This PDF file contains the front matter associated with SPIE Proceedings Volume 8047, including the Title Page, Copyright information, Table of Contents, Introduction, and Conference Committee listing.

The Defense Intelligence Agency (DIA) is developing Terra Harvest, an open, integrated battlefield unattended ground sensors (UGS) architecture that will employ multiple, flexible sensors via standards-based integration. The Terra Harvest open architecture separates the UGS system into fundamental components and standardizes internal and external interfaces to optimize interoperability. Other acquisition programs can take advantage of this open architecture to meet challenging mission requirements.

Under the Terra Harvest Program, the Defense Intelligence Agency (DIA) has the objective of developing a universal Controller for the Unattended Ground Sensor (UGS) community. The mission is to define, implement, and thoroughly document an open architecture that universally supports UGS missions, integrating disparate systems, peripherals, etc. The Controller's inherent interoperability with numerous systems enables the integration of both legacy and future Unattended Ground Sensor System (UGSS) components, while the design's open architecture supports rapid third-party development to ensure operational readiness. The successful accomplishment of these objectives by the program's Phase 3b contractors is demonstrated via integration of the companies' respective plug-'n-play contributions that include various peripherals, such as sensors, cameras, etc., and their associated software drivers. In order to independently validate the Terra Harvest architecture, L-3 Nova Engineering, along with its partner, the University of Dayton Research Institute (UDRI), is developing the Terra Harvest Open Source Environment (THOSE), a Java based system running on an embedded Linux Operating System (OS). The Use Cases on which the software is developed support the full range of UGS operational scenarios such as remote sensor triggering, image capture, and data exfiltration. The Team is additionally developing an ARM microprocessor evaluation platform that is both energyefficient and operationally flexible. The paper describes the overall THOSE architecture, as well as the implementation strategy for some of the key software components. Preliminary integration/test results and the Team's approach for transitioning the THOSE design and source code to the Government are also presented.

Unattended Ground Sensors (UGS) from a wide range of manufacturers have difficulty interoperating with each other, and common control and dissemination points. Therefore, an UGS controller that accepts data from a wide range of sensors and transmits this data coherently is essential. This paper proposes a packaged native data format for transmission. This format can be combined with Open Geospatial Consortium (OGC) Sensor Model Language (SensorML) sensor descriptions to provide sensor interoperability. A SensorML-enabled UGS controller that transmits packaged native data format sensor information is a powerful tool that can provide situational awareness and a common operational picture.

Our previous work has explored the application of enterprise middleware techniques at the edge of the network to address the challenges of delivering complex sensor network solutions over heterogeneous communications infrastructures. In this paper, we develop this approach further into a practicable, semantically rich, model-based design and analysis approach that considers the sensor network and its contained services as a service-oriented architecture. The proposed model enables a systematic approach to service composition, analysis (using domain-specific techniques), and deployment. It also enables cross intelligence domain integration to simplify intelligence gathering, allowing users to express queries in structured natural language (Controlled English).

In this paper, we present a method for a highly decentralized yet structured and flexible approach to achieve systems interoperability by orchestrating data and behavior across distributed military systems and assets with security considerations addressed from the beginning. We describe an architecture of a tool-based design of business processes called Decentralized Operating Procedures (DOP) and the deployment of DOPs onto run time nodes, supporting the parallel execution of each DOP at multiple implementation nodes (fixed locations, vehicles, sensors and soldiers) throughout a battlefield to achieve flexible and reliable interoperability. The described method allows the architecture to; a) provide fine grain control of the collection and delivery of data between systems; b) allow the definition of a DOP at a strategic (or doctrine) level by defining required system behavior through process syntax at an abstract level, agnostic of implementation details; c) deploy a DOP into heterogeneous environments by the nomination of actual system interfaces and roles at a tactical level; d) rapidly deploy new DOPs in support of new tactics and systems; e) support multiple instances of a DOP in support of multiple missions; f) dynamically add or remove run-time nodes from a specific DOP instance as missions requirements change; g) model the passage of, and business reasons for the transmission of each data message to a specific DOP instance to support accreditation; h) run on low powered computers with lightweight tactical messaging. This approach is designed to extend the capabilities of existing standards, such as the Generic Vehicle Architecture (GVA).

