Few argue with the need for modeling and simulation (M&S) to better or more completely represent current and expected military operations. The challenge is to decide where to make specific improvements in M&S representation and functionality within time, funding, technology, and research limitations. So, it is natural to select key areas - Grand Challenges - for a significant evolution in M&S where a major effort of many at considerable cost is needed to deal with the critical issues ahead. This paper selects three proposed and related Grand Challenges. First, M&S Depiction of Information and Effects-Based Operations, as a Grand Challenge, will assist in creating sufficiently realistic battlespaces for M&S users. Second, M&S Support to Crisis Response and Military Operations, as a Grand Challenge, is a key area that will help the Department of Defense meet transformation goals. Third, Effective Development of Future Simulations, as a Grand Challenge, will set the standards by which future M&S improvements and new M&S programs will be acquired to ensure needed simulations are delivered on time and at desired cost.

Modern anti-tank missiles and the requirement of rapid deployment are limiting the use of passive armour in protecting land vehicles. Vehicle survivability is becoming more dependent on sensors, computers and countermeasures to detect and avoid threats. The integration of various technologies into a Defensive Aids Suite (DAS) can be designed and analyzed by combining field trials and laboratory data with modeling and simulation. MATLAB is used as a quick prototyping tool to model DAS systems and facilitate transfer to other researchers. The DAS model can be transferred from MATLAB or programmed directly in ModSAF (Modular Semi-Automated Forces), which is used to construct the virtual battlefield. Through scripted input files, a fixed battle approach ensures implementation and analysis meeting the requirements of three different interests. These three communities include the scientists and engineers, military and operations research. This approach ensures the modelling of processes known to be important regardless of the level of information available about the system. A system can be modelled phenomenologically until more information is available. Further processing of the simulation can be used to optimize the vehicle for a specific mission. ModSAF will be used to analyze and plan trials and develop DAS technology for future vehicles. Survivability of a DAS-equipped vehicle can be assessed relative to a basic vehicle without a DAS. In later stages, more complete DAS systems will be analyzed to determine the optimum configuration of the DAS components and the effectiveness of a DAS-equipped vehicle for specific missions. These concepts and approach will be discussed in the paper.

The Simulation for Predictive Battlespace Awareness (SiPBA) will enhance Decision Support within the Air Force. Through the use of existing simulations and collaboration enabling tools such as workflows, resource agents and process sharing, the Air Force can augment multi-level simulation with repeatable processes and configuration control. SiPBA uses a forms-based approach that allows the simulation details to be removed from the individual posing the operational questions. Subsequently, the level of detail of the simulation required and the combination of simulations required to answer the question also becomes abstract. This paper will review the process of exploiting a forms-based approach within the framework of collaboration tools to support multi-level simulation.

One of the more difficult problems facing an analyst wishing to use a simulation is the task of collecting data and transforming it into a correctly formatted scenario. Raw data is often available from a variety of sources: multi-spectral force deployment (MSFD) documents, the electronic warfare integrated reprogramming database (EWIRDB), free text documents such as intelligence reports, pre-existing simulation scenarios, and scenarios taken from other simulations. The task of transforming this data into a usable scenario involves searching for the relevant information, followed by a manual transformation of the original format to the correct simulation format. This problem can be greatly alleviated by using a combination of three technologies: automatic parser generation, repository architectures using extensible markup language (XML), and information retrieval (IR) techniques. Automatic parser generation tools like JavaCC can automatically generate source code capable of reading data sources such as old Joint Integrated Mission Model (JIMM) or Suppressor input files. For simulations that regularly add scenario keywords to support changing needs, this can greatly reduce redevelopment time and cost for supporting tools. The objects parsed by this source can then be encapsulated in XML and stored into a repository. Using information retrieval techniques, objects can then be queried from the repository and transformed into the appropriate format for use in a scenario.

This paper introduces the concept of using simulation for both plan tracking and state estimation and prediction. Given some set of objectives the military commander must devise a sequence of actions that transform the current state to the desired one. The desire to do this in faster than real-time so that many courses of action can be considered motivates us to investigate modeling techniques that explicitly produce such courses of action. This class of problem can be modeled as a Markov decision process (MDP) whose principal solution is stochastic dynamic programming. In this paper we consider the extension of a MDP model of air operations to the partially observed case.

A key component of the Joint Air Operations (JAO) environment is the dynamic control of resources in the presence of uncertainty. This control involves the allocation of resources (e.g., different aircraft types) to prosecute targets and collect information while accounting for uncertain future events and partial, imperfect observations. The objective is to maximize the reward associated with the effective prosecution of targets, which is contingent on information collection, while minimizing loss of resources. In this paper, we extend an earlier formulation of an optimal dynamic resource allocation problem to explicitly include the dynamics of information collection and to identify the complexities involved. We then describe a simulation-based approach that was developed to solve the dynamic JAO control problem in the presence of partial and imperfect information.

Combat always involves uncertainty and uncertainty entails risk. To ensure that a combat task is prosecuted with the desired probability of success, the task commander has to devise an appropriate task force and then adjust it continuously in the course of battle. In order to do so, he has to evaluate how the probability of task success is related to the structure, capabilities and numerical strengths of combatants. For this purpose, predictive models of combat dynamics for combats in which the combatants fire asynchronously at random instants are developed from the first principles. Combats involving forces with both unlimited and limited ammunition supply are studied and modeled by stochastic Markov processes. In addition to the Markov models, another class of models first proposed by Brown was explored. The models compute directly the probability of win, in which we are primarily interested, without integrating the state probability equations. Experiments confirm that they produce exactly the same results at much lower computational cost.

