Erik Blasch - Profile on Academia.edu (original) (raw)
Papers by Erik Blasch
Target broker compression for multi-level fusion
Information Fusion consists of low-level information fusion (LLIF) of object-level assessment whi... more Information Fusion consists of low-level information fusion (LLIF) of object-level assessment which is subject to many operating conditions of the sensor type, environment conditions, and the targets. Likewise, high-level information fusion (HLIF) requires proactive management of sensor parameters. One example of a parameter that affects downstream information fusion tasks of target tracking and identification is that of upstream image compression. In this paper, we present a technique for analyzing the effects of image compression on the information fusion result. The compression selections are based on user needs, target type, and information fusion function, which is a subject of the operating conditions. Results are presented that modify the Video National Imagery Interpretability Ratio (VNIIRS) equations to include compression requirements for object recognition, fusion of results, and user selections. The target broker compression method would support image fusion system providing an exemplar of LLIF-HLIF interactions.
Domain adaptation is a technology enabling aided target recognition and other algorithms for envi... more Domain adaptation is a technology enabling aided target recognition and other algorithms for environments and targets with data or labeled data that is scarce. Recent advances in unsupervised domain adaptation have demonstrated excellent performance but only when the domain shift is relatively small. We proposed targeted adversarial discriminative domain adaptation (T-ADDA), a semi-supervised domain adaptation method that extends the ADDA framework. By providing at least one labeled target image per class, used as a cue to guide the adaption, T-ADDA significantly boosts the performance of ADDA and is applicable to the challenging scenario in which the sets of targets in the source and target domains are not the same. The efficacy of T-ADDA is demonstrated by cross-domain, cross-sensor, and cross-target experiments using the common digits datasets and several aerial image datasets. Results demonstrate an average increase of 15% improvement with T-ADDA over ADDA using just a few labeled images when adapting to a small domain shift and afforded a 60% improvement when adapting to large domain shifts.
On the Development of a Classification Based Automated Motion Imagery Interpretability Prediction
Lecture Notes in Computer Science, 2021
A hidden chamber detector based on a MIMO SAR
This paper presents a hidden chamber detector (HCD) using radio frequency (RF) signals’ penetrati... more This paper presents a hidden chamber detector (HCD) using radio frequency (RF) signals’ penetration and reflection characteristics. The sensor of the chamber detector is a linear frequency modulated continuous wave (LFMCW) time division multiple access (TDMA) multiple input multiple output (MIMO) synthetic aperture radar (SAR). The basic idea of the sensor system is to scan the target wall and form a 2-D image of the wall with a high depth resolution and a proper angular resolution (cross range resolution). When there is no hidden chamber behind the wall, the radar will receive reflected signals from the room wall. While if there is a hidden chamber behind the room wall, there will be reflected signals from the room wall and the hidden chamber walls. Thus, the hidden chamber can be detected using the reflected singles from the chamber walls. The gated LFMCW TDMA MIMO SAR is configured with multiple transmitting antennas and multiple receiving antennas. Each transmitting and receiving antenna pair constructs an equivalent virtual array element. All virtual array elements construct the virtual antenna array (or synthetic aperture). The virtual array oversees the angular resolution along the array direction (vertical direction), and the narrow antenna beam is in charge of the angular resolution in the cross-array direction (horizontal direction). The high depth resolution is obtained using the gated LFMCW TDMA radar. Simulations show that the hidden chamber detector can detect chambers larger than 20cm by 20cm by 20cm at a distance of 0.3m away from the room wall.
In this paper, a fused feature vector consisting of raw time series and texture feature informati... more In this paper, a fused feature vector consisting of raw time series and texture feature information is proposed for space object classification. The time series data includes historical orbit trajectories and asteroid light curves. The texture feature is derived from recurrence plots using Gabor filters for both unsupervised learning and supervised learning algorithms. The simulation results show that the classification algorithms using the fused feature vector achieve better performance than those using raw time series or texture features only.