The invention of the Microflown sensor has made it possible to measure acoustic particle velocity directly. An acoustic vector sensor (AVS) measures the particle velocity in three directions (the source direction) and the pressure. The sensor is a uniquely versatile battlefield sensor because its size is a few millimeters and it is sensitive to sound from 10Hz to 10kHz. This article shows field tests results of acoustic vector sensors, measuring rifles, heavy artillery, fixed wing aircraft and helicopters. Experimental data shows that the sensor is suitable as a ground sensor, mounted on a vehicle and on a UAV.

A major difficulty in classifying seismic events in the near field is the existence of multiple wave types and their lack of time to separate from one another. During an impulsive seismic event, as the seismic wave components travel through a medium, the difference in their velocities results in a superimposed signal that will look drastically different at varying distances. It would be most beneficial to detect, classify and localize targets creating impulsive events if seismic sensor data could be reduced to a single wave type that has an expected shape and consistent features that do not change as a function of distance. Research was conducted to determine if measuring seismic data from within enclosures of specific architectural design could be used to attenuate specific wave types while maintaining energy of other wave types. The resulting waves produced by these geophone enclosures were then subject to testing using various algorithms to determine their ability to detect, classify, and localize seismic targets.

A profiling sensor has been realized using a vertical column of sparse detectors with the sensor's optical axis configured perpendicular to the plane of the vertical column of detectors. Traditionally, detectors of the profiling sensor are placed in a sparse vertical column configuration. A subset of the detectors may be removed from the vertical column and placed at arbitrary locations along the anticipated path of the objects of interest, forming a custom detector array configuration. Objects passing through the profiling sensor's field of view have traditionally been classified via algorithms processed off-line. However, reconstruction of the object profile is impossible unless the detectors are placed at a known location relative to each other. Measuring these detector locations relative to each other can be particularly time consuming, making this process impractical for custom detector configuration in the field. This paper describes a method that can be used to determine a detector's relative location to other detectors by passing a known profile through the sensor's field of view as part of the configuration process. Real-time classification results produced by the embedded controller for a variety of objects of interest are also described in the paper.

Profiling sensor systems have been shown to be effective for detecting and classifying humans against animals. A profiling sensor with a 360 horizontal field of view was used to generate profiles of humans and animals for classification. The sensor system contains a long wave infrared camera focused on a smooth conical mirror to provide a 360 degree field of view. Human and animal targets were detected at 30 meters and an approximate height to width ratio was extracted for each target. Targets were tracked for multiple frames in order to segment targets from background. The average height to width ratio was used as a single feature for classification. The Mahalanobis distance was calculated for each target in the single feature space to provide classification results.

Data fusion plays a major role in assisting decision makers by providing them with an improved situational awareness so that informed decisions could be made about the events that occur in the field. This involves combining a multitude of sensor modalities such that the resulting output is better (i.e., more accurate, complete, dependable etc.) than what it would have been if the data streams (hereinafter referred to as 'feeds') from the resources are taken individually. However, these feeds lack any context-related information (e.g., detected event, event classification, relationships to other events, etc.). This hinders the fusion process and may result in creating an incorrect picture about the situation. Thus, results in false alarms, waste valuable time/resources. In this paper, we propose an approach that enriches feeds with semantic attributes so that these feeds have proper meaning. This will assist underlying applications to present analysts with correct feeds for a particular event for fusion. We argue annotated stored feeds will assist in easy retrieval of historical data that may be related to the current fusion. We use a subset of Web Ontology Language (OWL), OWL-DL to present a lightweight and efficient knowledge layer for feeds annotation and use rules to capture crucial domain concepts. We discuss a solution architecture and provide a proof-of-concept tool to evaluate the proposed approach. We discuss the importance of such an approach with a set of user cases and show how a tool like the one proposed could assist analysts, planners to make better informed decisions.

We have developed a flexible, target-driven, multi-modal, physics-based fusion architecture that efficiently searches sensor detections for targets and rejects clutter while controlling the combinatoric problems that commonly arise in datadriven fusion systems. The informational constraints imposed by long lifetime requirements make systems vulnerable to false alarms. We demonstrate that our data fusion system significantly reduces false alarms while maintaining high sensitivity to threats. In addition, mission goals can vary substantially in terms of targets-of-interest, required characterization, acceptable latency, and false alarm rates. Our fusion architecture provides the flexibility to match these trade-offs with mission requirements unlike many conventional systems that require significant modifications for each new mission. We illustrate our data fusion performance with case studies that span many of the potential mission scenarios including border surveillance, base security, and infrastructure protection. In these studies, we deployed multi-modal sensor nodes - including geophones, magnetometers, accelerometers and PIR sensors - with low-power processing algorithms and low-bandwidth wireless mesh networking to create networks capable of multi-year operation. The results show our data fusion architecture maintains high sensitivities while suppressing most false alarms for a variety of environments and targets.