A multi-agent system model of the origins of an archaic state is developed. Agent interaction is mediated by a collection of rules. The rules are mined from a related large-scale data base using two different techniques. One technique uses decision trees while the other uses rough sets. The latter was used since the data collection techniques were associated with a certain degree of uncertainty. The generation of the rough set rules was guided by Genetic Algorithms. Since the rules mediate agent interaction, the rule set with fewer rules and conditionals to check will make scaling up the simulation easier to do. The results suggest that explicitly dealing with uncertainty in rule formation can produce simpler rules than ignoring that uncertainty in situations where uncertainty is a factor in the measurement process.

Computer-based wargames have seen much improvement in recent years due to rapid increases in computing power. Because these games have been developed for the entertainment industry, most of these advances have centered on the graphics, sound, and user interfaces integrated into these wargames with less attention paid to the game's fidelity. However, for a wargame to be useful to the military, it must closely approximate as many of the elements of war as possible. Among the elements that are typically not modeled or are poorly modeled in nearly all military computer-based wargames are systematic effects, command and control, intelligence, morale, training, and other human and political factors. These aspects of war, with the possible exception of systematic effects, are individually modeled quite well in many board-based commercial wargames. The work described in this paper focuses on incorporating these elements from the board-based games into a computer-based wargame. This paper will also address the modeling and simulation of the systemic paralysis of an adversary that is implied by the concept of Effects Based Operations (EBO). Combining the fidelity of current commercial board wargames with the speed, ease of use, and advanced visualization of the computer can significantly improve the effectiveness of military decision making and education. Once in place, the process of converting board wargames concepts to computer wargames will allow the infusion of soft factors into military training and planning.

Project Albert is an initiative of the US Marine Corps which uses a series of new models and tools, multidisciplinary teams, and the scientific method to explore questions of interest to military planners. Project Albert attempts to address key areas that traditional modeling and simulation techniques often do not capture satisfactorily and uses two data management concepts, data farming and data mining, to assist in identifying areas of interest. The current suite of models used by Project Albert includes four agent-based models that allow agents to interact with each other and produce emergent behaviors. The 4th International Project Albert Workshop was held 6-9 August 2001 in Australia. Workshop participants split into five groups, each of which attempted to apply various combinations of the Project Albert models to answer a series of questions in five areas: Control Operations; Reconnaissance, Surveillance, and Intelligence Force Mix; Precision Maneuver; Mission Area Analysis; and Peace Support Operations. This paper focuses on the methodology used during the workshop, the results of the workshop, and a summary of follow-on work since the workshop.

The control of the Unmanned Combat Air Vehicle's swarm behavior is studied. One command string controls the motion of all Unmanned Combat Air Vehicles in a mission. Each Unmanned Combat Air Vehicle moves according to the control decoded from the same control command string. There is no explicit coordination among them. However, the decoding of a control command string partially depends on other Unmanned Combat Air Vehicles surrounding it. If the control command string is properly chosen, the motion of the swarm of Unmanned Combat Air Vehicles will perform well collectively. Genetic algorithm is used to evolve the control command string. The robustness of the control is studied. Monte Carlo simulation in conjunction with Genetic Algorithm is used to evolve the robust control when wind-gust disturbance exists. The results of different approached are compared.

In many critical applications such as airport operations (for capacity planning), military simulations (for tactical training and planning), and medical simulations (for the planning of medical treatment and surgical operations), it is very useful to conduct simulations within physically accurate and visually realistic settings that are represented by real video imaging sequences. Furthermore, it is important that the simulated entities conduct autonomous actions which are realistic and which follow plans of action or intelligent behavior in reaction to current situations. We describe the research we have conducted to incorporate synthetic objects in a visually realistic manner in video sequences representing a real scene. We also discuss how the synthetic objects can be designed to conduct intelligent behavior within an augmented reality setting. The paper discusses both the computer vision aspects that we have addressed and solved, and the issues related to the insertion of intelligent autonomous objects within an augmented reality simulation.

SPEEDES, the Synchronous Parallel Environment for Emulation and Discrete Event Simulation, is a software framework that supports simulation applications across parallel and distributed architectures. SPEEDES is used as a simulation engine in support of numerous defense projects including the Joint Simulation System (JSIMS), the Joint Modeling And Simulation System (JMASS), the High Performance Computing and Modernization Program's (HPCMP) development of a High Performance Computing (HPC) Run-time Infrastructure, and the Defense Modeling and Simulation Office's (DMSO) development of a Human Behavioral Representation (HBR) Testbed. This work documents some of the performance metrics obtained from benchmarking the SPEEDES Simulation Framework with respect to the functionality found in the summer of 2001. Specifically this papers the scalability of SPEEDES with respect to its time management algorithms and simulation object event queues with respect to the number of objects simulated and events processed.

