Browse Publications
2010 Reports
Note: Unless linked to the full text, reports are only available to NATO member nations from designated distribution centres.
Countermeasure effectiveness: how much protection is enough?. Kessel, R.T. NURC-FR-2010-016. November 2010
Among the foremost system requirements for protective countermeasures is their overall effectiveness. In practice, effectiveness is limited by socioeconomic and technical feasibility, forcing security providers to settle for admittedly imperfect countermeasures. The question, then, is how much countermeasure effectiveness is enough? The answer would constitute a minimum requirement for countermeasures to achieve. It is addressed from two vantage points: 1) risk analysis methodologies and 2) the attacker's perspective, with emphasis on the latter owing to the insight it provides. It is shown with examples (9/11, WMD, suicide attack, Mumbai) how the rational pursuit of objectives by the attacker imposes an upper bound on the risk that the mastermind behind an attack is willing to tolerate, which in turn constitutes a lower bound on the effectiveness of countermeasures when deterrence is taken into account. It is shown how low- to moderate-performance countermeasures can be effective owing to deterrence, and how the perspective adopted toward a countermeasure -whether from a defender's risk analysis, the attacker's expectation of success, or the audience at a security demonstration- changes the countermeasure effectiveness requirements. The prospect of 'over deterrence' is also considered. Practical implications are drawn throughout.
Numerical modelling of wave-current interactions in the Sea of Marmara. Chiggiato, J.; Rixen, M.; Besiktepe, S. NURC-FR-2010-015. October 2010
The synchronous 2-way coupled modelling system ROMS/SWAN has been configured in a realistic application in order to test wave-current interaction parameterizations for use in ocean modelling. Inclusion of these interactions is expected to improve model performance and representation of the dynamics of the upper ocean. The case study discussed here is an indirect assessment of the impact of these parameterizations in model simulations using available observations. The TSS09 trial provided the opportunity to test this modelling system thanks to the concurrent measurement of wind field, wave field and upper circulation as seen by surface lagrangian drifters. All these data have been used for model validation. The numerical exercise focuses on a selected windstorm during TSS09. Model results from the control run (i.e., no wave current interactions) show a fairly good agreement with known Marmara Sea hydrography and available observations. The inclusion of wave-current interactions physics weakly impacts on model results, without showing overall improvements, even if model results do show sensitivity to the inclusion of such parameterizations. The Marmara Sea is a very challenging natural laboratory, because of its limited size, the need for very high meteorological forcing and the impact of the outflow of the straits in the general hydrodynamics. This makes very difficulty to assess wave-current interactions physics implemented in the models because of the major impact of uncertainties associated to other driving mechanisms in the area.
Implementation of the two-way coupled ocean-atmosphere COAWST model in the Marmara Sea. Chiggiato, J.
NURC-FR-2010-014. October 2010
The Coupled Ocean Atmosphere Wave Sediment Transport model COAWST has been implemented in the Marmara Sea and undergone a preliminary assessment. In this first application at NURC the atmospheric and oceanic parts have been configured to run with 1-way coupling and 2-way coupling. The system has been applied to simulate a windstorm of north-easterlies over the Marmara Sea occurred during the TSS09 cruise (February 2009). Model results show the central role of the sea surface temperature in ocean-atmosphere coupled modeling. The performance of the two-way coupled system is generally better than the performance of the one-way coupled system, however some inconsistencies are still arising from the coupling procedure associated to the estimate of fluxes. In order to have a local closure of the system, the air-sea exchange of heat should be included in the coupling procedure in addition to the sea surface temperature.
Maritime Surveillance 2009 (MS09) Experiment Post-Analysis Workshop. Ranelli, P.; Bryan, K. NURC-CP-2010-001. September 2010.
This workshop report briefly overviews the meeting presentations and lists major findings and recommendations from the two-day event. Eighteen participants attended the meeting, including representatives from NURC, MOC Naples, the Turkish Navy, ACT NATO, and the European Space Agency (ESA). The following aspects of the experiment were addressed: * Results of the operational analysis; * Network architecture and data collection; * Moc Operations; * Multi-sensor fusion; * Anomaly detection; * Analysis of AIS intermittency; * SAR ship detection. Further, an overview on the recent HF radar experiment in the Ligurian Sea was given. A CD-ROM containing the workshop presentations can be requested from library@nurc.nato.int.
Transitioning the Multistatic Tactical Planning Aid (MSTPA) towards decision support software. Strode, C. NURC-FR-2010-013. September 2010.
