging task due to the limited available information about blur type,scenarios and level of blurriness.In this paper,we propose an effective method for blur detection and segmentation based on transfer learning concept.The proposed method consists of two separate steps.In the first step,genetic programming(GP)model is developed that quantify the amount of blur for each pixel in the image.The GP model method uses the multiresolution features of the image and it provides an improved blur map.In the second phase,the blur map is segmented into blurred and non-blurred regions by using an adaptive threshold.A model based on support vector machine(SVM)is developed to compute adaptive threshold for the input blur map.The performance of the proposed method is evaluated using two different datasets and compared with various state-of-the-art methods.The comparative analysis reveals that the proposed method performs better against the state-of-the-art techniques.
基金This work is supported by the BK-21 FOUR program and by the Creative Challenge Research Program(2021R1I1A1A01052521)through National Research Foundation of Korea(NRF)under Ministry of Education,Korea.
文摘Enormousmethods have been proposed for the detection and segmentation of blur and non-blur regions of the images.Due to the limited available information about blur type,scenario and the level of blurriness,detection and segmentation is a challenging task.Hence,the performance of the blur measure operator is an essential factor and needs improvement to attain perfection.In this paper,we propose an effective blur measure based on local binary pattern(LBP)with adaptive threshold for blur detection.The sharpness metric developed based on LBP used a fixed threshold irrespective of the type and level of blur,that may not be suitable for images with variations in imaging conditions,blur amount and type.Contrarily,the proposed measure uses an adaptive threshold for each input image based on the image and blur properties to generate improved sharpness metric.The adaptive threshold is computed based on the model learned through support vector machine(SVM).The performance of the proposed method is evaluated using two different datasets and is compared with five state-of-the-art methods.Comparative analysis reveals that the proposed method performs significantly better qualitatively and quantitatively against all of the compared methods.
基金This study is the result of a commissioned research project supported by the affiliated institute of ETRI(No.2021-026)partially supported by the NationalResearch Foundation of Korea(NRF)grant funded by the Korean government(MSIT)(No.2020R1F1A1061107)+2 种基金the Korea Institute for Advancement of Technology(KIAT)grant funded by the Korean government(MOTIE)(P0008703,The Competency Development Program for Industry Specialist)the MSIT under the ICAN(ICT Challenge and Advanced Network of HRD)program[grant number IITP-2022-RS-2022-00156310]supervised by the Institute of Information&Communication Technology Planning and Evaluation(IITP).
文摘Recently,with the normalization of non-face-to-face online environments in response to the COVID-19 pandemic,the possibility of cyberattacks through endpoints has increased.Numerous endpoint devices are managed meticulously to prevent cyberattacks and ensure timely responses to potential security threats.In particular,because telecommuting,telemedicine,and teleeducation are implemented in uncontrolled environments,attackers typically target vulnerable endpoints to acquire administrator rights or steal authentication information,and reports of endpoint attacks have been increasing considerably.Advanced persistent threats(APTs)using various novel variant malicious codes are a form of a sophisticated attack.However,conventional commercial antivirus and anti-malware systems that use signature-based attack detectionmethods cannot satisfactorily respond to such attacks.In this paper,we propose a method that expands the detection coverage inAPT attack environments.In this model,an open-source threat detector and log collector are used synergistically to improve threat detection performance.Extending the scope of attack log collection through interworking between highly accessible open-source tools can efficiently increase the detection coverage of tactics and techniques used to deal with APT attacks,as defined by MITRE Adversarial Tactics,Techniques,and Common Knowledge(ATT&CK).We implemented an attack environment using an APT attack scenario emulator called Carbanak and analyzed the detection coverage of Google Rapid Response(GRR),an open-source threat detection tool,and Graylog,an open-source log collector.The proposed method expanded the detection coverage against MITRE ATT&CK by approximately 11%compared with that conventional methods.
基金supported by a Korea Institute for Advancement of Technology(KIAT)Grant funded by theKorean Government(MOTIE)(P0008703,The Competency Development Program for Industry Specialists)the MSIT under the ICAN(ICT Challenge and Advanced Network ofHRD)program(No.IITP-2022-RS-2022-00156310)supervised by the Institute of Information Communication Technology Planning and Evaluation(IITP).
文摘With the introduction of 5G technology,the application of Internet of Things(IoT)devices is expanding to various industrial fields.However,introducing a robust,lightweight,low-cost,and low-power security solution to the IoT environment is challenging.Therefore,this study proposes two methods using a data compression technique to detect malicious traffic efficiently and accurately for a secure IoT environment.The first method,compressed sensing and learning(CSL),compresses an event log in a bitmap format to quickly detect attacks.Then,the attack log is detected using a machine-learning classification model.The second method,precise re-learning after CSL(Ra-CSL),comprises a two-step training.It uses CSL as the 1st step analyzer,and the 2nd step analyzer is applied using the original dataset for a log that is detected as an attack in the 1st step analyzer.In the experiment,the bitmap rule was set based on the boundary value,which was 99.6%true positive on average for the attack and benign data found by analyzing the training data.Experimental results showed that the CSL was effective in reducing the training and detection time,and Ra-CSL was effective in increasing the detection rate.According to the experimental results,the data compression technique reduced the memory size by up to 20%and the training and detection times by 67%when compared with the conventional technique.In addition,the proposed technique improves the detection accuracy;the Naive Bayes model with the highest performance showed a detection rate of approximately 99%.
