Modern sensors, such as multifunctional radars, comprise many settings and the number of controllable settings is increasing due to technological advance. Although having many settings allows many capabilities, it simultaneously requires an automatic manager to control them. This paper proposes mission-driven resource management to control reconfigurable sensing systems during run time. It works as follows. First, expected mission success is defined from an end-user point of view. Because such criterion can eventually be mathematically linked to the adaptable parameters of the reconfigurable systems, these parameters can be optimally selected. To illustrate, a case study is considered where many heterogeneous high-level operational tasks (i.e., air defense, weather alarm, crowd control, and drone flight) have to be supported by several fully reconfigurable radio-frequency antenna front ends. This paper analytically and numerically compares the proposed mission-driven method that maximizes end-user’s expectations for mission success, with a traditional task-driven one, that solely optimizes task performance characteristics. To conclude, by maximizing expected mission success, the systems are automatically adapted to new circumstances, and the end-user’s mission is constantly supported by the system in the most effective way. @en

This paper proposes feature extraction methods for object classification with passive acoustic sensor networks deployed in suburban environments. We analyzed the emitted acoustic signals of three object classes: 1) guns (muzzle blast); 2) vehicles (running piston engine); and 3) pedestrians (several footsteps). Based on the conducted analysis, methods are developed to extract the features that are related to the physical nature of the objects. In addition, a time-based location method is developed (based on a pseudo-matched-filter), because the object location is required for one of the feature extraction methods. As a result, we developed a proof-of-concept system to record and extract discriminative acoustic features. The performance of the features and the final classification are assessed with real measured data of the three object classes within suburban environment.@en

The spectrum of threats imposed upon modern society has drastically widened in the last decades. One of the required security means is sensor systems. There exist many types of sensors and their capabilities vary, but some of them are reconfigurable and able to provide many functions. Such sensors provide at different time instances a wide range of options. These options result in too many choices for humans to consider and they are usually defined in terms that are too strongly focused on complex aspects of sensor technology. Therefore, an automatic tool is required that can control such systems. However, there remain various fundamental challenges in the field of resource allocation in general and for sensor management specifically. The majority of current solutions is driven by the idea that execution of system tasks and/or sensing characteristics have to be optimized. A better option is to develop and configure systems in such a way that they contribute as much as possible to the end-user's mission. This is obtained with the following cross-disciplinary approach. The strategy of top-down thinking is used to first investigate what various types of end-users consider important, the hypothesis of expected-utility is used to describe mission success expectations, and the observe, orient, decide, and act loop is used to determine on a higher level the required functions and then how sensor technology can contribute to this. The reasoning of falsifiability is used to verify (i.e. self-critical) the concepts and develop supplementary solutions. The ratio between the expected success of the mission when using the developed approach and this expectation when using traditional methods is greater than or equal to 1. Thus, the performance cannot be worse, is at least equal, and potentially much better. The resource management solutions are demonstrated on a high operational level, because this allows to create a context in which the optimization can be discussed with end-users. End-users found the developed resource management solution very applicable for allocating resources during fictional deployment and operational phases. Because of their vivid interest, many focused questions, comments and compliments confirming the added value of the mission-driven approach, it can be concluded that the developed solution matches their operational understanding and needs. The developed mission-driven resource management solution directly defines the end-user's mission as the optimization objective for reconfigurable sensing systems. As a result, the process is not driven by technical characteristics, lookup tables, rules-of-thumb, task priorities and/or artificial quality measures, but by mission success. Any decision (e.g. trade-offs, graceful degradation) can be made based on a single objective function, making the optimization clear and transparent. The result of permanent maximization of mission success expectations is that systems adapt automatically, quickly and accurately to fast changing missions, environments and threats. @en