DIS AND HLA: CONNECTING PEOPLE, SIMULATIONS AND SIMULATORS IN THE MILITARY, SPACE AND CIVIL DOMAINS

Transcription

1 DIS AND HLA: CONNECTING PEOPLE, SIMULATIONS AND SIMULATORS IN THE MILITARY, SPACE AND CIVIL DOMAINS G.J. Jense, N.H.L. Kuijpers, and A.C.M. Dumay TNO Physics and Electronics Laboratory P.O. Box 96864, 2509 JG The Hague, The Netherlands Phone , Fax {jense kuijpers ABSTRACT During the past decade, requirements within the US defence community for simulation-based tools to support acquisition, planning, (team) training and analysis, have resulted in a standard for distributed simulation known as DIS (Distributed Interactive Simulation). The potential advantages of distributed simulation technologies are evident: increased flexibility, building on existing software and communications standards, maximisation of the use of existing simulation assets, and thus reduced costs. Although DIS was originally developed for military applications, the technology is well suited as a simulator interoperability standard for both the space domain and other civil application areas. The first international DIS experiments between space agencies have already taken place. As in other distributed computing areas, the military have expressed a need for standardisation on a higher abstraction level than the basic communication protocols such as DIS. This trend has led to a new standard, developed and supported by the US Defence Modelling and Simulation Office (DMSO) known as HLA (High Level Architecture). This paper describes three important activities: recent and current developments of the DIS and HLA standards, the potential benefits of DIS and HLA for the space domain, and issues related to spin-in, spin-off, and dual-use of distributed simulation technology and applications in the military and civil domains. Copyright 1997 by TNO Physics and Electronics Laboratory. Published by the American Institute of Aeronautics and Astronautics, Inc. with permission. Released to IAF/IAA/AIAA to publish in all forms. 1. INTRODUCTION Motivated by a requirement for simulation based tools to support acquisition, planning, training, and analysis efforts, the defence community - most notably the US Department of Defence (DoD) - has heavily invested in R&D programmes on distributed simulation. This has resulted in a suite of protocols, collectively known as DIS (Distributed Interactive Simulation) as a standard for interconnecting large numbers of heterogeneous simulators across local and wide area networks. The DIS protocols allow the combination of such diverse elements as real-time interactive human-in-the-loop simulators, computer generated autonomous agents, numerical simulations of physical processes, and live instrumented participants, to operate together in one shared synthetic environment. Several users in non-military domains have recognised the potential benefits of DIS, and have used it in their own distributed simulation applications. In order to address several shortcomings of the DIS standards, to enable the interoperability of simulators with widely different architectures, and to further promote the reuse of existing simulation assets, the US DoD has now committed to a successor of DIS, which is known as the High Level Architecture (HLA). A significant difference in the way the HLA standard is being developed, refined, and extended, as compared to the way DIS was developed, is that contributions from the civil domain are actively sought by the Defence Modelling and Simulation Office (DMSO), who are the co-ordinating agency for HLA developments. This has resulted in the use of technologies, originally developed in the civil 1

2 domain, for the implementation of HLA support software and tools, and in the early adaptation of HLA for civil applications. configuration interconnecting a number of different simulators. The remainder of this paper is organised as follows. Section 2 introduces the DIS and HLA concepts and terminology, and includes a comparison of these two interoperability standards. Section 3 describes TNO s Advanced Simulation Framework, an abstraction layer designed to facilitate the migration from DIS to HLA. Next, in Section 4, the relevance of DIS and HLA for international space programmes is indicated by presenting some possible applications and mentioning the R&D work to achieve standardised interoperability in the space domain. Finally, in Section 5, conclusions are drawn and spin-in/spin-off of the DIS/HLA technology is discussed. 2.1 The DIS approach 2. DIS AND HLA The underlying technical approach of the DIS protocols is based on the following principles 1 : Object/event architecture: all dynamic entities in the synthetic environment inform all other entities of their status and actions through the transmission of standard information packets, socalled Protocol Data Units (PDUs); Autonomy of simulation nodes: all events caused by an individual simulator are broadcast and available to all other simulators. It is the receiving simulator's responsibility to determine the effects on its own state caused by the event; Transmission of state change information only: simulators transmit only changes in the state of the entity they simulate. Continuous activities are transmitted at a reduced update rate. Each simulator locally extrapolates the last reported states of other simulators until the next state update is received. This is referred to as deadreckoning. These principles allow simulators to minimise network traffic, make optimal use of local processing power, and enter and leave the synthetic environment without disrupting it. Figure 1 shows a typical DIS Simulator 1 Scenario Manager Computer Simulation Monitor (3D Stealth) Generated Force DIS protocol Simulator 2 Logger ISDN interface Figure 1: Interconnecting simulators through DIS. 2.2 Strengths and weaknesses of DIS DIS is a proven concept and is extensively used as standard protocol for simulator interoperability. An advantage of the DIS concept is that all DIS compliant simulators can operate within one simulated exercise. Stand-alone simulators become open systems. However, DIS has some drawbacks. Three features in the underlying data transport mechanism cause problems. Firstly, messages can get lost or arrive in the wrong order due to the use of the UDP/IP protocol. Secondly, the messages sent are part of standardised, fixed-sized Protocol Data Units (PDUs), although generic PDUs exist to communicate any type of data. Finally, due to the broadcast mechanism, the scaleability is rather limited. In the case that simulation experiments have to be repeatable, reliable data transfer is crucial. However, since DIS was originally developed for training purposes, unreliability was never recognised as a problem. 2 The fixed-sized PDUs in combination with the deadreckoning algorithms are well-suited to communicate state-updates and events between simulators (short messages). However, PDUs are relatively large and entire PDUs need to be broadcast even if just one Ethernet ISDN Remote Sites 2

