In situation analysis, an agent observing a scene receives information from heterogeneous sources of information including for example remote sensing devices, human reports and databases. The aim of this agent is to reach a certain level of awareness of the situation in order to make decisions. For the purpose of applications, this state of awareness can be conceived as a state of knowledge in the classical epistemic logic sense. Considering the logical connection between belief and knowledge, the challenge for the designer is to transform the raw, imprecise, conflictual and often paradoxical information received from the different sources into statements understandable by both man and machines. Situation analysis applications need frameworks general enough to take into account the different types of uncertainty and information present in the situation analysis context, doubled with a semantics allowing meaningful reasoning on situations. The aim of this chapter is to evaluate the capacity of neutrosophic logic and Dezert-Smarandache theory (DSmT) to cope with the ontological and epistemic problems of situation analysis.

We present a methodology to construct optimal visibility graphs from vector and raster terrain data based on the integration of Geographic Information Systems, computational geometry, and integer linear programming. In an emergency situation, the ability to observe an environment, completely or partially, is crucial when searching an area for survivors, missing persons, intruders or anomalies. We first analyze inter-visibility using computational geometry and GIS functions. Then, we optimize the visibility graphs by choosing vertices in a way to either maximize coverage with a given number of watchers or to minimize the number of watchers needed for full coverage.

A wireless sensor network (WSN) comprises a large number of wireless sensor nodes. Wireless sensor nodes are battery-powered devices with limited processing and transmission power. Therefore, energy consumption is a critical issue in system designs of WSNs. In-network data fusion and clustering have been shown to be effective techniques in reducing energy consumption in WSNs. However, clustering can introduce bottlenecks to a network, which causes extra delays in a data aggregation process. The problem will be more severe when in-network data fusion does not yield any size reduction in outgoing data. Such problems can be greatly alleviated by modifying the network structure. In this paper, a delay-aware network structure for WSNs with in-network data fusion is proposed. The proposed structure organizes sensor nodes into clusters of different sizes so that each cluster can communicate with the fusion center in an interleaved manner. An optimization process is proposed to optimize intra-cluster communication distance. Simulation results show that, when compared with other existing aggregation structures, the proposed network structure can reduce delays in data aggregation processes and keep the total energy consumption at low levels provided that data are only partially fusible.

The aim of this paper is to assess the impact of the uncertainty representation for retrieving items with missing data. The problem of information retrieval from incomplete incident databases is addressed in this paper. After a brief survey on the problem of missing data with an emphasis on the information retrieval application, we propose a novel approach for retrieving case records with missing data. The general idea of the proposed data driven approach is to model the uncertainty pertaining to this missing data. We chose the general model of belief functions as it encompasses as special cases both classical and probability models. Several uncertainty models are then compared based on (1) an expressiveness criterion (non-specificity or randomness) and (2) objective measures of performance typical to the Information Retrieval domain. The results are illustrated on several datasets and a simulation controlled missing data mechanism.

In this paper, we propose a novel pattern matching approach for vehicle identification based on belief functions. Distances are computed within a belief decision space rather than directly in the feature space as traditionally done. The main goal of the paper is to compare performances obtained when using several distances between belief functions recently introduced by the authors. Belief functions are modeled using the outputs of a set of modality-based-1-NN classifiers, two distinct uncertainty modeling techniques and are combined with Dempster's rule. Results are obtained on real data gathered from sensor nodes with 4 signal modalities and for 4 classes of vehicles (pedestrian, bicycle, car, truck). Main results show the importance of the uncertainty technique used and the interest of the proposed pattern matching approach in terms of performance and expressiveness.

Designing an optimal video camera network for monitoring ground activities is a critical problem to ensure sufficient performance in video surveillance system. Several authors have addressed this issue in the past, and considered this problem to be essentially a coverage optimization problem. Because of its nature and complexity, this problem is considered to be a multiobjective global optimization problem. In this paper we propose a basic concept for the design of a decision support tool for the deployment of C4ISR systems in general and Ground Air Traffic Control systems in particular. A novel and general formulation of coverage based on distance degraded ROC curves is proposed. This representation allows the user to define in a flexible as well as expressive way a wide variety of constraint combinations to build a sensor placement solution. Visualizations are provided to the user for a meaningful selection among the optimal solution computed. First results with the proposed approach are presented on a Ground Air Traffic Control case study aiming at the positioning of camera networks for the detection and localization of moving objects on runways.

The Support Vector Machine (SVM) is a very powerful technique for general pattern recognition purposes but its efficiency in practice relies on the optimal selection of hyper-parameters. A naïve or ad hoc choice of values for these can lead to poor performance in terms of generalization error and high complexity of the parameterized models obtained in terms of the number of support vectors identified. The task of searching for optimal hyper-parameters with respect to the aforementioned performance measures is the so-called SVM model selection problem. In this paper we propose a strategy to select optimal SVM models in a dynamic fashion in order to address this problem when knowledge about the environment is updated with new observations and previously parameterized models need to be re-evaluated, and in some cases discarded in favor of revised models. This strategy combines the power of swarm intelligence theory with the conventional grid search method in order to progressively identify and sort out potential solutions using dynamically updated training datasets. Experimental results demonstrate that the proposed method outperforms the traditional approaches tested against it, while saving considerable computational time.