The Multistatic Tactical Planning Aid is a tool currently in development at NURC which may be used to model the performance of a given multistatic sensor network in terms of the probability of detection of a threat, the ability to hold a track, and whether such a track could be correctly classi ed as a threat. The tool therefore considers the entire chain of events from an initial calculation of signal excess, the generation of a contact considering localization errors, followed by the subsequent tracking and classi cation process. In its current form the tool may be used to plan a particular multistatic scenario through operational analysis of many Monte Carlo simulations. The future development of MSTPA will transition towards a real time decision support tool to assist operators and planners at sea. This study investigates a number of generic decision support techniques which may be wrapped around the MSTPA tool. In essence, MSTPA becomes a central metric calculator within a greater scheme of decision support processes. These processes include data mining, game theory, and optimisation techniques. The study considers each in turn and provides examples as to how they may be applied to tactical multistatic sonar scenarios.
Improved meteorological data for naval fire support. Schmidt, R.; Rixen, M. NURC-FR-2010-012. August 2010.
There is an enduring requirement for the engagement of land targets from naval guns. Although precision gun munitions are being developed, conventional munitions will be used for many missions, particularly suppression. However, there is also a need to reduce the risk of collateral damage. One of the principal causes of inaccurate fire is poor meteorological data – a problem that has increasingly significant effects as the range of conventional gun munitions increases. This study describes R&T developments of recent years and presents future gun fire support proposals for the maritime domain by recommending methodologies to improve deliverables.
Implications of behavioural rules on autonomous vehicle (AUV) trajectories: some examples of their limiting behaviour. Harrison, C. NURC-MR-2010-002. June 2010.
Some unexpected AUV behaviour during the experiment CLUTTER'09 was caused by its navigation algorithm attempting to follow waypoints arranged in a horizontal circle but with each way point being prematurely timed out. This leads to a 'pursuit' problem where the AUV track, that ought to be a circle, becomes a looping cycloid- or spiral-like curve. This report investigates the curve first in the Earth's frame then in the frame of the effectively moving pursued waypoint. In the former frame the curve overlaps itself many times, finally converging on a smaller circle. In the latter frame the curve is a true spiral that converges on a point and never overlaps itself. Differential equations for these spirals are found in three polar coordinate systems: 1) centred on the circle centre; 2) centred on the waypoint; 3) centred on the final convergence point. Although the solutions are a hybrid of the logarithmic spiral and those formed of piece-wise circular arcs (Golden Section), no analytical solution has been found.
Exploratory Research Final Report: Advanced Data Fusion. Coraluppi, S.; Carthel, C. NURC-FR-2010-011. June 2010.
Text Track-oriented multiple-hypothesis tracking (MHT) is arguably the most powerful current methodology in multi-sensor multi-target tracking. In this report, we provide an accessible introduction to this methodology, and discuss our result on some of the challenges in tracking theory, namely the target-death problem and the greedy-target problem. Though centralized MHT often works well, there are robustness challenges and fundamental performance limitations. These limitations are observed when dealing with difficult surveillance challenges that include multi-scale data, slowly-crossing targets, large sensor networks, dim targets, heavy clutter, unresolved measurements, and fragmented track data. Centralized MHT is challenged due to its inability to maintain a rich enough set of track hypotheses over long time intervals. Distributed MHT provides a straight-forward methodology to advanced hypothesis management that provides a means to address these challenges. This report describes our application of the distributed MHT paradigm to synthetic and experimental data in undersea and maritime surveillance settings.
Synthetic Aperture Radar Wind Field Retrieval. Horstmann, J.; Falchetti, S. NURC-FR-2010-009. May 2010.
In this report the latest methodologies for wind field retrieval from SAR images acquired at C-band in vertical and horizontal polarization are introduced. The methodologies are applied to several SAR data to show the possibilities as well as remaining open questions regarding SAR wind retrieval. In addition new geophysical model functions (GMFs) for X-band and L-band SAR data are introduced. Preliminary investigations of several X-band SAR data acquired by the German satellite TerraSAR-X, shows that the GMF is well suited for high resolution X-band SAR wind field retrieval for both horizontal and vertical polarization. Investigation of several Alos PalSAR data showed too large uncertainties in the normalized radar cross section (NRCS), as well as other SAR processing artifacts, which did not allow to test the L-band GMF sufficiently. Furthermore, several SAR data products have been investigated with respect to certainty of the NRCS, incidence angle and other image parameters that are relevant for SAR wind retrieval. Comparison of different SAR processors applied to the same Envisat ASAR raw data lead to significant differences, which affect the SAR wind retrieval. Therefore, great care has to be taken when selecting SAR data as well as the best suited SAR processor. Last but not least a sensitivity study was performed for C-band SAR with respect to wind speed retrieval. The effect of uncertainties in incidence angle, wind direction and NRCS was investigated for the GMFs CmodIfr2 and Cmod5 with respect to wind speed, wind direction and incidence angle. This study shows that there is a strong dependence on certainty of a SAR retrieved wind speed on incidence angle, wind speed and wind direction. These investigations suggest to include a uncertainty estimate in future SAR retrieved wind speeds.