3 attribute is changed. Furthermore, PDUs are generally ill-suited to transfer environmental state updates, as these often involve the transfer of large amounts of data. For instance, if weather plays a role in a distributed simulation, changes in wind direction and force are broadcast as meshes of wind vectors. This has lead to the current situation where new PDUs are being defined for all kinds of data exchange. Basically, DIS is nothing more than a protocol. No requirements for the software architecture of DIS compliant simulators are defined. Software packages can be purchased which provide the programmer with primitives for DIS based communication. However, using this type of software, specific in-depth knowledge about DIS is still required. A basic software layer which takes care of the communication and with which a simulation can communicate on a higher level would be a solution for programmers who are not familiar with DIS. 2.3 The High Level Architecture Over the past years, the DIS protocol standard has provided a solid base for real-time distributed simulation. DIS has been widely accepted, both by industrial and governmental users. The Aggregate Level Simulation Protocol (ALSP), which supports event-driven simulation, has been successfully applied in the field of Wargaming. 3 In order to unify these domains and extend the scope of the existing standards, the US Defence Modelling and Simulation Office (DMSO) in 1995 initiated a development to establish a new standard for modelling and simulation. This new standard, called High Level Architecture (HLA), is currently being defined. The HLA development is targeted at two main objectives. The first is to facilitate interoperability amongst heterogeneous simulations, e.g., real-time virtual simulations, aggregate level event driven simulations, non real-time constructive and live instrumented entities. The second objective is to promote reuse of simulations and their components. HLA is an attempt to capture the best from both DIS and ALSP and provide a standard architecture for simulator software at the same time. Figure 2: Functional View of the High level Architecture.. The HLA standard promotes reuse of simulations and their components. It attempts to specify the general structure of the interfaces between simulations without making specific demands on the implementation of each simulation. The standard is being developed in a co-operative, consensus-based forum of developers. 4 In the HLA baseline documents four basic concepts are defined: 1. The Run-Time Infrastructure (RTI) is an implementation of a distributed operating system which will be the basis software layer for HLA applications. The software layer takes care of communication between all simulation models. 2. The Interface Specification is a formal, functional description of the interface between the HLA application and the Run-Time Infrastructure. 3. A set of technical rules is defined to which an HLA participant has to comply. 4. The Object Model Templates are standardised formats to define the functionality of simulation models and the interaction between models. In this way the capabilities of all simulation models are defined before the actual simulation takes place. 2.4 Differences between DIS and HLA 3