Cognitive radio testbed for Digital Beamforming of satellite communication
Digital Beamforming (DB) in satellite communication (SATCOM) has drawn significant attention in r... more Digital Beamforming (DB) in satellite communication (SATCOM) has drawn significant attention in recent years. DB design is flexible and adaptive among frequencies, bandwidths, and interferences. In this work, we design a cognitive radio testbed for SATCOM for digital beamforming design and implementation. The testbed is capable of simultaneously capturing signals from multiple antennas. Considering the conditions of jamming or interference, the design implements spectrum detection and compressive sensing for channel condition monitoring in a real time. Using the analyzed jammer and interference information, the incoming signal can be moved into a different frequency where the channel is vacant or at least meets acceptable conditions. The beamforming design is also adaptive to the frequency where each beam is weighted differently for a given channel condition. In addition, the satellite antenna array is used to determine the signal direction of arrival (DoA), The DB is capable of forming the beam in the direction that avoids interference and jamming. Moreover, the detection of the jamming signal direction can help localize the jammer. In the testbed design, a Universal Software Radio Peripheral (USRP) is used as a radio frequency (RF) front end component and GNU radio connecting to computer as signal processing back end. The system architecture includes the transponder receiver. And a USRP ×310 with twinRx receiver is capable of obtaining multiple copies of the signal captured by multiple receiving antennas.
IEEE Aerospace and Electronic Systems Magazine, Feb 1, 2018
2015 IEEE Aerospace Conference, 2015
Space object tracking and maneuver detection play an essential role in space situation awareness ... more Space object tracking and maneuver detection play an essential role in space situation awareness (SSA). The ordinary Kalman filter and its variants may give large error due to the maneuver of the space object. In this paper, to consistently track a maneuvering space object, the interacting multiple model (IMM) filter is utilized. Multiple Models with different process noise levels are used to distinguish the maneuvering effects. The IMM cubature Kalman filter (IMM-CKF) is used to track the maneuvering space object which considers of the geometric relations between the space object, space based optical (SBO) sensor, and the sun. The geometric relation highly affects the quality of the observation. A scenario which contains a target spacecraft and four SBO sensors is used to test performance of the IMM-CKF. We also compare the IMM-CKF and the ordinary cubature Kalman filter (CKF). The results indicate that IMM-CKF is more robust than the CKF when the space object undergoes a maneuver. In addition, the detection of a maneuver can be obtained by using the IMM-CKF. Hence, IMM-CKF could facilitate future SBO based SSA mission awareness.
Cyber-physical systems (CPS) are systems with a tight coupling of the cyber aspects of computing ... more Cyber-physical systems (CPS) are systems with a tight coupling of the cyber aspects of computing and communications with the physical aspects of dynamics and engineering that abide by the laws of physics. The real-time monitoring provided by wireless sensor networks (WSNs) is essential for CPS, as it provides rich and pertinent information on the condition of physical systems. In WSNs, the attackers could inject false measurements to the controller through compromised sensor nodes, which not only threaten the security of the system, but also consume significant network resources and pose serious threats to the lifetime of sensor networks. To mitigate false data injection (FDI) measurement attacks, a number of situation aware en-route filtering schemes to filter false data inside the networks have been developed. In this book chapter, the authors first review those existing situation aware en-route filter mechanisms such as: Statistical En-route Filtering (SEF), Location-Based Resilient Secrecy (LBRS), Location-ware End-to-end Data Security (LEDS), and Dynamic En-route Filtering Scheme (DEFS). The authors then compare the performance of those schemes via both the theoretical analysis and simulation study. These extensive simulations validate findings that most of the schemes can filter out false data within few hops, and the filtering efficiency increases as the number of hops increases and the filtering efficiency of most schemes decreases rapidly as the number of compromised nodes increases.