MSA experimentation 2009. Final analysis report. Bryan, K.; Ranelli, P.; Lewis, C.; Coraluppi, S.; Carthel, C.; Grasso, R.; Baldacci, A.; Neumayer, J. (ACT). NURC-FR-2010-010. May 2010.
The Maritime Situational Awareness experiment (MS09) took place 17-22 October in the Eastern Mediterranean. MS09 was the combination of two NURC-led experiments executed for ACT which examined the utility of scientific developments at NURC to increase operators’ situational awareness in the Maritime Operations Centre at MCC-Naples. The first experiment concerned multi-sensor data fusion and anomaly detection and the second the ability to predict coverage of an Automatic Identification System (AIS) network based on a receiver locations and meteorological data. This report examines and discusses the analysis objectives of the two experiments.
Rapid Contact Designation and Warning for Port Protection: EPOW Quick-Look Report. Kessel, R.; Pastore, T. NURC-MR-2010-001. March 2010.
Key elements of a new concept for automatic threat designation and warning for port protection were exercised in military experimentation funded by NATO ACT EPOW 2009. This is apparently the first experiment with the automatic control and activation of non-lethal first response against small boats in maritime security, occasioned on the one hand by the long-standing vulnerability of expeditionary NATO forces and shipments to terrorist attack while in foreign ports, and on the other hand by the likely constraints on response; namely, the (*) duty to warn, prove hostile intent, and use proportional force; (*) short response time available to security forces; (*) inevitable lapses human vigilance and readiness during long uneventful watches; (*) risk of collateral damage when using lethal force. Experimentation was designed by NURC and executed with support by the ITA Navy, from 30-Nov-2009 to 15-Dec-2009, in La Spezia. A first look at the experimentation and its results is given here as required by ACT.
A pilot 3D super-ensemble algorithm: evaluation using LSCV08 data. Mourre, B.; Lenartz, F.; Rixen, M. NURC-FR-2010-008. March 2010.
The 3D super-ensemble technique (3DSE) is proposed to improve short-term ocean state forecast skills compared to individual ocean models and their Ensemble Mean. The three-dimensional spatial variability of model weights is the innovative aspect of the 3DSE compared to previous implementations of super-ensemble methods. The predictive skills are evaluated using outputs from three ocean circulation models and data from the LSCV08 oceanographic campaign carried out in the Ligurian Sea in October 2008. The 3DSE forecast is shown not only to significantly reduce the bias against observations during a 96-hour simulation period, but also to improve other pattern-related forecast skills. Sensitivity studies show that (i) the weight error covariance matrix has to be properly initialized, preferably using model-based correlations, (ii) the consideration of an additional constant model does not bring any significant improvement in the present example.
Atmospheric corrections algorithms for high resolution remote sensing (SPOW 2009 document). Trees, C.; Pennucci, G. NURC-FR-2010-007. March 2010.
Airplane and satellite remote sensing is a very useful observational tool for providing operational information on environment properties over large areas. The maritime zone is a highly dynamic region where hydrodynamic and morphodynamic processes change on a wide range of spatial and temporal scales. For those reasons, remote sensing of the coastal zone is very challenging and it needs improvement. The experiments LSCV08 (Ligurian Sea Cal-Val 2008) and BP09 (Battlespace Preparation 2009) addressed specific problems associated with remote sensing (RS) of coastal littorals, specifically to improve the quality of the optical properties derived from RS in coastal environments. The main objective of the experiment was to assist in the calibration and validation of large-scale ocean color sensors (MODIS, 1 km), medium scale sensors (MERIS, 300m), and small scale sensors (GeoEye-1 and hyperspectral airplane over-flights, 1 and 2 meters, respectively) using in situ measurements. During this study, the integration of remote sensing and in water sensors was tested to provide on-scene near shore battlespace characterization (in spatial and spectral dimensions). In the present report we first focus on the atmospheric correction and then on the integration of RS data with in situ measurements to test optical models. The objective is to tune models of the Ligurian Sea, in particular in the near shore areas.
Non-Lethal Response Technologies for Port Protection: 2009 Survey Report and Options. Kessel, R. NURC-FR-2010-006. February 2010.
In defensive security measures during operations other than war, low levels of force may be required to meet obligations to warn, prove hostile intent, and use proportional force. Non-lethal response technologies are therefore being considered for use in port protection, especially for the enforcement of security exclusion zones above and under the water in the vicinity of protected assets, to counter underwater intruders and small boats. Concepts for non-lethal capabilities and technology readiness assessment are given here, followed by a review of selected response technologies for port protection, including an underwater air gun and sparker, marine mammals, above and underwater loud hailing devices, dazzlers, boat entanglement barriers, and projected energy. The intent was familiarization with new technologies that could be operational within 3 to 5 years, and identification of gaps where further research is required to advance them toward fielded capability. The review and context given here should be of interest to security providers, technology developers, system integrators, and researchers.