The Joint Modeling And Simulation System (JMASS) is a Tri-Service simulation environment that supports engineering and engagement-level simulations. As JMASS is expanded to support other Tri-Service domains, the current set of modeling services must be expanded for High Performance Computing (HPC) applications by adding support for advanced time-management algorithms, parallel and distributed topologies, and high speed communications. By providing support for these services, JMASS can better address modeling domains requiring parallel computationally intense calculations such clutter, vulnerability and lethality calculations, and underwater-based scenarios. A risk reduction effort implementing some HPC services for JMASS using the SPEEDES (Synchronous Parallel Environment for Emulation and Discrete Event Simulation) Simulation Framework has recently concluded. As an artifact of the JMASS-SPEEDES integration, not only can HPC functionality be brought to the JMASS program through SPEEDES, but an additional HLA-based capability can be demonstrated that further addresses interoperability issues. The JMASS-SPEEDES integration provided a means of adding HLA capability to preexisting JMASS scenarios through an implementation of the standard JMASS port communication mechanism that allows players to communicate.

The Air Force is developing a Distributed Information Enterprise Modeling and Simulation (DIEMS) framework under sponsorship of the High Performance Computer Modernization Office Common High Performance Computing Software Support Initiative (HPCMO/CHSSI). The DIEMS framework provides a design analysis environment for deployable distributed information management systems. DIEMS establishes the necessary analysis capability allowing developers to identify and mitigate programmatic risk early within the development cycle to allow successful deployment of the associated systems. The enterprise-modeling framework builds upon the Synchronous Parallel Environment for Emulation and Discrete-Event Simulation (SPEEDES) foundation. This simulation framework will utilize 'Challenge Problem' class resources to address more than five million information objects and hundreds of thousands of clients comprising the future information based force structure. The simulation framework will be capable of assessing deployment aspects such as security, quality of service, and fault tolerance. SPEEDES provides an ideal foundation to support simulation of distributed information systems on a multiprocessor platform. SPEEDES allows the simulation builder to perform optimistic parallel processing on high performance computers, networks of workstations, or combinations of networked computers and HPC platforms.

A metamodel is relatively small, simple model that approximates the behavior of a large, complex model. A common and superficially attractive way to develop a metamodel is to generate data from a number of large-model runs and to then use off-the-shelf statistical methods without attempting to understand the models internal workings. This paper describes research illuminating why it is important and fruitful, in some problems, to improve the quality of such metamodels by using various types of phenomenological knowledge. The benefits are sometimes mathematically subtle, but strategically important, as when one is dealing with a system that could fail if any of several critical components fail. Naive metamodels may fail to reflect the individual criticality of such components and may therefore be quite misleading if used for policy analysis. Na*ve metamodeling may also give very misleading results on the relative importance of inputs, thereby skewing resource-allocation decisions. By inserting an appropriate dose of theory, however, such problems can be greatly mitigated. Our work is intended to be a contribution to the emerging understanding of multiresolution, multiperspective modeling (MRMPM), as well as a contribution to interdisciplinary work combining virtues of statistical methodology with virtues of more theory-based work. Although the analysis we present is based on a particular experiment with a particular large and complex model, we believe that the insights are more general.

CACI, Inc.-Federal has built, tested, and demonstrated the use of a JMASS-JWARS HLA Federation that supports multi- resolution modeling of a weapon system and its subsystems in a JMASS engineering and engagement model environment, while providing a realistic JWARS theater campaign-level synthetic battle space and operational context to assess the weapon system's value added and deployment/employment supportability in a multi-day, combined force-on-force scenario. Traditionally, acquisition analyses require a hierarchical suite of simulation models to address engineering, engagement, mission and theater/campaign measures of performance, measures of effectiveness and measures of merit. Configuring and running this suite of simulations and transferring the appropriate data between each model is both time consuming and error prone. The ideal solution would be a single simulation with the requisite resolution and fidelity to perform all four levels of acquisition analysis. However, current computer hardware technologies cannot deliver the runtime performance necessary to support the resulting extremely large simulation. One viable alternative is to integrate the current hierarchical suite of simulation models using the DoD's High Level Architecture in order to support multi- resolution modeling. An HLA integration eliminates the extremely large model problem, provides a well-defined and manageable mixed resolution simulation and minimizes VV&A issues.

The Mixed Resolution Modeling Aide (MRMAide) technology is an effort to semi-automate the implementation of Mixed Resolution Modeling (MRM). MRMAide suggests ways of resolving differences in fidelity and resolution across diverse modeling paradigms. The goal of MRMAide is to provide a technology that will allow developers to incorporate model components into scenarios other than those for which they were designed. Currently, MRM is implemented by hand. This is a tedious, error-prone, and non-portable process. MRMAide, in contrast, will automatically suggest to a developer where and how to connect different components and/or simulations. MRMAide has three phases of operation: pre-processing, data abstraction, and validation. During pre-processing the components to be linked together are evaluated in order to identify appropriate mapping points. During data abstraction those mapping points are linked via data abstraction algorithms. During validation developers receive feedback regarding their newly created models relative to existing baselined models. The current work presents an overview of the various problems encountered during MRM and the various technologies utilized by MRMAide to overcome those problems.

