Browse Publications
2007 Reports
Note: Unless linked to the full text, reports are only available to NATO member nations from designated distribution centres.
The NURC Distributed Multi-Hypothesis Tracker (DMHT): Scientific User Documentation. Carthel, C.; Coraluppi, S. NURC-FR-2007-025. December 2007.
NURC automatic tracker technology has been developed over the last several years, with a focus on multistatic active sonar. The technology is applicable to other problem domains, including maritime surveillance. The DMHT (distributed multi-hypothesis tracker) provides an essential component in the signal and information processing chain that yields a surveillance picture based on multistatic active sonar assets, including both fixed sonobuoys and mobile platforms. The DMHT combines high track detection and localization performance, computational efficiency, and architectural flexibility. This report provides a quick reference guide to using the DMHT. Further, it provides a complete, annotated listing of all DMHT technical documentation, containing detailed algorithmic descriptions and performance results.
Simulation of trial DEMUS04 using the multistatic tactical planning aid (MSTPA). Strode, C.; Wathelet, A.; Vermeij, A. NURC-FR-2007-021. December 2007.
The multistatic tactical planning aid (MSTPA) is a simulation tool that aims to provide a fast assessment of multistatic sonar performance. This report describes a comparative study between an at sea multistatic active sonar experiment, DEMUS04, and the equivalent experiment simulated using MSTPA. The results are reported and serve to validate MSTPA prior to its use in multistatic tactical development.
Doppler-aided multistatic sonar tracking. Coraluppi, S.; Carthel, C.; Hughes, D.; Micheli, M.. NURC-FR-2007-024. November 2007.
This report extends the NURC distributed multi-hypothesis tracker (DMHT) technology for multistatic active sonar data to include Doppler-sensitive processing based on continuous waveform (CW) contacts. This enhancement requires extensions to our past measurement modelling that accounts for system and environmental errors, as well as extensions to our nonlinear recursive filter design. Testing and validation of the enhanced DMHT, with a centralized processing architecture, is performed with simulated multistatic data as well as with sea trial data from the recent pre-DEMUS’06 trial, which involved a synthetic target (echo repeater).
An active acoustic tripwire for simultaneous detection and localization of multiple underwater intruders. Folegot, T.; Martinelli, G.; Guerrini, P.; Stevenson, J.M.. NURC-FR-2007-023. November 2007.
An algorithm allowing simultaneous detection and localization of multiple submerged targets crossing an acoustic tripwire and based on forward scattering is proposed and has been demonstrated at sea. The paper shows an excellent agreement between the theoretical performances and the processing of the data gathered at sea for crossings at several depth and ranges. Targets crossing the acoustic field produce shadows on the receivers of the barrier, for specific sensors and for specific acoustic paths. In post-processing, a model is involved to assign the expected sound paths to the observed shadow. This process gives a triangulation of the target’s range and depth inside the acoustic field. The quality of the localization is achieved taking advantage of the multi path propagation structure of the transmitted signal, together with the diversity of the source and receiver locations. Due to the use of an array of sources, the robustness of the method for various environmental configurations is demonstrated by simulation.
Unmanned surface vehicles for harbour protection. Pastore, T.J. NURC-FR-2007-022. November 2007.
NURC has conducted a technology survey of Unmanned Surface Vehicles (USVs) as part of the Scientific Programme of Work in Harbour Protection. The assessment was based on vendor information, trade publications, scientific journals, and personal contacts with developers. The report includes an extensive bibliography and a tabular presentation of the characteristics of approximately fifty USVs. The utility of USVs in the context of Harbour Protection missions was assessed, and conclusions were drawn based on results reported to date in the field.
MCM AUV autonomy: requirements, approach and system implementation. Evans, B.S.; Williams, D.P. NURC-FR-2007-020. October 2007.
This report documents the first step in examining the application of intelligent decisional autonomy to MCM AUVs. The report examines and describes decisional autonomy requirements for MCM use and identifies a behavioural-based approach as being the most suitable approach for delivering effective decisional autonomy for an MCM AUV. A design for a decisional autonomy system is defined and the extent to which the design matches the MCM AUV decisional autonomy requirement is reviewed.
Countering underwater intruders: ITA MINEX 2007 NATO force-protection exercise. Kessel, R.T. NURC-FR-2007-019. October 2007.