Model based decision support for Glider operations. Grasso, R.; Cecchi, D.; Trees, C.; Rixen, M.; Alvarez, A.; Strode, C. NURC-FR-2010-005. March 2010.
A technique to forecast the glider horizontal state position in a given current random field, based on the unscented transform (UT) and the Unscented Kalman Filter (UKF), is detailed and tested by using a reference validation data set collected during the Ligurian Sea Calibration Validation 2008 (LSCV08) experiment. This technique makes use of a two dimensional kinematic glider model having stochastic sea water current field inputs. The depth averaged two dimensional current field is modeled as a bi-variate Gaussian random vector with given mean and covariance matrix. An ensemble technique is used to provide an estimate of the current mean and covariance at each prediction step given a set of different current forecasts. The comparison with reference glider tracks from LSCV08 confirms that the proposed methodology is able to predict the second order statistics of the glider horizontal position so that the reference tracks fall within the 95% error ellipse of the predicted state. The algorithm is the base for a series of applications in the field of decision support systems for glider operation planning and risk assessment. To this purpose, the report provides an example of how to build a risk measure to monitor the position of the glider which has to be constrained to operate within a given allowed area.
The OEX SLITA Signal Processing Chain: Current Progress and Issues. Hamilton, M. NURC-FR-2010-004. February 2010.
This report summarizes research and work performed from September 2009 through September 2010 at NURC on the \real-time" towed array processing system for the SLITA array installed on the OEX AUV. In particular the system used aboard the OEX in the GLINT09 sea trial will be documented, as well as subsequent improvements. In this technical report some low level details of the source code and computer science approaches to the development are documented. In general most of the explicit code references have been moved into the appendices. Particular pieces of software developed at NURC and MIT, and explicit modifications to them, are documented there for the use of current developers working on the Cooperative ASW program.
Anomaly detection during the MS09 experiment . Baldacci, A.; Cappelletti, M.; Coraluppi, S.; Carthel, C.; Bryan, K.; Arcieri, G.; Ranelli, P. NURC-FR-2010-001. January 2010.
The Maritime Surveillance '09 (MS09) sea trial was the first NURC sea trial fully dedicated to maritime surveillance. During the experiment, for the first time, the modules of the Maritime Surveillance System (MSS) were integrated and applied to real data feeds. In order to evaluate the performance of the MSS in real-life situations, a number of scenarios were designed and executed. Two supporting ships acted as cooperative targets. This report focuses on anomaly detection (AD). The AD modules tested during the experiment are: (a) AIS transponder anomalies; (b) kinematic anomalies (radical changes in speed or course, offshore stops); (c) multiple-target joint anomalies (rendez-vous at sea). All the experiment goals were met. The positive aspects of the experiment were: (a) the successful execution of algorithms and positive first-look impression on results (good probability of detection, moderate false alarm rate); (b) the successful integration of the AD part with the other modules of the MSS; (c) the collection of a rich dataset. The negative aspects were: (a) estimation of data-driven AIS coverage from NRV Alliance was challenging due to very sparse data; (b) frequent drop-outs in AIS data significantly affected performance of the on-off anomaly detector; (c) post-trial performance assessment was limited due to problems in ground-truth reconstruction.
Analysis of AIS intermittency and vessel characterization using a Hidden Markov Model. Guerriero, M.; Coraluppi, S.; Carthel, C. NURC-FR-2010-002. January 2010
In this report we perform a statistical analysis of the Automated Identification System (AIS) communication channel. We base our study on a Hidden Markov Model (HMM). We model the transition between different states of the channel as a Markov Chain (MC); the probability that a message sent by a AIS transmitter will be lost is associated to each state. The MC is not directly observed, but the received time stamps of the AIS reports provide some statistical information about the current state of the channel as well as some information about the parameters of the model. Additionally, the statistic characteristics of the AIS channel are used in a batch anomaly detection algorithm that characterizes vessel as anomalous if their (hidden) transponder state is estimated to be in the off state for too high a fraction of the surveillance time.
Real-time wide-area tracking with intermittent multi-sensor data. Carthel, C.; Coraluppi, S.; Bryan, K.; Arcieri, G. NURC-FR-2010-003. January 2010.
This report describes the data fusion algorithms that constitute the core functionality of the NURC Maritime Surveillance System (MSS). Key features of this capability include modular, multi-hypothesis tracking logic as well as wide-area track fusion logic amenable to active (scan-based) and passive (time-series) data streams from cooperative and non-cooperative sensor systems. We illustrate the performance and robustness of the MSS with simulated sensor data. Additionally, we provide performance results for real-time at-sea processing during a recent NURC sea trial, Maritime Surveillance 09 (MS09).