Modeling of real systems relies on the arduous task of describing the physical phenomena in terms of mathematical models, which often require excessive amounts of computation time when used in simulations. In the last few years there has been a growing acceptance of model abstraction whose emphasis rests on the development of more manageable models. Abstraction refers to the intelligent capture of the essence of the behavior of a model, without all the details. In the past, model abstraction techniques have been applied to complex models, such as Advanced Low Altitude Radar Model (ALARM) to simplify analysis. The scope of this effort is to apply model abstraction techniques to ALARM; a DoD prototype radar model for simulating the volume detection capability of low flying targets within a digitally simulated environment. Due to the complexity of these models it is difficult to capture and assess the relationship between the model parameters and the performance of the simulation. Under this effort ALARM parameters were modified and/or deleted and the impact on the simulation run time assessed. In addition, several meta-models were developed and used to assess the impact of ALARM parameters on the simulation run time. This paper establishes a baseline for ALARM from which additional meta-models can be compared and analyzed.

The Air Force Hierarchy of Models, often referred to as the Great Pyramid, depicts the four disparate levels of resolution in which models are typically categorized. These levels range from an Engineering/Component level at the bottom, to Theater/Campaign level at the apex of the pyramid. Today, the landscape of simulations has evolved from uni-purpose, stove-piped simulations to those that provide a Joint Vision encompassing a much broader scope. Within the simulation community, there exists the desire for model reuse, particularly when it involves the reuse of validated legacy codes. Much effort has been put forth to integrate existing models into a federated system. Integrating models of similar resolution is difficult enough; yet, even more difficult is the more prevalent situation where models are represented at different levels of resolution. Often referred to as Mixed Resolution Modeling (or Multiresolution Modeling), it is arguably the most pressing problem facing the simulation research community today. This paper will describe an attempt to address the MRM problem by applying model abstraction techniques to reduce the complexity of a detailed model without sacrificing the essence of the model. This surrogate version of the detailed model will then be able to play within a more aggregate simulation environment. To demonstrate, JSAF (Joint Semi- Automated Forces) will be used to simulate the behavior of models at both the detailed and abstract levels. The results will be compared to demonstrate the impact and utility of model abstraction.

Variable Resolution Modeling is a collection of techniques designed to make higher fidelity models operate dynamically with lower fidelity level simulations. Previously, we have illustrated an example of these techniques with DeLoRes, a set of software tools developed to perform the state variable analysis, data base management, functional abstraction, and algorithmic implementation. We previously showed two classes of functional abstraction, multivariate, multidimensional linear and nonlinear interpolation and neural network representations. Now we extend the functional abstraction techniques to include stochastic information in some of the state variables. We illustrate this process by using the mean value at each interpolation node and fit the data with various linear multivariate interpolation techniques and then illustrate hypothesis testing of the stochastic data for consistency with a Gaussian (normal) distribution and discuss how this can be used in variable resolution modeling.

Effects-Based Operations (EBO) is a mindset, a philosophy and an approach for planning, executing and assessing military operations for the effects they produce rather than the targets or even objectives they deal with. An EBO approach strives to provide economy of force, dynamic tasking, and reduced collateral damage. The notion of EBO is not new. Military Commanders certainly have desired effects in mind when conducting military operations. However, to date EBO has been an art of war that lacks automated techniques and tools that enable effects-based analysis and assessment. Modeling and simulation is at the heart of this challenge. The Air Force Research Laboratory (AFRL) EBO Program is developing modeling techniques and corresponding tool capabilities that can be brought to bear against the challenges presented by effects-based analysis and assessment. Effects-based course-of-action development, center of gravity/target system analysis, and wargaming capabilities are being developed and integrated to help give Commanders the information decision support required to achieve desired national security objectives. This paper presents an introduction to effects-based operations, discusses the benefits of an EBO approach, and focuses on modeling and analysis for effects-based strategy development. An overview of modeling and simulation challenges for EBO is presented, setting the stage for the detailed technical papers in the subject session.

Models of adversaries' reasoning can be constructed to inform development of adaptive strategies, including strategies that include effects-based operations. Such models can apply to individual leaders or to groups that one seeks to influence. This paper describes an approach to building such models. The results are top-down, highly structured, driven by theory, designed with multiresolution methods that permit zooming in on issues, and suitable for use in high-level decision meetings. The models have been qualitative and non-automated, but the methodology could usefully be incorporated into a more general computer-supported decision-support environment, where it would supplement other tools for decision support.

Operational Concepts for planning, executing and assessing Effects-Based Operations (EBO) have been evolving over the past 5 years. The EBO concept is based on relating actions in a battle plan to overall effects. Evaluating these relationships requires the synthesis of a number of modeling approaches. A prototype system to assist in developing Courses of Action (COAs) for Effects-Based Operations and evaluating them in terms of the probability of achieving the desired effects has been developed and is called CAESAR II/EB. The tool supports both static and dynamic evaluation of COAs by integrating influence nets with discrete event systems modeling techniques. One challenge being addressed is how to use this type of modeling capability in real world command and control environments. Preliminary operational concepts were tested during the Naval War College Global 2000 and 2001 war games. The organizational location of the modeling team and its interaction with the multiple command and control cells in the game using network centric collaborative tools was tested and examined. Experiences with building and using the models as a decision support system for the war game are described. In addition requirements for enhancements to the modeling techniques generated from this experience are discussed.