基金supported by a National Research Foundation of Korea (NRF)grant funded by the Ministry of Science and ICT (MSIT) (No.2020R1F1A1061107)the Korea Institute for Advancement of Technology (KIAT)grant funded by the Korean Government (MOTIE) (P0008703,The Competency Development Program for Industry Specialists)the MSIT under the ICAN (ICT Challenge and Advanced Network of HRD)program (No.IITP-2022-RS-2022-00156310)supervised by the Institute of Information&Communication Technology Planning and Evaluation (IITP).
文摘With the development of the 5th generation of mobile communi-cation(5G)networks and artificial intelligence(AI)technologies,the use of the Internet of Things(IoT)has expanded throughout industry.Although IoT networks have improved industrial productivity and convenience,they are highly dependent on nonstandard protocol stacks and open-source-based,poorly validated software,resulting in several security vulnerabilities.How-ever,conventional AI-based software vulnerability discovery technologies cannot be applied to IoT because they require excessive memory and com-puting power.This study developed a technique for optimizing training data size to detect software vulnerabilities rapidly while maintaining learning accuracy.Experimental results using a software vulnerability classification dataset showed that different optimal data sizes did not affect the learning performance of the learning models.Moreover,the minimal data size required to train a model without performance degradation could be determined in advance.For example,the random forest model saved 85.18%of memory and improved latency by 97.82%while maintaining a learning accuracy similar to that achieved when using 100%of data,despite using only 1%.
基金supported by grants NRF-2018R1A6A1A03025526 under the Priority Research Program through the National Research Foundation of Korea(NRF)under the Ministry of Education,and in partby NRF-2021R1A2C1004540 under the Ministry of Science and ICT,and in part by the BK-21 Four program through National Research Foundation of Korea(NRF)under Ministry of Education.
文摘Fly Ormia ochracea ears have been well-studied and mimicked to achieve subwavelength directional sensing,but their efficacy in sound source localization in three dimensions,utilizing sound from the X-,Y-,and Z-axes,has been less explored.This paper focuses on a mm-sized array of three Ormia ochracea ear-inspired piezoelectric MEMS directional microphones,where their in-plane directionality is considered a cue to demonstrate sound source localization in three dimensions.In the array,biomimetic MEMS directional microphones are positioned in a 120°angular rotation;as a result,six diaphragms out of three directional microphones keep a normal-axis relative to the sound source at six diferent angles in the azimuth plane starting from 0°to 360°in intervals of±30°.In addition,the cosine-dependent horizontal component of the applied sound gives cues for Z-axis directional sensing.The whole array is frst analytically simulated and then experimentally measured in an anechoic chamber.Both results are found to be compliant,and the angular resolution of sound source localization in three dimensions is found to be±2°at the normal axis.The resolution at the azimuth plane is found to be±1.28,and the same array shows a±4.28°resolution when sound is varied from the elevation plane.Looking at the scope within this area combined with the presented results,this work provides a clear understanding of sound source localization in three dimensions.
基金L.W.acknowledges funding from the Chinese Scholarship Council(CSC,No.201906260277)The work of L.D.was part of theÉcole Européenne d’Ingénieurs en Génie des Matériaux(EEIGM)carried out at Saarland University.Work at the Molecular Foundry was supported by the Office of Science,Office of Basic Energy Sciences,of the U.S.Department of Energy(No.DE-AC02-05CH11231)We acknowledge support for the eLiRec project by the European Union from the European Regional Development Fund(EFRE)and the State of Saarland,Germany.S.J.Z.acknowledges support from Tulane University.
文摘Heavy metal pollution is a key environmental problem.Selectively extracting heavy metals could accomplish water purification and resource recycling simultaneously.Adsorption is a promising approach with a facile process,adaptability for the broad concentration of feed water,and high selectivity.However,the adsorption method faces challenges in synthesizing highperformance sorbents and regenerating adsorbents effectively.FeOOH is an environmentally friendly sorbent with low-cost production on a large scale.Nevertheless,the selectivity behavior and regeneration of FeOOH are seldom studied.Therefore,we investigated the selectivity of FeOOH in a mixed solution of Co^(2+),Ni^(2+),and Pb^(2+)and proposed to enhance the capacity of FeOOH and regenerate it by using external charges.Without charge,the FeOOH electrode shows a Pb^(2+)uptake capacity of 20 mg/g.After applying a voltage of-0.2/+0.8 V,the uptake capacity increases to a maximum of 42 mg/g and the desorption ratio is 70%-80%.In 35 cycles,FeOOH shows a superior selectivity towards Pb^(2+)compared with Co^(2+)and Ni^(2+),with a purity of 97%±3%in the extracts.The high selectivity is attributed to the lower activation energy for Pb^(2+)sorption.The capacity retentions at the 5^(th)and the 35^(th)cycles are ca.80%and ca.50%,respectively,comparable to the chemical regeneration method.With industrially exhausted granular ferric hydroxide as the electrode material,the system exhibits a Pb^(2+)uptake capacity of 37.4 mg/g with high selectivity.Our work demonstrates the feasibility of regenerating FeOOH by charge and provides a new approach for recycling and upcycling FeOOH sorbent.