Multi sensor fusion techniques are widely employed in several surveillance applications (e.g., battlefield monitoring, air traffic control, camp protection, etc). The necessity of tracking the elements of a dynamic system usually requires combining information from heterogeneous data sources in order to overcome the limitations of each sensor. The gathered information might be related to the target kinematics (position, velocity), its physical features (shape, size, composition) or intentions (route plan, friend/foe, engaged sensor modes, etc). The combination of such heterogeneous sensor data proved to benefit from the exploitation of context information, i.e., static and dynamic features of the scenario, represented in a Knowledge Base (KB). A Geographic Information System (GIS) is a typical example for a KB that can be exploited for the enhancement of multi sensor data fusion. The present paper describes potential strategies for "knowledge-based" data fusion in the area of Maritime Situational Awareness (MSA). MSA is founded on the data from heterogeneous sources, including radars, Navigation Aids, air- and space-based monitoring services, and recently-conceived passive sensors. Several strategies for optimally fusing two or more of these information data flows have been proposed for MSA applications. Relevant KB information comprises port locations, coastal lines, preferred routes, traffic rules, and potentially a maritime vessel database. We propose mathematical models and techniques to integrate kinematic constraints, e.g., in terms of navigation fields, and different object behaviour into a data fusion approach. For an exemplary sensor suite, we evaluate performance measures in the framework of centralised and decentralised fusion architectures.

Persistent surveillance applications require unattended sensors deployed in remote regions to track and monitor some physical stimulant of interest that can be modeled as output of time varying stochastic process. However, the accuracy or the trustworthiness of the information received through a remote and unattended sensor and sensor network cannot be readily assumed, since sensors may get disabled, corrupted, or even compromised, resulting in unreliable information. The aim of this paper is to develop information theory based metric to determine sensor trustworthiness from the sensor data in an uncertain and time varying stochastic environment. In this paper we show an information theory based determination of sensor data trustworthiness using an adaptive stochastic reference sensor model that tracks the sensor performance for the time varying physical feature, and provides a baseline model that is used to compare and analyze the observed sensor output. We present an approach in which relative entropy is used for reference model adaptation and determination of divergence of the sensor signal from the estimated reference baseline. We show that that KL-divergence is a useful metric that can be successfully used in determination of sensor failures or sensor malice of various types.

Small xed-wing UAS (SUAS) such as Raven and Unicorn have limited power, speed, and maneuverability. Their missions can be dramatically hindered by environmental conditions (wind, terrain), obstructions (buildings, trees) blocking clear line of sight to a target, and/or sensor hardware limitations (xed stare, limited gimbal motion, lack of zoom). Toyon's Sensor Guided Flight (SGF) algorithm was designed to account for SUAS hardware shortcomings and enable long-term tracking of maneuvering targets by maintaining persistent eyes-on-target. SGF was successfully tested in simulation with high-delity UAS, sensor, and environment models, but real- world ight testing with 60 Unicorn UAS revealed surprising second order challenges that were not highlighted by the simulations. This paper describes the SGF algorithm, our rst round simulation results, our second order discoveries from ight testing, and subsequent improvements that were made to the algorithm.

Techniques were developed to localize acoustic quasiperiodic signals using microphone arrays located on the ground and on an aerostat. The direction of arrival (DOA) was computed at each array and then the position of the source was estimated using algorithms based upon triangulation. Differential time delays between the microphones in a tetrahedral array were estimated in the frequency domain, and then DOA estimates were calculated using a weighted least squares approach. The location of the target was calculated by minimizing the weighted squared error of a cost function for different combinations of DOA estimates. The algorithms were tested offline using data collected by the U.S. Army Research Laboratory on an aircraft. The ground-truth position of the target was recorded using a GPS system as it maneuvered and compared to the results obtained from the localization algorithms. The algorithms performed well when estimating the x and y positions, but had difficulty obtaining consistently good z positions, or equivalently, height estimates.