IEEE Aerospace and Electronics Systems Magazine, 2021
Proceedings of SPIE, May 13, 2016
This paper investigates weather effects on a satellite communication (SATCOM) link communication ... more This paper investigates weather effects on a satellite communication (SATCOM) link communication channel model. Specifically, rain attenuation in the Ka band and X band of the SATCOM link for both uplink and downlink scenarios are presented. The weather model for the SATCOM link uses a Markov chain model with an average probability and transition probability for different states of weather, to investigate the impact of dynamic weather on the SATCOM link channel propagation model. Also, a power control method is proposed to achieve the required carrier to noise ratio in a SATCOM scenario using a Bayesian Network in Netica. The Bayesian Network models the space-ground link geometry and transmit power control to adapt to the dynamic weather. Simulations are implemented for the weather states during relatively long and short periods, path loss variations, and transmit power distributions over different scenarios. The simulation results demonstrate the effectiveness of the proposed weather model, Markov chain model, and the power control method for SATCOM.
Analysis of thermal Infrared (IR) imagery is critical to many law enforcement and military missio... more Analysis of thermal Infrared (IR) imagery is critical to many law enforcement and military missions, particularly for operations at night or in low-light conditions. Transmitting the imagery data from the sensor to the operator often relies on limited bandwidth channels, leading to information loss. This paper develops a method, known as the Compression Degradation Image Function Index (CoDIFI) framework, that predicts the degradation in interpretability associated with the specific image compression method and level of compression. Quantification of the image interpretability relies on the National Imagery Interpretability Ratings Scale (NIIRS). Building on previously reported development and validation of CoDIFI operating on electro-optical (EO) imagery collected in the visible region, this paper extends CoDIFI to imagery collected in the mid-wave infrared (MWIR) region, approximately 3 to 5 microns. For the infrared imagery application, the IR NIIRS is the standard for quantifying image interpretability and the prediction model rests on the general image quality equation (GIQE). A prediction model using the CoDIFI for IR imagery is established with empirical validation. By leveraging the CoDIFI in operational settings, mission success ensures that the compression selection is achievable in terms of the NIIRS level of imagery data delivered to users, while optimizing the use of scarce data transmission capacity.
Tailoring image compression to mission needs: Predicting NIIRS loss due to image compression
Transmission and analysis of imagery for law enforcement and military missions is often constrain... more Transmission and analysis of imagery for law enforcement and military missions is often constrained by the capacity of available communications channels. Nevertheless, achieving success in operational missions requires acquisition and analysis of imagery that satisfies specific interpretability requirements. By expressing these requirements in terms of the National Imagery Interpretability Ratings Scale (NIIRS), we have developed a method for predicting the NIIRS loss associated with various methods and levels of imagery compression. Our method, known as the Compression Degradation Image Function Index (CoDIFI) framework automatically predicts the NIIRS degradation associated with the specific image compression method and level of compression. In this paper, we first review NIIRS and methods for predicting it followed by the presentation of the CoDIFI framework and we put our emphasis on the results of the empirical validation experiments. By leveraging CoDIFI in operational settings, our goal is to ensure mission success in terms of the NIIRS level of imagery data delivered to users, while optimizing the use of scarce data transmission capacity.
Deep Learning Based Domain Adaptation with Data Fusion for Aerial Image Data Analysis
Lecture Notes in Computer Science, 2021
Anomaly detection of unstructured big data via semantic analysis and dynamic knowledge graph construction
There is an increasing need for both governments and businesses to discover latent anomalous acti... more There is an increasing need for both governments and businesses to discover latent anomalous activities in unstructured publicly-available data, produced by professional agencies and the general public. Over the past two decades, consumers have begun to use smart devices to both take in and generate a large volume of open-source text-based data, providing the opportunity for latent anomaly analysis. However, real-time data acquisition, and the processing and interpretation of various types of unstructured data, remains a great challenge. Recent efforts have focused on artificial intelligence / machine learning (AI/ML) solutions to accelerate the labor-intensive linear collection, exploitation, and dissemination analysis cycle and enhance it with a data-driven rapid integration and correlation process of open-source data. This paper describes an Activity Based Intelligence framework for anomaly detection of open-source big data using AI/ML to perform semantic analysis. The proposed Anomaly Detection using Semantic Analysis Knowledge (ADUSAK) framework includes four layers: input layer, knowledge layer, reasoning layer, and graphical user interface (GUI)/output layer. The corresponding main technologies include: Information Extraction, Knowledge Graph (KG) construction, Semantic Reasoning, and Pattern Discovery. Finally, ADUSAK was verified by performing Emerging Events Detection, Fake News Detection, and Suspicious Network Analysis. The generalized ADUSAK framework can be easily extended to a wide range of applications by adjusting the data collection, modeling construction, and event alerting.