Currently light infantry soldiers do not have access to many of their cyber resources the moment they depart the forward operating base (FOB). Commanders with recent combat experience have reported on the dearth of computing abilities once a mission is underway [14]. To address this, our group seeks to develop a tactical, mobile cloud implemented on a swarm of semi-autonomous robots. We provide two contributions with this work. First, provide a formal definition of the problem followed by a description of our approach to vulnerable state identification based on pattern recognition techniques. Second, we present an awareness definition as it pertains to our domain.

The aim of this paper is to detail further a Vehicle-Borne IED scenario proposed as an uncertainty modeling challenge to the information fusion community by the Evaluation of Techniques for Uncertainty Representation (ETUR) working group. This enrichment of the basic scenario is partly based on the careful comparison of formalizations published thus far by four uncertainty modeling experts as well as on the authors expertise in surveillance system design and risk assessment. The main additions reside in the exploitation of the temporal and spatial dimensions of the IED scenario initial statement. The authors show that the compromise between the expected risk, the time to certainty and time to intervene is central to the modeling of this very basic scenario and should be exploited further. According to the analysis of the formalizations of the VBIED scenario already published is seems also of interest to introduce the notion of agents to clarify the definition of state spaces. From the proposed model elements the authors expect that the scenario can be extended for more practical uses by allowing the addition of historical datasets from which a priori knowledge can be extracted, measurements be made from maps, and available resources balanced against expected risk.

: Currently light infantry soldiers do not have access to many of their cyber resources the moment they depart the forward operating base (FOB). Commanders with recent combat experience have reported on the dearth of computing abilities once a mission is underway. To address this, our group seeks to develop a tactical, mobile cloud implemented on a swarm of semi-autonomous robots. In this paper, we propose a pattern recognition approach to network vulnerability assessment applied to a tactical swarm of robots to enhance their strategy for surveillance coverage. Our work enhances network-enabled persistent surveillance within a dynamic, mobile domain via the implementation of sensor awareness concepts.

A Wireless Sensor Networks (WSN) usually consists of numerous wireless devices deployed in a region of interest, each able to collect and process environmental information and communicate with neighbouring devices. The problem of sensor deployment becomes non-trivial when we consider environmental factors, such as terrain elevation. We propose a Crowd-Out Dominance search (CODS) that makes use of topographic terrain information and inter-sensors relationship information to facilitate the search of the best sensor positions. The proposed method demonstrates better robustness to terrain complexity compared with traditional heuristic methods.

This paper addresses the general problem of bridging the gap between sensor output and the structured reports needed and handled by end users in several practical applications. Sensors perform measurements used to characterize world entities according to their physical features, such as weight or speed or even more abstract features extracted from signal statistics such as spectral higher-order moments. The sensor outputs are usually provided to a human operator, allowing him to build, communicate and share with other humans a picture of a given situation. However in practice, humans often handle entities by considering their functions as it is the case in military intelligence reports rather than limiting their attention to physical or statistical properties. A possible solution to such a mismatch is to augment appropriately the vocabulary used to deliver the sensor outputs, allowing the operators to manipulate it efficiently in practical applications such as military intelligence analysis and dissemination. This paper presents a semantic approach developed for relating two complementary descriptions of data, one built upon experimental data collected using the SASNet system, the other being a military intelligence ontology called ONTO-CIF. The proposed method is based on ontology mapping and allows us to augment sensor data in order to cope with user needs.

The primary concern of a bio-surveillance system is to analyze and interpret data as they are collected and then decide whether further investigation is required. Decision makers need to know whether the data in the current test interval are sufficiently different from expected counts to cause an alert. Despite the fact that a number of detection methods have been proposed, we notice in the literature the users of current systems can still experience extremely high false alarm rate. We propose a novel measure that takes into account both anomaly magnitude and anomaly frequencies for bio-surveillance, and experimental results show that the proposed measure performs better than conventional measures for bio-surveillance.

A fundamental problem when performing incremental learning is that the best set of a classification system's parameters can change with the evolution of the data. Consequently, unless the system self-adapts to such changes, it will become obsolete, even if the application environment seems to be static. To address this problem, we propose a dynamic optimization approach in this paper that performs incremental learning in an adaptive fashion by tracking, evolving, and combining optimum hypotheses overtime. The approach incorporates various theories, such as dynamic particle swarm optimization, incremental support vector machine classifiers, change detection, and dynamic ensemble selection based on classifiers' confidence levels. Experiments carried out on synthetic and real-world databases demonstrate that the proposed approach actually outperforms the classification methods often used in incremental learning scenarios. (C) 2011 Wiley Periodicals, Inc.

: When positioning a number of static or mobile agents (human or autonomous) into an area for the purpose of surveillance, the positioning and motion strategies assigned to these agents impact on the situation awareness gained. This paper presents a toolbox that evaluates motion and placement strategies in a realistic surveillance context, using the Interpreted Systems semantics. Its five components implement relevant theoretical concepts: simulation of neutral and opposing forces? behavior, geometrical discretization of the Area of Interest, state space generation, state space analysis and visualization.