Capability against underwater intruders in ports requires surveillance of the underwater domain and the means for response when intruders are detected. The Italian MINEX 2007 (March 2007) exercise was an opportunity to experiment for the first time with a side-looking sonar for the reacquisition and classification of an intruder by a response team vectored into position by an large-area surveillance sonar. The results, analysis, and implications of the exercise are reported here. It was found that one could 1) reacquire an underwater contact with ease (i.e., high probability and high confidence); 2) classify a contact as a diver with ease if the diver was using open-circuit breathing equipment; and 3) classify a contact as a diver with some difficulty (moderate confidence) if the diver was wearing closed-circuit breathing equipment. Mission-relevant probabilities and times windows were estimated. The advantage of side-looking sonar over forward-looking was considered. The exercise prepares the methodology to be applied in tests with other sonars, side-looking and forward, in upcoming trials at NURC.
SUPREMO operation at sea during BASE'07. Prior, M.K. NURC-FR-2007-018. October 2007.
The operation of the SUPREMO sonar performance prediction model, during a demonstration of an environmentally adaptive sonar system, is described. The way in which the model uses environmental and sonar data to predict optimum sonar settings is described. Predictions made during the BASE 07 experiment are detailed and the impact on target detectability of using these settings is described. It is shown that adapting pulse bandwidth and centre frequency leads to a reduction in sonar clutter. This, in turn, reduces the number of false tracks produced by the system’s tracker.
Environmentally adapted active sonar tracking and multi-sensor integration for maritime surveillance in BASE’07. Coraluppi, S.; Carthel, C.; Haralabus, G. NURC-FR-2007-017. October 2007.
Fusion and tracking are essential components in effective surveillance technology. Here, we apply the NURC distributed multi-hypothesis tracker (DMHT) to contact-level data that results from interleaved standard and environmentally adapted LFM pulses. Significantly improved tracking performance was achieved with real-time system adaptation during BASE’07. Additionally, surface data acquired during BASE’07 (AIS, ship-based radar, and satellite-based imagery) demonstrates the potential for multi-sensor maritime surveillance and for improved undersea surveillance.
Analysis of model-based automatic target recognition using the statistical theory of shape. Myers, V.; Couillard, M. NURC-FR-2007-016. August 2007.
This report investigates the use of shape theory in automated target recognition (ATR). The acoustic shadows cast by proud or partially buried objects on the seafloor are compared to theoretical shadows obtained with a ray tracing model. The shape of these shadows is treated as a mathematical object and statistical tools are used to extract information from a large dataset obtained during the Citadel sea trial in October 2005. Results are obtained for ATR with nearest neighbour and support vector machine classifiers using a feature which is the Procrustes shape similarity distance between a measured shadow and one obtained by the ray-trace model. The Hausdorff and Hamming distance measures are added as features to improve classification performance. Then, some hypothesis testing using the multivariate T2 test is performed to analyse populations of shapes of the shadows of truncated cones and rocks. The tests show that the mean measured rock and truncated cone shapes are statistically different from each other; however the same is true for the measured and modelled populations of cones . This multivariate test also underlines the inherent difficulties of comparing real images to modelled ones.
MX 04 Trial Report: Black Sea Port & Harbour Protection with Autonomous Underwater Vehicles. Yip, H.; Baralli, F.; Bovio, E.; Cernich, F.; Reed, S.; Rehr, B. NURC-FR-2007-014. August 2007.
This report documents the activities and results from a Military Experimentation that took place in the harbours of Varna and Burgas, Bulgaria. The experimentation was funded by Allied Command Transformation (ACT), and conducted in partnership with the Bulgarian Navy and the Bulgarian Institute ofMetal Sciences. The aim of the experiment was to demonstrate and assess the effectiveness of Commercial-Off-The-Shelf (COTS) Autonomous Underwater Vehicles (AUV) combined with Computer Aided Detection (CAD) and related technologies to counter terrorist placement of UnderWater Improvised Explosive Devices (UWIED) in ports and harbours.
NATO Harbour Protection Trials 2006 (HPT06) Analyst Report: Harbour Surveillance Systems. Kessel, R.T. NURC-FR-2007-004. August 2007.