Effects-Based Operations (EBO) is an approach to planning, executing, and assessing military operations where lower-level specific actions are derived from higher-level desired effects. Implementing EBO will require decision support capabilities that can model complex interactions and relationships, and then present the results to decision makers. Sponsored by the Air Force Research Laboratory Information Directorate (AFRL/IF), Science Applications International Corporation (SAIC) is developing a prototype collaborative Course-Of-Action (COA) decision-support system for real-time analysis of multiple possible COAs against multiple possible enemy COAs. This tool builds upon the SAIC-developed Geospatial Force Planning Tool (GFPT). GFPT is a distributed visual planning and collaboration tool within the Adaptive COA tool suite. GFPT allows geographically dispersed planners representing different planning echelons to collectively create, modify, and view a visual representation of an operational plan on top of a digital map. GFPT is being extended to support entry and visualization of both blue and red COAs. Using a COA-versus-COA simulation, expected results of various combinations of blue and red COAs will be analytically compared and visually presented. This paper will discuss the development process, the architecture, and the current results of this effort.

This paper describes the experimental approach for understanding requirements needed to satisfy the objectives of the Joint Synthetic Battlespace (JSB) for Simulation Based Acquisition (SBA). The JSB (SBA) is an envisioned system comprised of constructive and virtual simulations and supportive tools and databases which will provide an analysis capability that supports the full acquisition life cycle of a weapon system. The challenge for developing a robust capability is to clearly understand the requirements of the JSB (SBA) that can support a broad number of different types of users, which includes not only government organizations but also industry, laboratories, and academia. A key element in understanding the JSB (SBA) requirements is to implement a series of prototypes prior to attempting to develop the actual JSB (SBA). The prototypes will not only provide insight into the JSB (SBA) requirements but will also allow us to understand whether existing technologies and legacy systems are adequate for attaining the envisioned JSB (SBA) objectives or whether a paradigm shift in our current approach to simulation developments today is needed.

Military air operations in the European theater require U.S. and NATO participants to send various mission experts to 10 Combined Air Operations Centers (CAOCs). Little or no training occurs prior to their arrival for tours of duty ranging between 90 days to 3 years. When training does occur, there is little assessment of its effectiveness in raising CAOC mission readiness. A comprehensive training management system has been developed that utilizes traditional and web based distance-learning methods for providing instruction and task practice as well as distributed simulation to provide mission rehearsal training opportunities on demand for the C2 warrior. This system incorporates new technologies, such as voice interaction and virtual tutors, and a Learning Management System (LMS) that tracks trainee progress from academic learning through procedural practice and mission training exercises. Supervisors can monitor their subordinate's progress through synchronous or asynchronous methods. Embedded within this system are virtual tutors, which provide automated performance measurement as well as tutoring. The training system offers a true time management savings for current instructors and training providers that today must perform On the Job Training (OJT) duties before, during and after each event. Many units do not have the resources to support OJT and are forced to maintain an overlap of several days to minimally maintain unit readiness. One CAOC Commander affected by this paradigm has advocated supporting a beta version of this system to test its ability to offer training on-demand and track the progress of its personnel and unit readiness. If successful, aircrew simulation devices can be connected through either Distributed Interactive Simulation or High Level Architecture methods to provide a DMT-C2 air operations training environment in Europe. This paper presents an approach to establishing a training, testing and decision aid capability and means to assess its effectiveness within a comprehensive distributed Mission Training System (MTS).

This paper presents an exploration of the Simulation Based Research and Development through an in-house technology assessment of a Sensorcraft concept. The goals of SBR&D are: to reduce the time and cost for developing and maturing promising technology, to integrate the technologist and the warfighter into the Science and Technology (S&T) acquisition process, and to provide analytical input into the Air Force S&T planning process. SBR&D combines a variety of critical research and technology-development capabilities, including engineering-level modeling, design, and analysis tools, mission- and campaign-level simulations, cost analysis tools, and database tools in a networked, distributed environment. Early SBR&D capabilities combine high fidelity manned and unmanned vehicle simulations to create a common synthetic battlespace for technology assessment in a mission environment. The simulation environment is being combined with engineering models, design tools, and an intelligent database to allow differing degrees of fidelity to be used at different times and in different parts of a simulation analysis. The study presented here represents an attempt to show the SBR&D process in action and to identify deficiencies in the process. Once established, the SBR&D process will provide the capability for researchers to evaluate the impact of different technologies in a warfighting environment, providing a link between AFRL technologies and warfighter mission needs.

The past decade has produced significant changes in the conduct of military operations: increased humanitarian missions, asymmetric warfare, the reliance on coalitions and allies, stringent rules of engagement, concern about casualties, and the need for sustained air operations. Future mission commanders will need to assimilate a tremendous amount of information, make quick-response decisions, and quantify the effects of those decisions in the face of uncertainty. Integral to this process is creating situational assessment-understanding the mission space, simulation to analyze alternative futures, current capabilities, planning assessments, course-of-action assessments, and a common operational picture-keeping everyone on the same sheet of paper. Decision support tools in a distributed collaborative environment offer the capability of decomposing these complex multitask processes and distributing them over a dynamic set of execution assets. Decision support technologies can semi-automate activities, such as planning an operation, that have a reasonably well-defined process and provide machine-level interfaces to refine the myriad of information that is not currently fused. The marriage of information and simulation technologies provides the mission commander with a collaborative virtual environment for planning and decision support.