Due to increased surveillance, information, and exploitation assets, and the wide variety of interfaces, protocols, etc. that these systems use, the interactions between these systems is rapidly growing more complex. Likewise, integrating a new component into existing systems is no longer a trivial challenge. In order to make modification and integration of components into a larger system easier, the Air Force Research Labs have developed Sensor Processing Architecture for Data Exploitation (SPADE). The contribution of this paper is to discuss the successful integration of a vehicle tracker into the SPADE architecture, using Pursuer as the user interface.

This paper will look at using open source tools (Blender © [17], LuxRender © [18], and Python © [19]) to build an image processing model for exploring combinations of sensors/platforms for any given image resolution. The model produces camera position, camera attitude, and synthetic camera data that can be used for exploitation purposes. We focus on electro-optical (EO) visible sensors to simplify the rendering but this work could be extended to use other rendering tools that support different modalities. Due to the computational complexity of ray tracing we employ the Amazon Elastic Cloud Computer to help speed up the generation of large ray traced scenes. The key idea of the paper is to provide an architecture for layered sensing simulation which is modular in design and constructed on open-source off-the-shelf software. This architecture shows how leveraging existing open-source software allows for practical layered sensing modeling to be rapidly assimilated and utilized in real-world applications. In this paper we demonstrate our model output is automatically exploitable by using generated data with an innovative video frame mosaic algorithm.

Tactical networks should be optimized to deliver the maximum amount of useful information from which decisions may be made. This requires that both the quality and amount of information be considered. The quality of information may be judged by both intrinsic and contextual attributes. We define the operational information content capacity (OICC) as the measure of the amount of useful information a network can deliver. In this paper we discuss several ways to quantify OICC and determine the residual information content capacity in a network based on a set of information requests. We first define functions which relate specific metrics to the quality of a piece of information to be used for a certain purpose. From this we determine the amount of data required to deliver the information to its recipient and the resultant "information bits" which can be derived. We then provide two illustrative examples highlighting the use of OICC.

Sensor networks find application in many tactical ISR/ISTAR processes and applications. However, these processes and applications depend on reliable collection, distribution and delivery of information that, typically, travels over multiple interconnecting nodes to reach processing centres, and are susceptible to various disruptions such as the ones caused caused by message drops, packet loss and loss of connectivity due to high traffic volumes and noise on the wireless medium. In this paper, we investigate and present approaches to pro-actively adapt routing over such networks by forecasting potential faulty regions of the network based on previous trends and reorganising routing paths. We have prototyped this approach in the ITA Sensor Fabric, an evolving middleware infrastructure for sensor networks. We, further, provide some preliminary results based on simulations.

Broadcast scheduling has been extensively studied in wireless environments, where a base station broadcasts data to multiple users. Due to the sole wireless channel's limited bandwidth, only a subset of the needs may be satisfiable, and so maximizing total (weighted) throughput is a popular objective. In many realistic applications, however, data are dependent or correlated in the sense that the joint utility of a set of items is not simply the sum of their individual utilities. On the one hand, substitute data may provide overlapping information, so one piece of data item may have lower value if a second data item has already been delivered; on the other hand, complementary data are more valuable than the sum of their parts, if, for example, one data item is only useful in the presence of a second data item. In this paper, we define a data bundle to be a set of data items with possibly nonadditive joint utility, and we study a resulting broadcast scheduling optimization problem whose objective is to maximize the utility provided by the data delivered.

The net-centric ISR/ISTAR networks are expected to play a crucial role in the success of critical tasks such as base perimeter protection, border patrol and so on. To accomplish these tasks in an effective and expedient manner, it is important that these networks have the embedded capabilities to discover, delegate, and gather relevant information in a timely and robust manner. In this paper, we present a system architecture and an implementation that combines a service based reasoning mechanism with a sensor middleware infrastructure so that tasks can be executed efficiently and effectively. A knowledge base, utilising the Semantic Web technologies, provides the foundation for reasoning mechanism that assists users to discover, identify and allocate resources that are made available through the middleware, in order to satisfy the needs of tasks. Once resources are allocated to any given task, they can be accessed, controlled, shared, and their data feeds consumed through the Fabric middleware. We use the semantic descriptions from the knowledge base to annotate the resources (types, capabilities, etc.) in the sensor middleware so that they can be retrieved for reasoning during the discovery and identification phases. The reasoner is implemented as a HTTP web service, with the following characteristics: 1. Computational intensive operations are off-loaded to dedicated nodes, preserving the resources in the ISR/ISTAR networks. 2. HTTP services are accessible through a standard set of APIs irrespective of the reasoner technology used. 3. Support for seamless integration of different reasoners into the system.