NATO’s Harbour Protection Trials (HPT06), La Spezia, 3-7 April 2006, demonstrated systems for hull search, underwater pier inspection, mine and improvised explosive device (IED) detection, and harbour surveillance. An analyst team lead by NURC was struck to assess in particular the technology readiness of the systems. This report is the first of several planned, giving introductory background to HPT06 and analyzing the six underwater surveillance systems demonstrated in the trials. Three technology groups were represented in this portion of the demonstrations: active sonar (4 systems), passive sonar (1), and one passive magnetic “trip-wire” (1). The systems were installed near the end of a pier in the inner harbour. Divers staged simulated intrusions toward each system over the course of two days. A NATO Technology Readiness Level was assigned to each, and performance was assessed in terms of the probability of detection insofar as it could be determined from the observations. The methodology and results are described.
MX 05 Trial Report: MCM with Autonomous Underwater Vehicles & the Standing NATO MCM Group 1 in Open Spirit 06. Yip, H.; Baralli, F.; Bovio, E.; Cernich, F.; Reed, S.; Rehr, B. NURC-FR-2007-003. August 2007.
This report documents the activities and results from a military experimentation that took place in OPEN SPIRIT 06, Estonia. Allied Command Transformation (ACT) funded this experimentation. The NATO Research Centre conducted the experimentation in partnership with the Standing NATO MCM Group 1, the Estonian Navy, and ACT. The aims of the experimentation were to assess the effectiveness of contact acquisition with multiple-look analysis on the MCM efficiency, and to demonstrate the effectiveness of COTS AUV used in conjunction with NATO MCMV in an MCM environment with historical ordnances.
Detection of NATO Classified Information Leaks: A Proof-of-concept for Counter-intelligence. Wathelet, A.; Leonard, M. NURC-MR-2007-003. July 2007.
This report documents a brief study of a proof-of-concept to detect unauthorised release of NATO classified information. The approach uses the information content of emails and machine learning algorithms. Accuracies of various algorithms in categorising emails into NATO Unclassified and NATO Restricted classes are reported. Recommendations to inform decisions regarding future research and experimentation proposals as well as broader decisions on improving NATO information security are reported.
An example of integrated decision support for amphibious landing operations (demonstrated during Steadfast Jaguar 06 experimentation). Grasso, R.; Ranelli, P.; Schmidt, R.; Teixeira, J.; Conley, D.; Pennucci, G.; Fabiani, A.; Giannecchini, S. NURC-MR-2007-001. July 2007.
Expeditionary Operations need Environmental Information in particular when operating in an area with denied access due to political or military reasons. To provide support for decision makers this paper describes a new architecture to obtain Environmental Information (EI) by fusing several existing high resolution forecast models and determining parameters of atmosphere, ocean and hydrography and combining to a Tactical Decision Aid (TDA). The several components of the decision support architecture will be described and discussed and demonstrated taking an example for an amphibious landing operation in a denied or unknown area. The research was performed during Steadfast Jaguar Experimentation Exercise (SFJG 06 XREP) in June 2006.
Remote deployment of NURC Ground Station (2-16 November 2005). McCarthy, E.; Stenvoll, R.; Turgutcan, F.; Ranelli, P. NURC-FR-2007-013. July 2007.
The NURC Remote Sensing Ground Station was deployed over a 2 week period in November 2005. The antenna was dismantled at the NURC site and transported by flatbed truck to a remote site at an Italian military installation in Aulla, Italy where it was re-assembled. Within 24 hours of re-assembly, it was fully operational. All scheduled passes were successfully down linked. Over 20 passive and active images were received in real-time from several satellites including RADARSAT, TERRA, AQUA and ENVISAT. The Ground Station operated without event from 10-16 November. It was then dismantled within 24 hours and re-assembled at NURC. The remote deployment of such large equipment (the antenna is 5.4 meters in diameter) requires a significant amount of planning. To this end, this report details for potential users the effort required to remotely deploy the ground station, the logistics and support required, as well as the processes. This includes site preparation, power requirements, environmental conditions, and communications necessary.
SUPREMO v.2.0 program description. Prior, M.K. NURC-FR-2007-011. June 2007.
A program description for version 2.0 of the SUPREMO multistatic sonar performance model is provided. SUPREMO allows scientists and engineers engaged in sonar research to predict underwater acoustic signal and interering background intensities. Program operation is explained in a step-by-step manner along with relevant program assumptions and approximations.
SUPREMO v.2.0 User Guide. Prior, M.K. NURC-FR-2007-009. June 2007.
A user guide for version 2.0 of the SUPREMO multistatic sonar performance model is provided. Input data entry and model execution are explained in a step-by-step manner and the display of model predictions is described. Examples of automated program execution are given, showing how the model may be used to assess the impact of environmental input data uncertainty and to help in tactical considerations such as the identification of best deployment depths for sonars.