Optical Engineering, Mar 12, 2019
Multisource AI scorecard table analysis of AMIGO
Various techniques, applications, and tools for space situational awareness (SSA) have been devel... more Various techniques, applications, and tools for space situational awareness (SSA) have been developed for specific functions that can provide decision support tools. The generality of tools to enable a user-defined operating picture (UDOP) enables analysis across a wide variety of applications. This paper explores the Multisource AI Scorecard Table (MAST) for artificial intelligence/machine learning methods. Using the MAST categories, the Adaptive Markov Inference Game Optimization (AMIGO) SSA tool is presented as an example. The analysis reveals the importance of human interaction in the task, user, and technology operations. Recent advances in artificial intelligence (AI) have led to an explosion of multimedia applications (e.g., computer vision (CV) and natural language processing (NLP)) for different domains such as commercial, industrial, and intelligence. In particular, the use of AI applications is often problematic because the opaque nature of most systems leads to an inability for a human to understand how the results came about. A reliance on “black boxes” to generate predictions and inform decisions but requires explainability. This paper explores how MAST can support human-machine interactions to support the design and development of SSA tools. After describing the elements of MAST, the use case for AMIGO explains the general rating concept for the community to consider and modify for the interpretability of advanced data analytics that support various elements of data awareness.
Secure spectrum-efficient frequency hopping for return link of protected tactical satellite communications
Protected tactical waveform (PTW) provides cost-efficient, jamming-resistant communications via b... more Protected tactical waveform (PTW) provides cost-efficient, jamming-resistant communications via both government and commercial satellites over multiple frequency bands such as the C-, Ku-, Ka-, and X-band. Frequency hopping (FH) was originally developed for secure communications by exploiting time-frequency diversity over a large spectrum. The anti-jamming feature of the frequency hopping technique helps communication systems improve the reliability of transmission. In this paper, we consider a secure channel group hopping (CGH) scheme in order to satisfy the needs of protected tactical satellite communication system. This scheme is proposed to ensure that each channel group hops to a new hopping spectrum in a pseudo-random manner during each hopping period, and different channel groups always transmit over non-overlapping frequency bands. Our analysis suggests that the proposed CGH scheme achieves collision-free frequency hopping among channel groups, such that a higher spectrum efficiency of the system can be obtained. Simulation results are provided to validate the performance of the proposed CGH scheme for PTW system.
Class disagreement detection with its application to EO-SAR fusion
RODAD: Resilience Oriented Decentralized Anomaly Detection for Urban Air Mobility Networks
Target broker compression for multi-level fusion
Information Fusion consists of low-level information fusion (LLIF) of object-level assessment whi... more Information Fusion consists of low-level information fusion (LLIF) of object-level assessment which is subject to many operating conditions of the sensor type, environment conditions, and the targets. Likewise, high-level information fusion (HLIF) requires proactive management of sensor parameters. One example of a parameter that affects downstream information fusion tasks of target tracking and identification is that of upstream image compression. In this paper, we present a technique for analyzing the effects of image compression on the information fusion result. The compression selections are based on user needs, target type, and information fusion function, which is a subject of the operating conditions. Results are presented that modify the Video National Imagery Interpretability Ratio (VNIIRS) equations to include compression requirements for object recognition, fusion of results, and user selections. The target broker compression method would support image fusion system providing an exemplar of LLIF-HLIF interactions.