The Intelligence Data Object Server (IDOS) has been developed under the Air Force Research Laboratory Global Information Base Branch (AFRL/IFED) Global Awareness Virtual Testbed project to provide automated mechanisms for using Military Intelligence Data in modeling and simulation experiments. The IDOS software allows information from multiple data sources to be published in exercises using the High Level Architecture (HLA) or other object-oriented formats. IDOS uses the AFRL/IFEB Broadsword Gatekeeper for data source access. IDOS has been used in simulation-based acquisition experiments designed and carried out among distributed AFRL sites. This paper describes the IDOS architecture and capabilities including the use of the eXtensible Markup Language (XML) to provide a common representation for data objects, and application of IDOS to visualization of Intelligence Information.

The visualization of and interaction with decision quality information is critical for effective decision makers in today's data rich environments. The generation and presentation of intuitively meaningful decision support information is the challenge. In order to investigate various visualization approaches to improve the timeliness and quality of Commander decisions, a robust, distributed virtual simulation environment, based on AFRL's Global Awareness Virtual Testbed (GAVTB), is being developed to represent an Air Operations Center (AOC) environment. The powerful Jview visualization technology is employed to efficiently and effectively utilize the simulation products to experiment with various decision quality representations and interactions required by military commanders.

Information fusion is a critical problem for science and engineering. There is a need to fuse information content specified as either data or model. We frame our work in terms of fusing dynamic and geometric models, to create an immersive environment where these models can be juxtaposed in 3D, within the same interface. The method by which this is accomplished fits well into other eXtensible Markup Language (XML) approaches to fusion in general. The task of modeling lies at the heart of the human-computer interface, joining the human to the system under study through a variety of sensory modalities. I overview modeling as a key concern for the Defense Department and the Air Force, and then follow with a discussion of past, current, and future work. Past work began with a package with C and has progressed, in current work, to an implementation in XML. Our current work is defined within the RUBE architecture, which is detailed in subsequent papers devoted to key components. We have built RUBE as a next generation modeling framework using our prior software, with research opportunities in immersive 3D and tangible user interfaces.

Relatively recent advances in computer technology enable us to create three-dimensional (3D) dynamic models and simulate them within a 3D web environment. The use of such models is especially valuable when teaching simulation, and the concepts behind dynamic models, since the models are made more accessible to the students. Students tend to enjoy a construction process in which they are able to employ their own cultural and aesthetic forms. The challenge is to create a language that allows for a grammar for modeling, while simultaneously permitting arbitrary presentation styles. For further flexibility, we need an effective way to represent and simulate dynamic models that can be shared by modelers over the Internet. We present an Extensible Markup Language (XML)-based framework that will guide a modeler in creating personalized 3D models, visualizing its dynamic behaviors, and simulating the created models. A model author will use XML files to represent geometries and topology of a dynamic model. Model Fusion Engine, written in Extensible Stylesheet Language Transformation (XSLT), expedites the modeling process by automating the creation of dynamic models with the user-defined XML files. Modelers can also link simulation programs with a created model to analyze the characteristics of the model. The advantages of this system lie in the education of modeling and simulating dynamic models, and in the exploitation of visualizing the dynamic model behaviors.

SimPackJ/S is the JavaScript and Java version of SimPack, which means SimPackJ/S is a collection of JavaScript and Java libraries and executable programs for computer simulations. The main purpose of creating SimPackJ/S is that we allow existing SimPack users to expand simulation areas and provide future users with a freeware simulation toolkit to simulate and model a system in web environments. One of the goals for this paper is to introduce SimPackJ/S. The other goal is to propose translation rules for converting C to JavaScript and Java. Most parts demonstrate the translation rules with examples. In addition, we discuss a 3D dynamic system model and overview an approach to 3D dynamic systems using SimPackJ/S. We explain an interface between SimPackJ/S and the 3D language--Virtual Reality Modeling Language (VRML). This paper documents how to translate C to JavaScript and Java and how to utilize SimPackJ/S within a 3D web environment.

There exist various model types to represent dynamic systems but they arení»t generally reused for new modeling methods. The DXL (Dynamic eXchange Language) represents system using a simple block diagram defined by XML (eXtensible Markup Language), where each block has codes for either JavaScript or Java. The DXL were designed for being parsed from various existing models represented by MXL (Multimodel eXchange Language), and plays a role of basic unit layer for simulating and modeling in the rube. Models denoted by this DXL are produced to actual simulation codes used in rube through a translator using DOM (Document Object Model). These simulation codes use a SimpackJ/S toolkit as a target library for simulation.

The framework for JView, a Java based runtime re-configurable simulation visualizer, was described in two previous publications1. Many augmentations and substitutions have taken place in the JView API, brought about by working closely with customers from various agencies as well using the API on internal projects. However, the core mantra that JView is based upon has made it through these alterations unscathed and with more concrete proof of its utility. JView demystifies the world of 3D graphics programming, allowing users to concentrate solely on the task of visualization instead of concentrating their efforts on the art of complicated 3D graphics. It is a cross platform technology that is engineered to save time, money, and effort while meeting a variety of visualization needs. Its Java implementation, which provides cross-platform functionality while utilizing the OpenGL API, allows for platform dependent hardware acceleration. This paper contains concepts that the JView architecture utilizes as well as a brief introduction to its new 2D engine concepts.