Deployable network setup for Remote Sensing Satellite Ground Sation (RSSGS): architecture and performance considerations. Berni, A. ; Merani, D.; Gorra, E.; Leonard, M.; Ranelli, P.; Sciarrone, M.; Stenvoll, R. NURC-FR-2007-010. June 2007.
This document reports on the trials performed with the NURC Remote Sensing Satellite Ground Station (RSSGS) in 2005, as part of project “Covert Remote Sensing in Support of Expeditionary Operations” funded by Allied Command Transformation with the objective of demonstrating and evaluating the utility of remotely sensed METOC data to support expeditionary operations under the Recognized Environmental Picture (REP) concept. The experiment provided proof of concept and enabled the development of procedures and communications infrastructures that will enable future deployments of the RSSGS. Additionally, the application-level performance of the RSSGS communication system in different configurations was tested and validated.
Stochastic nature of physical parameterizations in ensemble prediction: a stochastic convection parameterization. Teixeira, J.; Reynolds, C.A. June 2007. NURC-FR-2007-012.
The goal of the ‘Air-Sea Interaction Effects on Expeditionary Force Operations’ project is to improve littoral METOC forecast capabilities to support Expeditionary Operations (EO) and to be used in Tactical Decision Aids (TDAs). Environmental forecasting systems often possess a substantial amount of uncertainty. A more refined knowledge of the uncertainty associated with the forecast of any variable would be of great help to support NATO operations. In order to answer this type of questions ensemble prediction systems have been developed to produce estimates of the uncertainty of model forecasts. The development of stochastic parameterizations is a fundamental stepping-stone for improved ensemble prediction systems. In this report it is argued that ensemble systems can be devised, in which parameterizations of sub-grid scale motions are utilized in a stochastic manner. The equations used to estimate the variance of a generic variable are discussed and a simplified algorithm is proposed as a preliminary attempt. Results from the implementation of this stochastic scheme show that this method is able to produce a substantial ensemble spread. Simulations with initial-condition and stochastic convection perturbations show a promising increase in ensemble spread and decrease in the number of outliers. The research devoted to this stochastic scheme has matured to the point of operational transition. As an example, this method is currently being tested in the US Navy operational ensemble. These methods will also be tested in the high-resolution re-locatable system used at NURC, as part of an environmental prediction package for EO support in the context of the Recognized Environmental Picture (REP).
Evaluation of Minehunting Concepts Based on AUVs. Rothenbach, M.; Sodhi, M.; Bovio, E. NURC-FR-2007-008. May 2007.
The report evaluates by computer simulation and analytical methods the potential role of different classes of AUVs in a variety of MCM scenarios. The five scenarios selected are: a strategic sea line of communications (SLOC); a strategic port; fleet operating areas; amphibious operation areas; and inspection of own harbour. The four classes of vehicles range from large AUVs equipped with long range/high resolution Synthetic Aperture Sonar (SAS) and large battery capacity, to medium size vehicles equipped with SAS or high performance Side Scan Sonar (SSS), ending with fleets of small light weight AUVs equipped with entry level SSS. The report describes the criteria, the computer simulations and analytical tools used to compare the performance of the various systems. In addition to the study of the above scenarios, the report examines the influence of the bathymetry for the AUV mission planning, and analyses optimal search techniques for multi vehicle operations. The study concludes that the required performance can be achieved in all scenarios by at least one type of vehicle, but one type only does not perform well in all scenarios. In general large areas require large vehicles; confined spaces are better handled by small vehicles. Overall the medium AUV with SAS is the preferred solution.
AUV Navigation Systems for MCM Operations. Baralli, F.; Evans, B.; Coiras,E.; Bellettini, A. NURC-FR-2007-007. May 2007.
This report documents the analysis performed within the project 3G1 on the accuracy of COTS Navigation Systems for AUVs. Performance improvements in using innovative navigation aids like, SAS-micro-navigation and contact matching are also assessed. Accuracy assessments are conducted by analysing data collected by different vehicles during trials/exercises organised by NURC in different environments/areas. AUVs are divided into two classes depending on their navigation systems: Aided Inertial Navigation and LBL aided dead-reckoning. Particular attention is taken in the definition of the performance of AUV navigation systems for the execution of military missions such as mine disposal in shallow water, change detection and port protection.