Domain adaptation is a technology enabling aided target recognition and other algorithms for envi... more Domain adaptation is a technology enabling aided target recognition and other algorithms for environments and targets with data or labeled data that is scarce. Recent advances in unsupervised domain adaptation have demonstrated excellent performance but only when the domain shift is relatively small. We proposed targeted adversarial discriminative domain adaptation (T-ADDA), a semi-supervised domain adaptation method that extends the ADDA framework. By providing at least one labeled target image per class, used as a cue to guide the adaption, T-ADDA significantly boosts the performance of ADDA and is applicable to the challenging scenario in which the sets of targets in the source and target domains are not the same. The efficacy of T-ADDA is demonstrated by cross-domain, cross-sensor, and cross-target experiments using the common digits datasets and several aerial image datasets. Results demonstrate an average increase of 15% improvement with T-ADDA over ADDA using just a few labeled images when adapting to a small domain shift and afforded a 60% improvement when adapting to large domain shifts.
On the Development of a Classification Based Automated Motion Imagery Interpretability Prediction
Lecture Notes in Computer Science, 2021
A hidden chamber detector based on a MIMO SAR
This paper presents a hidden chamber detector (HCD) using radio frequency (RF) signals’ penetrati... more This paper presents a hidden chamber detector (HCD) using radio frequency (RF) signals’ penetration and reflection characteristics. The sensor of the chamber detector is a linear frequency modulated continuous wave (LFMCW) time division multiple access (TDMA) multiple input multiple output (MIMO) synthetic aperture radar (SAR). The basic idea of the sensor system is to scan the target wall and form a 2-D image of the wall with a high depth resolution and a proper angular resolution (cross range resolution). When there is no hidden chamber behind the wall, the radar will receive reflected signals from the room wall. While if there is a hidden chamber behind the room wall, there will be reflected signals from the room wall and the hidden chamber walls. Thus, the hidden chamber can be detected using the reflected singles from the chamber walls. The gated LFMCW TDMA MIMO SAR is configured with multiple transmitting antennas and multiple receiving antennas. Each transmitting and receiving antenna pair constructs an equivalent virtual array element. All virtual array elements construct the virtual antenna array (or synthetic aperture). The virtual array oversees the angular resolution along the array direction (vertical direction), and the narrow antenna beam is in charge of the angular resolution in the cross-array direction (horizontal direction). The high depth resolution is obtained using the gated LFMCW TDMA radar. Simulations show that the hidden chamber detector can detect chambers larger than 20cm by 20cm by 20cm at a distance of 0.3m away from the room wall.
In this paper, a fused feature vector consisting of raw time series and texture feature informati... more In this paper, a fused feature vector consisting of raw time series and texture feature information is proposed for space object classification. The time series data includes historical orbit trajectories and asteroid light curves. The texture feature is derived from recurrence plots using Gabor filters for both unsupervised learning and supervised learning algorithms. The simulation results show that the classification algorithms using the fused feature vector achieve better performance than those using raw time series or texture features only.
Cognitive radio testbed for Digital Beamforming of satellite communication
Digital Beamforming (DB) in satellite communication (SATCOM) has drawn significant attention in r... more Digital Beamforming (DB) in satellite communication (SATCOM) has drawn significant attention in recent years. DB design is flexible and adaptive among frequencies, bandwidths, and interferences. In this work, we design a cognitive radio testbed for SATCOM for digital beamforming design and implementation. The testbed is capable of simultaneously capturing signals from multiple antennas. Considering the conditions of jamming or interference, the design implements spectrum detection and compressive sensing for channel condition monitoring in a real time. Using the analyzed jammer and interference information, the incoming signal can be moved into a different frequency where the channel is vacant or at least meets acceptable conditions. The beamforming design is also adaptive to the frequency where each beam is weighted differently for a given channel condition. In addition, the satellite antenna array is used to determine the signal direction of arrival (DoA), The DB is capable of forming the beam in the direction that avoids interference and jamming. Moreover, the detection of the jamming signal direction can help localize the jammer. In the testbed design, a Universal Software Radio Peripheral (USRP) is used as a radio frequency (RF) front end component and GNU radio connecting to computer as signal processing back end. The system architecture includes the transponder receiver. And a USRP ×310 with twinRx receiver is capable of obtaining multiple copies of the signal captured by multiple receiving antennas.