A team consisting of Teledyne Scientific Company, the University of California at Santa Barbara and the Army Research Laboratory* is developing technologies in support of automated data exfiltration from heterogeneous battlefield sensor networks to enhance situational awareness for dismounts and command echelons. Unmanned aerial vehicles (UAV) provide an effective means to autonomously collect data from a sparse network of unattended ground sensors (UGSs) that cannot communicate with each other. UAVs are used to reduce the system reaction time by generating autonomous collection routes that are data-driven. Bio-inspired techniques for search provide a novel strategy to detect, capture and fuse data. A fast and accurate method has been developed to localize an event by fusing data from a sparse number of UGSs. This technique uses a bio-inspired algorithm based on chemotaxis or the motion of bacteria seeking nutrients in their environment. A unique acoustic event classification algorithm was also developed based on using swarm optimization. Additional studies addressed the problem of routing multiple UAVs, optimally placing sensors in the field and locating the source of gunfire at helicopters. A field test was conducted in November of 2009 at Camp Roberts, CA. The field test results showed that a system controlled by bio-inspired software algorithms can autonomously detect and locate the source of an acoustic event with very high accuracy and visually verify the event. In nine independent test runs of a UAV, the system autonomously located the position of an explosion nine times with an average accuracy of 3 meters. The time required to perform source localization using the UAV was on the order of a few minutes based on UAV flight times. In June 2011, additional field tests of the system will be performed and will include multiple acoustic events, optimal sensor placement based on acoustic phenomenology and the use of the International Technology Alliance (ITA) Sensor Network Fabric (IBM).

The U.S. Army Research Laboratory (ARL) has recently concluded a research experiment to study the benefits of multimodal sensor fusion for improved hostile-fire-defeat (HFD) in an urban setting. This joint effort was led by ARL in partnership with other R&D centers and private industry. The primary goals were to detect hostile fire events (small arms, mortars, rockets, IEDs) and hostile human activities by providing solutions before, during, and after the events to improve sensor networking technologies; to develop multimodal sensor data fusion; and to determine effective dissemination techniques for the resultant actionable intelligence. Technologies included ultraviolet, infrared, retroreflection, visible, glint, Laser Detection and Ranging (LADAR), radar, acoustic, seismic, E-field, magnetic, and narrowband emission technologies; all were found to provide useful performance. The experiment demonstrated that combing data and information from diverse sensor modalities can significantly improve the accuracy of threat detections and the effectiveness of the threat response. It also demonstrated that dispersing sensors over a wide range of platforms (fixed site, ground vehicles, unmanned ground and aerial vehicles, aerostat, Soldier-worn) added flexibility and agility in tracking hostile actions. In all, the experiment demonstrated that multimodal fusion will improve hostile event responses, strike force efficiency, and force protection effectiveness.

In distributed, heterogeneous, multi-agent teams, agents may have different capabilities and types of sensors. Agents in dynamic environments will need to cooperate in real-time to perform tasks with minimal costs. Some example scenarios include dynamic allocation of UAV and UGV robot teams to possible hurricane survivor locations, search and rescue and target detection. Auction based algorithms scale well because agents generally only need to communicate bid information. In addition, the agents are able to perform their computations in parallel and can operate on local information. Furthermore, it is easy to integrate humans and other vehicle types and sensor combinations into an auction framework. However, standard auction mechanisms do not explicitly consider sensors with varying reliability. The agents sensor qualities should be explicitly accounted. Consider a scenario with multiple agents, each carrying a single sensor. The tasks in this case are to simply visit a location and detect a target. The sensors are of varying quality, with some having a higher probability of target detection. The agents themselves may have different capabilities, as well. The agents use knowledge of their environment to submit cost-based bids for performing each task and an auction is used to perform the task allocation. This paper discusses techniques for including a Bayesian formulation of target detection likelihood into this auction based framework for performing task allocation across multi-agent heterogeneous teams. Analysis and results of experiments with multiple air systems performing distributed target detection are also included.