Sensor Resource Management Research Articles

Feedback-Assisted Automatic Target and Clutter Discrimination Using a Bayesian Convolutional Neural Network for Improved Explainability in SAR Applications

In this paper, a feedback training approach for efficiently dealing with distribution shift in synthetic aperture radar target detection using a Bayesian convolutional neural network is proposed. After training the network on in-distribution data, it is tested on out-of-distribution data. Samples that are classified incorrectly with high certainty are fed back for a second round of training. This results in the reduction of false positives in the out-of-distribution dataset. False positive target detections challenge human attention, sensor resource management, and mission engagement. In these types of applications, a reduction in false positives thus often takes precedence over target detection and classification performance. The classifier is used to discriminate the targets from the clutter and to classify the target type in a single step as opposed to the traditional approach of having a sequential chain of functions for target detection and localisation before the machine learning algorithm. Another aspect of automated synthetic aperture radar detection and recognition problems addressed here is the fact that human users of the output of traditional classification systems are presented with decisions made by “black box” algorithms. Consequently, the decisions are not explainable, even to an expert in the sensor domain. This paper makes use of the concept of explainable artificial intelligence via uncertainty heat maps that are overlaid onto synthetic aperture radar imagery to furnish the user with additional information about classification decisions. These uncertainty heat maps facilitate trust in the machine learning algorithm and are derived from the uncertainty estimates of the classifications from the Bayesian convolutional neural network. These uncertainty overlays further enhance the users’ ability to interpret the reasons why certain decisions were made by the algorithm. Further, it is demonstrated that feeding back the high-certainty, incorrectly classified out-of-distribution data results in an average improvement in detection performance and a reduction in uncertainty for all synthetic aperture radar images processed. Compared to the baseline method, an improvement in recall of 11.8%, and a reduction in the false positive rate of 7.08% were demonstrated using the Feedback-assisted Bayesian Convolutional Neural Network or FaBCNN.

Adaptive resource management using many-core processing for fault tolerance based on cyber–physical cloud systems

With the increasing utilization of cloud computing and cyber–physical systems (CPSs), which allow the expression and control of the real world in a virtual environment, researches related to these subjects are being actively conducted in various areas. The convergence of CPS and cloud computing is being researched primarily because of their high availability, high-performance computing, and high-throughput computing. CPS consisting of numerous sensors, actuators, controllers, and control managers requires optimized modeling, simulation, and resource management technologies to integrate physical elements with computing elements for processing, which will provide high-throughput computing and high-reliability services. But the main problem of sensor resource management is that information of sensors cannot be approached in case that a sensor failure occurs at the sensing target area. Thus, various researches have been done to reconstruct the topology, but the self-topology configuration of sensors causes unnecessary events and battery consumption from various sensor nodes.In this paper, adaptive resource management (ARM) is proposed to 1) minimize information loss due to the irregular lifespan of resources, such as sensors and actuators; and 2) quickly respond to any problems. ARM uses the many-core of GPU to speed up fault handling, parallelizes the sensor information to select an alternate node of the fault node, and presents the performance evaluation results of the execution time of CPU and GPU.

Developing Mathematical Formulations for the Integrated Problem of Sensors, Weapons, and Targets

Although sensor resource management and weapon–target assignment problems have been well studied, there is currently no well-posed mathematical formulation that captures the interactions of these problems in the context of a missile defense scenario. This paper gives a representative example to illustrate this fact; then, it poses a method for navigating the abstraction space for the integrated problem of sensors, weapons, and targets based on the specification of information sharing between subproblems. This method begins by outlining a set of formulation classes to specify the way in which information is shared; from this set, some formulation schemata (particular instantiations of a mathematical models that subscribe to the rules of a formulation class) are suggested. A set of goodness criteria for selecting which formulation classes and formulation schemata best address the problem-absent simulation results is considered. Last, the benefits and shortcomings of these formulation classes and formulation schemata, as well as areas for future work, are discussed.

Sensor grid resource management: Model and implementation issues

This paper studies optimal sensor resource management in sensor grids. We formalize the problem using nonlinear optimization theory, which incorporates sensor resource constraint, energy, and expense budget. The paper also presents a pricing-based iterative algorithm for sensor management which balances the sensor user׳ QoS requirements to achieve a sensor system optimization based on the preference of the sensor service users. The paper discusses implementation issues of sensor management. Simulations reveal that the proposed sensor management algorithms can obtain better performance than a previous approach.