Benchmarking scattering from spheres and cylinders near the seafloor: A numerical comparitive study. Jensen, F.B.; M. Zampolli, M.; Tesei, A. NURC-FR-2007-005. May 2007.
A workshop was organized to assess the state-of-affairs in target scatter modeling, both for low-frequency (0.1–10 kHz) and high-frequency (>=100kHz) applications. The problem set included free-field scenarios as well as scattering from targets that were proud, half-buried or fully buried in the sediment. Ten different organizations provided solutions to the various test problems. Low-frequency results are abundant for most test problems, whereas high-frequency solutions were provided only for the free-space cylinder with different material parameters and fillings. A detailed comparison between the various solutions for each test problem are presented and low-frequency reference solutions are identified for all but one of the eleven test problems.
Automatic Target Recognition in Synthetic Aperture Sonar Imagery using Template Deformation and Gaussian Processes. Williams, D.; Myers, V. NURC-FR-2007-006. March 2007.
This report describes a method for performing automatic target recognition (ATR) of targets of interest in synthetic aperture sonar (SAS) images. ATR is considered an enabling technology for mine countermeasures (MCM) operations using autonomous underwater vehicles (AUVs). The ability to detect and classify targets in side- and forward-looking highfrequency sonars can be used to significantly augment the ability for AUVs for onboard decision-making, resulting in reduced mission time. This is achieved by intelligently making decisions about mission parameters in real time and relying less on human intervention and the limited communications bandwidth underwater.
A method for sea floor clutter suppression in side scan sonar images. R. Grasso, F. Spina. NURC-FR-2007-002. February 2007.
A clutter suppression algorithm for improving detection of unknown spatially distributed small targets on a textured sea floor background in side scan sonar images is analysed and tested. The algorithm is based on a multi-resolution transform of the sonar data and a projection pursuit (PP) algorithm, which is used to enhance signal-to-clutter ratio in the multi-resolution space. The sea floor background in a sub-image is modelled by a set of texture features estimated from the image data by a two dimensional un-decimated wavelet transform. Each image sample is associated with a vector of features which are composed by the local mean energy resulting from a bank of multi-resolution oriented bandpass filters. The target points in the feature space appear as a small set of outliers compared to the feature vector distribution of the background. The feature vector data set is projected, by a PP algorithm, along the direction of kurtosis maximisation of the data which is sensitive to the outliers in a data set. The projected signal is used in a threshold test for detecting anomalies in sea floor background signals. An ad hoc gradient algorithm to maximise the kurtosis iteratively was studied and developed. An efficient procedure to find a first guess solution to initialise the gradient algorithm was proposed and tested on a side scan sonar data set acquired during the NURC BP02 cruise by a ship towed sensor. The improvement in the signal-to-clutter ratio observed in this case ranges between 5 and 8 dB.
Characterization of local seabed properties using synthesized horizontal array data. NURC-FR-2007-001. January 2007.
In April 2002 NURC conducted the BOUNDARY 2002 experiment on the Malta Plateau, Mediterranean Sea, with the aim of collecting high-quality environmental and acoustic data. A part of these data was directly intended for geoacoustic inversion characterizing the bottom properties by acoustic measurements. The acoustic data were collected by using towed sources covering a frequency band 180-7000Hz and received on a bottom- moored vertical array. The sources were towed about 1 km on both sides of the array and passing the shallowest hydrophone at a range of 21m. The transmission rate was every 3 s resulting in high spatial sampling of the acoustic field. The high repetition rate of the transmissions provide the possibility of building a 1-km synthetic horizontal array with a virtual hydrophone spacing of about 5.9m. The principle of reciprocity, i.e. interchanging the source and receiver positions, is applied to build the synthetic horizontal array. The acoustic data collected from this synthetic horizontal array are used to characterize the bottom properties locally by matched-field geoacoustic inversion and to extract the reflection coefficient from a spherical wave interacting with the bottom. The reflection coefficient is calculated by time-gating of the direct and bot tom arrivals of the matched-filtered signal. The results from the inversion are compared with independent “ground truth” data collected at the experimental site. Both the matched-field inversion and the determination of the reflection co efficient are strong candidates for rapid characterization of bottom properties locally over large areas. This implies the use of fast and agile platforms like Autonomous Underwater Vehicles with broad-band sound sources and arrays and eventually a supporting sono-buoy field.