IEEE Aerospace and Electronic Systems Magazine, Feb 1, 2018
2015 IEEE Aerospace Conference, 2015
Space object tracking and maneuver detection play an essential role in space situation awareness ... more Space object tracking and maneuver detection play an essential role in space situation awareness (SSA). The ordinary Kalman filter and its variants may give large error due to the maneuver of the space object. In this paper, to consistently track a maneuvering space object, the interacting multiple model (IMM) filter is utilized. Multiple Models with different process noise levels are used to distinguish the maneuvering effects. The IMM cubature Kalman filter (IMM-CKF) is used to track the maneuvering space object which considers of the geometric relations between the space object, space based optical (SBO) sensor, and the sun. The geometric relation highly affects the quality of the observation. A scenario which contains a target spacecraft and four SBO sensors is used to test performance of the IMM-CKF. We also compare the IMM-CKF and the ordinary cubature Kalman filter (CKF). The results indicate that IMM-CKF is more robust than the CKF when the space object undergoes a maneuver. In addition, the detection of a maneuver can be obtained by using the IMM-CKF. Hence, IMM-CKF could facilitate future SBO based SSA mission awareness.
Cyber-physical systems (CPS) are systems with a tight coupling of the cyber aspects of computing ... more Cyber-physical systems (CPS) are systems with a tight coupling of the cyber aspects of computing and communications with the physical aspects of dynamics and engineering that abide by the laws of physics. The real-time monitoring provided by wireless sensor networks (WSNs) is essential for CPS, as it provides rich and pertinent information on the condition of physical systems. In WSNs, the attackers could inject false measurements to the controller through compromised sensor nodes, which not only threaten the security of the system, but also consume significant network resources and pose serious threats to the lifetime of sensor networks. To mitigate false data injection (FDI) measurement attacks, a number of situation aware en-route filtering schemes to filter false data inside the networks have been developed. In this book chapter, the authors first review those existing situation aware en-route filter mechanisms such as: Statistical En-route Filtering (SEF), Location-Based Resilient Secrecy (LBRS), Location-ware End-to-end Data Security (LEDS), and Dynamic En-route Filtering Scheme (DEFS). The authors then compare the performance of those schemes via both the theoretical analysis and simulation study. These extensive simulations validate findings that most of the schemes can filter out false data within few hops, and the filtering efficiency increases as the number of hops increases and the filtering efficiency of most schemes decreases rapidly as the number of compromised nodes increases.
IEEE Aerospace and Electronics Systems Magazine, 2021
Proceedings of SPIE, May 13, 2016
This paper investigates weather effects on a satellite communication (SATCOM) link communication ... more This paper investigates weather effects on a satellite communication (SATCOM) link communication channel model. Specifically, rain attenuation in the Ka band and X band of the SATCOM link for both uplink and downlink scenarios are presented. The weather model for the SATCOM link uses a Markov chain model with an average probability and transition probability for different states of weather, to investigate the impact of dynamic weather on the SATCOM link channel propagation model. Also, a power control method is proposed to achieve the required carrier to noise ratio in a SATCOM scenario using a Bayesian Network in Netica. The Bayesian Network models the space-ground link geometry and transmit power control to adapt to the dynamic weather. Simulations are implemented for the weather states during relatively long and short periods, path loss variations, and transmit power distributions over different scenarios. The simulation results demonstrate the effectiveness of the proposed weather model, Markov chain model, and the power control method for SATCOM.