The purpose of Scientific Visualization is to provide qualitative information by means of graphic methods, bringing quick and intuitive data interpretation. In stationary vector flows and non-linear and chaotic dynamic systems, visualization is specially useful due to the impossibility to find closed analytic solutions to the system's behavior. Numerous methods have been proposed for representing vector fields, of which streamlines an LIC are the most important. In streamlines, we perform a numerical evaluation of the flow by means of trajectories that are tangent to the vector field. This technique is adequate for real-time visualization, but has several drawbacks in accuracy. On the other hand, the technique known as LIC (line integral convolution) is a texture-based method, in which the trajectories originated at every pixel in the phase portrait, are advected along an input texture to find the final color of that pixel. The LIC overcomes most of the drawbacks of the streamlines, but is too slow to be useful as a real-time visualization tool. In this work we will show a new visualization method, called CLIC (cumulative line integral convolution), which effectively combines the advantages of streamlines and LIC. We further discuss the implementation of a real-time visualization tool that is adequate for bringing an Internet based visualization service.

Our LG-based overall model of a battlespace stems from a new concept of the LG hypergames. A hypergame is a collection of several inter-linked concurrent abstract board games (ABG). It may include a number of military ABG with different boards of various space-time resolutions. A move in one of the ABG may change the state of the rest of the ABG included in the hypergame. For example, the WWII Normandy Invasion hypergame includes three ABG: Allies' troop-ships invasion through the English Channel, Battle on the Normandy beaches, and Allied Air Force missions to cut German supply routs. In addition, an LG hypergame may include non-military games, such as transportation, politics, or economy. Multiple ABG included in the hypergame are separate games bounded by their boards and pieces, though linked through the mappings. However, LG zones may stretch through these boundaries to cover several ABG. As a result, the LG strategies for the players in a hypergame fully account for drastic mutual influence of multiple subgames included in the hypergame.

The LG-WGT approach to EBO may be summarized as follows. 1) Causes and Effects will be defined as game state properties. 2) LG algorithms will automatically generate strategies to attain desired effects. The strategies will be generated through LG Zones. LG will model effects as properties of the game pieces and relations among the pieces and the board. 3) The overall Engagement Theater will be modeled as LG hypergame, that is several concurrent abstract board games (ABG) linked together via inter-linking mappings (ILM). LG will represent indirect effects in a related game linked with the game of interest via several ILMs. With LG-WGT, a commander will observe the entire operation as an omnipresent ghost with a virtual camera. He/she would be able to view the operation from the cockcpit of a fighter flying on a SEAD mission, from the cabin of an amphibious vehicle, through the periscope of an attack submarine, or from a virtual AWACS flying over the entire battlefield. Even a normally invisible element, like damages to adversarial infrastructure or political changes, will be made visible in virtual reality together with the chain of events causing this effect. The LG-WGT will provide explanation for all the decisions made employing probabilities of kill, integrated probabilities of survival, threshold for retreat, etc.

The LG approach permits a game representation of the Engagement Theater in two modes, called the planning mode and the execution mode. The planning mode assumes that the computations are to be performed well before the actual engagement and thus a sophisticated (but slower) planning component of the LG hypergame is to be activated. On the planning stage, the emphasis is on the following. 1) distributing the battlefield resources to optimize some battlefield related criteria. Important optimization criteria are battlefield related, ones such as probabilities to achieve desired ends and survivability of the friendly forces; 2) developing of preliminary strategies for the commander to achieve certain desired end; 3) playing several what if situations to develop alternative strategies. Whereas the COA options for the resource distribution game could influence the COA options for selection of the preliminary and alternative strategies, the strategies are used to validate the effectiveness of the resource distribution, thus creating a feedback loop tying together the resources and battlefield actions. On the execution stage, the emphasis is on protecting the commander from possible errors caused by the information blizzard. The commander could have so many things on his/her mind that some negative consequences of a COA being decided upon may be overlooked. A faster execution version of the LG-based tool is intended to catch the undesired consequence and warn the commander.

LG-CONSTRUCTOR is a knowledge acquisition and construction component intended to facilitate rapid deployment of the LG- based tools. LG-CONSTRUCTOR will supply warfighters at all levels with the patterns of games. A warfighter will be able to consider and visualize the most viable patterns and quickly convert them into the LG hypergame, most adequate to the current mission. The adequacy of construction will be tested on the fly by playing and re-playing semi-finished hypergames. In order to construct a hypergame, LG- CONSTRUCTOR will define the board (with specific space-time scale), the pieces, the variety of legal movements and other activities of pieces, additional gaming constraints that define legal moves like winning conditions, rules of engagement, abort conditions, etc. LG-CONSTRUCTOR will store a number of pattern-ABG and complete pattern-hypergames developed earlier. After deployment to the mission, the military personnel will be able to play an appropriate pattern-hypergame. During this play, a military analyst will dynamically adjust this game to the real state of affairs. The required knowledge acquisition by pattern-game playing and game adjustment will be controlled by LG-CONSTRUCTOR. Step-by-step, by interacting with an analyst, LG-CONSTRUCTOR will generate a new hypergame, a network of interlinked ABG. For this generation, it may combine a number of pattern-ABG and complete pattern-hypergames. LG-CONSTRUCTOR will be capable of the real time construction due to transparency of LG game representation and high computational efficiency of LG-Strategist, an LG strategy-generating component. It will assist LG-CONSTRUCTOR in testing new games. This construction by playing and adjusting certain game-patterns will allow rapid deployment of LG-STRATEGIST during the mission. In this paper, we will consider the details of required knowledge acquisition and construction.