A Game Theoretic Sensor Resource Allocation Using Fuzzy Logic

A sensor resource management system that employs fuzzy logic to provide the utility functions to a game theoretic approach is developed. The application looks at a virtual fence problem where several unattended ground sensors are placed in remote locations to act as virtual sentries. The goal of the approach is to maximize the battery life while tracking targets of interest. This research also considers the incorporation of uncertainty into the fuzzy membership functions. Both type-2 fuzzy logic and the use of conditional fuzzy membership function are employed. The type-2 fuzzy logic is employed in the case of acoustical sensor tracking accuracy degradation, while the condition-based membership functions are used to adapt to different conditions, such as environmental conditions and sensor performance degradation, over time. The resource management process uses fuzzy logic to determine which of the sensor systems on a sensor pod is used to provide initial classification of the target and which sensor or sensors are to be used in tracking and better classifying the target if it is determined to be of value to the mission. The three different approaches are compared to determine when the best times for the more complex approaches are warranted.

Performance Measures of Covariance and Information Matrices in Resource Management for Target State Estimation

In target tracking, sensor resource management (SRM) assigns to each target a best combination of sensors, which requires performance analysis of track filter updates. Two popular implementations of track filters are the Kalman filter (or covariance filter) and the information filter. SRM with Kalman filters attempts to minimize the estimation error covariance matrix-based scalar performance measures, whereas SRM with information filters aims to maximize the information matrix-based counterpart. In this paper, we investigate issues related to scalar performance measures and, in particular, compare the use of trace, determinant, and eigenvalues of the covariance matrix or information matrix as scalar performance measures. The study demonstrates which matrix measures are appropriate for resource management applications. Furthermore, the study shows when the matrix measures lead to equivalent goals. While this analysis is agnostic to the type of measurement, the paper demonstrates how to accommodate bearing and range measurements. Overall, the analysis provides insight about how sensor measurements best reduce uncertainty so that we can properly exploit performance measures to satisfy requirements of practical tracking and SRM applications.

An energy-efficient self-deployment with the centroid-directed virtual force in mobile sensor networks

The deployment of sensors at their intended locations enables not only adequate sensing coverage of the area of interest, but also efficient sensor resource management. However, for static wireless sensor networks, it is sometimes impossible to manually deploy the sensors in those locations as they can be distributed in unexploited, hostile, or disaster areas. Nevertheless, if each sensor has locomotive capability, they can re-deploy themselves using the location information of neighboring sensors. In particular, in our previous study, we showed that the coverage area can be efficiently expanded by having sensors move to the centroids of their Voronoi polygon generated using the location information of neighboring sensors. In this paper, we present an energy-efficient self-deployment scheme to utilize the attractive force generated from the centroid of a sensor’s local Voronoi polygon as well as the repulsive force frequently used in self-deployment schemes using the potential field. We also provide the design and implementation of the simulator used to analyze the performance of the proposed approach as compared with existing self-deployment schemes. The simulation results show that our scheme can achieve a higher coverage and enables less sensor movements in shorter times than self-deployment schemes using the traditional potential field.

Decentralized Sensor Selection for Large-Scale Multisensor-Multitarget Tracking

The problem of sensor resource management for multitarget tracking in decentralized tracking systems is considered here. Inexpensive sensors available today are used in large numbers to monitor wide surveillance regions. However, due to frequency, power, and bandwidth limitations, there is an upper limit on the number of sensors that can be used by a fusion center (FC) at any one time. The problem is then to select the sensor subsets to be used at each sampling time in order to optimize the tracking performance (i.e., maximize the tracking accuracy of existing tracks and detect new targets as quickly as possible) under the given constraints. The architecture considered in this paper is decentralized in which there is no central fusion center (CFC); each FC communicates only with the neighboring FCs, as a result communication is restricted. In such cases each FC has to decide which sensors should be used by itself at each sampling time by considering which sensors can be used by neighboring FCs. An efficient optimization-based algorithm is proposed here to address this problem in real time. Simulation results illustrating the performance of the proposed algorithms are also presented to support its efficiency. The novelty lies in the discrete optimization formulation for large-scale sensor selection in decentralized networks.

Sensor Resource Management driven by threat projection and priorities

This paper extends previous Sensor Resource Management (SRM) work by addressing information flow from sensor inputs to SRM, through four levels of the US DoD’s Joint Directors of Laboratories (JDL) sensor fusion model. The method flexibly adapts to several domains/problems. Human situation awareness information needs are linked to sensor control in a manner similar to perception management. The key to effective integration of JDL levels is the timely determination of the highest priorities via threat projection accomplished via Probabilistic Accumulative Situation Calculus (PASC), which quantifies threat intent using an appropriate level of automated context-based reasoning. The accuracy of the threat projection is improved over time using self-learning techniques. The multiple sensor system levels are unified primarily using the structure of quantified priorities. Algorithms are presented for a radar sensor resource allocation and adjustment method in which the dwell time per track parameter is the key radar sensor resource to be managed. A developed application of the method to an Integrated Air Defense System (IADS) sensor system problem is detailed, with simulation results shown to demonstrate the effectiveness of the method.