Analysis of thermal Infrared (IR) imagery is critical to many law enforcement and military missio... more Analysis of thermal Infrared (IR) imagery is critical to many law enforcement and military missions, particularly for operations at night or in low-light conditions. Transmitting the imagery data from the sensor to the operator often relies on limited bandwidth channels, leading to information loss. This paper develops a method, known as the Compression Degradation Image Function Index (CoDIFI) framework, that predicts the degradation in interpretability associated with the specific image compression method and level of compression. Quantification of the image interpretability relies on the National Imagery Interpretability Ratings Scale (NIIRS). Building on previously reported development and validation of CoDIFI operating on electro-optical (EO) imagery collected in the visible region, this paper extends CoDIFI to imagery collected in the mid-wave infrared (MWIR) region, approximately 3 to 5 microns. For the infrared imagery application, the IR NIIRS is the standard for quantifying image interpretability and the prediction model rests on the general image quality equation (GIQE). A prediction model using the CoDIFI for IR imagery is established with empirical validation. By leveraging the CoDIFI in operational settings, mission success ensures that the compression selection is achievable in terms of the NIIRS level of imagery data delivered to users, while optimizing the use of scarce data transmission capacity.
Tailoring image compression to mission needs: Predicting NIIRS loss due to image compression
Transmission and analysis of imagery for law enforcement and military missions is often constrain... more Transmission and analysis of imagery for law enforcement and military missions is often constrained by the capacity of available communications channels. Nevertheless, achieving success in operational missions requires acquisition and analysis of imagery that satisfies specific interpretability requirements. By expressing these requirements in terms of the National Imagery Interpretability Ratings Scale (NIIRS), we have developed a method for predicting the NIIRS loss associated with various methods and levels of imagery compression. Our method, known as the Compression Degradation Image Function Index (CoDIFI) framework automatically predicts the NIIRS degradation associated with the specific image compression method and level of compression. In this paper, we first review NIIRS and methods for predicting it followed by the presentation of the CoDIFI framework and we put our emphasis on the results of the empirical validation experiments. By leveraging CoDIFI in operational settings, our goal is to ensure mission success in terms of the NIIRS level of imagery data delivered to users, while optimizing the use of scarce data transmission capacity.
Deep Learning Based Domain Adaptation with Data Fusion for Aerial Image Data Analysis
Lecture Notes in Computer Science, 2021
Anomaly detection of unstructured big data via semantic analysis and dynamic knowledge graph construction
There is an increasing need for both governments and businesses to discover latent anomalous acti... more There is an increasing need for both governments and businesses to discover latent anomalous activities in unstructured publicly-available data, produced by professional agencies and the general public. Over the past two decades, consumers have begun to use smart devices to both take in and generate a large volume of open-source text-based data, providing the opportunity for latent anomaly analysis. However, real-time data acquisition, and the processing and interpretation of various types of unstructured data, remains a great challenge. Recent efforts have focused on artificial intelligence / machine learning (AI/ML) solutions to accelerate the labor-intensive linear collection, exploitation, and dissemination analysis cycle and enhance it with a data-driven rapid integration and correlation process of open-source data. This paper describes an Activity Based Intelligence framework for anomaly detection of open-source big data using AI/ML to perform semantic analysis. The proposed Anomaly Detection using Semantic Analysis Knowledge (ADUSAK) framework includes four layers: input layer, knowledge layer, reasoning layer, and graphical user interface (GUI)/output layer. The corresponding main technologies include: Information Extraction, Knowledge Graph (KG) construction, Semantic Reasoning, and Pattern Discovery. Finally, ADUSAK was verified by performing Emerging Events Detection, Fake News Detection, and Suspicious Network Analysis. The generalized ADUSAK framework can be easily extended to a wide range of applications by adjusting the data collection, modeling construction, and event alerting.