Asymmetric operations represent conflict where one of the sides would apply military power to influence the political and civil environment, to facilitate diplomacy, and to interrupt specified illegal activities. This is a special type of conflict where the participants do not initiate full-scale war. Instead, the sides may be engaged in a limited open conflict or one or several sides may covertly engage another side using unconventional or less conventional methods of engagement. They may include peace operations, combating terrorism, counterdrug operations, arms control, support of insurgencies or counterinsurgencies, show of force. An asymmetric conflict can be represented as several concurrent interlinked games of various kinds: military, transportation, economic, political, etc. Thus, various actions of peace violators, terrorists, drug traffickers, etc., can be expressed via moves in different interlinked games. LG tools allow us to fully capture the specificity of asymmetric conflicts employing the major LG concept of hypergame. Hypergame allows modeling concurrent interlinked processes taking place in geographically remote locations at different levels of resolution and time scale. For example, it allows us to model an antiterrorist operation taking place simultaneously in a number of countries around the globe and involving wide range of entities from individuals to combat units to governments. Additionally, LG allows us to model all sides of the conflict at their level of sophistication. Intelligent stakeholders are represented by means of LG generated intelligent strategies. TO generate those strategies, in addition to its own mathematical intelligence, the LG algorithm may incorporate the intelligence of the top-level experts in the respective problem domains. LG models the individual differences between intelligent stakeholders. The LG tools make it possible to incorporate most of the known traits of a stakeholder, i.e., real personalities involved in the conflict with their specific individual style.

Robust exploitation of tracking and surveillance data will provide an early warning and cueing capability for military and civilian Law Enforcement Agency operations. This will improve dynamic tasking of limited resources and hence operational efficiency. The challenge is to rapidly identify threat activity within a huge background of noncombatant traffic. We discuss development of an Automated Anomaly Detection Processor (AADP) that exploits multi-INT, multi-sensor tracking and surveillance data to rapidly identify and characterize events and/or objects of military interest, without requiring operators to specify threat behaviors or templates. The AADP has successfully detected an anomaly in traffic patterns in Los Angeles, analyzed ship track data collected during a Fleet Battle Experiment to detect simulated mine laying behavior amongst maritime noncombatants, and is currently under development for surface vessel tracking within the Coast Guard's Vessel Traffic Service to support port security, ship inspection, and harbor traffic control missions, and to monitor medical surveillance databases for early alert of a bioterrorist attack. The AADP can also be integrated into combat simulations to enhance model fidelity of multi-sensor fusion effects in military operations.

Antenna diversity promises to provide high data rates with low error probabilities in wireless communication. An important step for realizing these promises centers around the design of efficient space-time codes for multiple-antenna links. The case of mobile communication is particularly challenging. The use of complex representations of finite algebraic groups has recently been shown by some researchers to be a promising avenue for the design of space-time codes. This paper provides some background on the interplay between group representation theory and wireless communication. Then, it examines the potential contributions of finite groups of Lie type and finite unitary geometry to space-time coding. Java technologies are used for the implementation of our numerical experimentation platform.

It has been postulated that organizations can be categorized into one of three perspectives that represent the mind-set of managers within organizations with respect to their organization and organizational learning. These are the normative, the developmental and the capability perspectives. Each of these reflects variations among organizational features such as the source of organizational learning, the timeframe for organizational learning and the relationship between organizational learning and organizational culture. However, much like the dynamics experienced by teams, i.e. various stages such as forming, norming, storming and performing, organizations can move through various learning stages, i.e. the three 'perspectives,' often stopping and restarting at different points in their cycles. This means that the three perspectives can be simply viewed as different modes of organizational learning. All organizations operate within one of the three perspectives all the time. And, the perspective through which the organization is best viewed at any point in time changes over time. Because organizations are complex, adaptive systems these modes can be mathematically represented using the output from a neural network model of complex, adaptive systems. This paper briefly describes the organizational science, the neural network model, and the mathematics required to determine critical points in these modes.

The Decision Integration and Support Environment (DISE) is a Bayesian network (BN) based modeling and simulation of the target nomination and aircraft tasking decision processes. DISE operates in event driven interactions with FTI's AOC model, being triggered from within the Time Critical Target (TCT) Operations cell. As new target detections are received by the AOC from off-board ISR sources and processed by the Automatic Target Recognition (ATR) module in the AOC, DISE is called to determine if the target should be prosecuted, and if so, which of the available aircraft should be tasked to attack it. A range of decision criteria, with priorities established off-line and input into the tool, are associated with this process. DISE, when running in its constructive mode, automatically selects the best-suited aircraft and assigns the new target. In virtual mode, with a human operator, DISE presents the user with a suitability ranked list of the available aircraft for assignment. Recent DISE enhancements are applying this concept to the prioritization and scheduling of ISR support requests from Users to support both latent and dynamic tasking and scheduling of both space-based and airborne ISR assets.