Optical Engineering, Mar 12, 2019
Multisource AI scorecard table analysis of AMIGO
Various techniques, applications, and tools for space situational awareness (SSA) have been devel... more Various techniques, applications, and tools for space situational awareness (SSA) have been developed for specific functions that can provide decision support tools. The generality of tools to enable a user-defined operating picture (UDOP) enables analysis across a wide variety of applications. This paper explores the Multisource AI Scorecard Table (MAST) for artificial intelligence/machine learning methods. Using the MAST categories, the Adaptive Markov Inference Game Optimization (AMIGO) SSA tool is presented as an example. The analysis reveals the importance of human interaction in the task, user, and technology operations. Recent advances in artificial intelligence (AI) have led to an explosion of multimedia applications (e.g., computer vision (CV) and natural language processing (NLP)) for different domains such as commercial, industrial, and intelligence. In particular, the use of AI applications is often problematic because the opaque nature of most systems leads to an inability for a human to understand how the results came about. A reliance on “black boxes” to generate predictions and inform decisions but requires explainability. This paper explores how MAST can support human-machine interactions to support the design and development of SSA tools. After describing the elements of MAST, the use case for AMIGO explains the general rating concept for the community to consider and modify for the interpretability of advanced data analytics that support various elements of data awareness.
Secure spectrum-efficient frequency hopping for return link of protected tactical satellite communications
Protected tactical waveform (PTW) provides cost-efficient, jamming-resistant communications via b... more Protected tactical waveform (PTW) provides cost-efficient, jamming-resistant communications via both government and commercial satellites over multiple frequency bands such as the C-, Ku-, Ka-, and X-band. Frequency hopping (FH) was originally developed for secure communications by exploiting time-frequency diversity over a large spectrum. The anti-jamming feature of the frequency hopping technique helps communication systems improve the reliability of transmission. In this paper, we consider a secure channel group hopping (CGH) scheme in order to satisfy the needs of protected tactical satellite communication system. This scheme is proposed to ensure that each channel group hops to a new hopping spectrum in a pseudo-random manner during each hopping period, and different channel groups always transmit over non-overlapping frequency bands. Our analysis suggests that the proposed CGH scheme achieves collision-free frequency hopping among channel groups, such that a higher spectrum efficiency of the system can be obtained. Simulation results are provided to validate the performance of the proposed CGH scheme for PTW system.
Class disagreement detection with its application to EO-SAR fusion
RODAD: Resilience Oriented Decentralized Anomaly Detection for Urban Air Mobility Networks
– A human presented with a variety of displays is expected to fuse data to obtain information. An... more – A human presented with a variety of displays is expected to fuse data to obtain information. An effective presentation of information would assist the human in fusing data. This paper describes a multisensor-multisource information decision making tool that was designed to augment human cognitive fusion.
Blasch, Erik Philip. Ph.D., Department of Engineering, Wright State University, 1999. Derivation ... more Blasch, Erik Philip. Ph.D., Department of Engineering, Wright State University, 1999. Derivation of a Belief Filter for High Range Resolution Radar Simultaneous Target Tracking and Identification. Standard multitarget tracking algorithms employ Bayesian updating to associate the highest measurement probability to target tracks. Limitations of traditional tracking algorithms are that measurement-to-track associations do not account for target-type uncertainties, target identities, or incomplete knowledge. The dissertation develops the Joint Belief-Probabilistic Data Association (JBPDA) and the Set-Based Data Association (SBDA) simultaneous tracking and identification algorithms to: advance data association tracking techniques, provide an alternative to Bayesian tracking methods, and demonstrate a fusion of stochastic and set-based uncertainties. The JBPDA algorithm fuses kinematic-continuous and identification-discrete features to track and identify targets. The set of features includes moving-target indicator (MTI) position hits and highrange resolution (HRR) radar range-bin locations and amplitudes. Kinematic states include target positions, velocities, and poses, where pose is the aspect angle of the target. Using MTI positions and estimated track-pose information, HRR features are extracted to obtain a target belief-ID with an associated belief-pose. To facilitate the combination of continuous-probabilistic and discrete-set mathematics, a belief-probabilistic uncertainty calculus is devised to combine track-pose and belief-pose. The intersection, association, and filtering of track and identification beliefs results in a recursive simultaneous tracking and identification algorithm.