4 The following list summarises the major differences between the new HLA standard and its DIS predecessor: HLA is an architecture (allowing the use of different components, such as communication protocols), while DIS specifies one single protocol; HLA is much more scaleable than DIS, because network traffic is greatly reduced due to its declaration management facilities (publication/subscription); HLA supports a much wider range of simulations, e.g., real-time virtual (human-in-theloop) simulations, constructive (non real-time) simulations, aggregate level (event driven non real-time) simulations, and instrumented live entities, where DIS was chiefly targeted at the first category only; HLA decouples the simulation application from the underlying communication protocol. Thus, HLA simulations can be developed by experts in the application domain, rather than by DIS experts; HLA is based on efficient and reliable point-topoint communication (TCP/IP could be used) using multicast grouping, while DIS relies on broadcast communication (TCP/UDP); HLA still has to prove itself, whereas DIS already has a track record of numerous successful implementations and applications. These comparisons indicate that HLA will be a major step forward with respect to DIS, but also that the transition from DIS to HLA is going to be a gradual one. Next, we describe possible applications of DIS and HLA in international space programmes, and how TNO realises the migration from DIS to HLA in its military R&D programme for distributed simulation. 3. DIS/HLA FOR INTERNATIONAL SPACE PROGRAMMES As in the defence area, modelling and simulation are indispensable tools in the space domain. A recent paper provides an overview of the use of simulation for verification and testing tasks in space projects. 5 This paper mentions a number of tasks related to space systems engineering, such as: concept study; requirements analysis; design; system integration; system testing. In fact, the use of modelling and simulation in the space domain is increasing. This is motivated by several reasons: In the space domain there is a continuing trend toward more international collaboration, requiring the interoperability of independently developed simulation facilities as well as the distribution of project test-beds that currently are mostly centralised; Several phases of space projects rely on extensive verification and validation exercises by combining hardware-in-the-loop and software based simulation; Also in the space domain there is an increasing drive toward further cost reduction by promoting the reuse of existing simulation assets. Because of these reasons, the use of DIS and HLA technology in the space domain, and any other civil domain for which similar characteristics can be identified, should be seriously considered. On the other hand, space applications pose specific demands that are not addressed by the basic DIS and HLA standards. This would mean that additional research and development efforts are required in order for DIS and HLA to become usable in the space domain Space applications of DIS/HLA Team Training Although training to prepare for space missions usually involves only a few individuals (controllers, mission specialists, astronauts) as compared to military training of hundreds or sometimes thousands of soldiers, the fidelity requirements for the simulated environment are often higher than in defence applications. A reconfigurable project test-bed, consisting of highly accurate constructive simulations, hardware-in-theloop components, event-driven simulations as well as 4

5 real-time virtual simulations would allow the training of operational procedures. Initial training is possible at an early stage in the project, when the target space system is still in its conceptual phase. Results from these initial training exercises could feed into the requirements analysis phase. Later, as more and more components are designed and implemented, the training environment would consist of validated simulations of the final system components, including hardware. Basing the project test-bed on a standardised distributed architecture, for instance based on DIS or HLA, would in principle allow the test-bed components to be geographically dispersed, and to be easily adaptable as the project progresses. A first experiment in this area was performed in a joint exercise of ESA/ESTEC and GSTC, and has been reported in. 7 Collaborative engineering Another area where distributed simulation technology can potentially be used to great advantage is collaborative virtual environments. Collaborative VEs have been proposed for several applications, including training and design engineering. 8 A collaborative VE would allow a geographically dispersed team of engineers to use existing simulation and database assets and link them through a standard communication infrastructure. An existing distributed simulation infrastructure for the space community could also mean increased accessibility of spacecraft (subsystem) data, e.g., simulation models, while these are still at contractors sites. Spacecraft control A virtual environment for spacecraft operations would enable the monitoring of spacecraft operations by a team of ground control operators. Based upon the telemetry data that is received, the virtual environment system reflects the status, ongoing activities and changes of the spacecraft and its subsystems. Ground operators are able to monitor the spacecraft from within a shared virtual environment. The virtual environment gives each operator an appropriate view on relevant onboard subsystems that renders its status. The virtual environment enables the operator to monitor both the subsystem(s) he or she is responsible for, as well as to maintain insight in the overall status of the spacecraft and relevant environmental conditions. An architecture for a spacecraft control environment based on advanced distributed simulation concepts such as DIS or HLA, supports distribution of operators across geographically separated sites; distribution of simulation modules across geographically separated sites; separation between ground control site and space link site. 3.2 Technical issues of DIS/HLA for space applications A number of technical issues that are directly or indirectly relevant for the use of DIS and HLA technology in the space domain have already been tackled. The following descriptions provide a few examples of these issues. Dead reckoning of spacecraft entities An example of a technical issue that can be encountered in space applications of DIS is described in. 9 This paper examines the effects of using the standard DIS dead reckoning (DR) algorithms to predict the motion of satellites in orbit. Because the standard algorithms are based on linear motion in a Cartesian co-ordinate system, their accuracy for the prediction of curvilinear motion of spacecraft entities is generally not sufficient. The paper goes on to compare the accuracy of the standard DR algorithms with that of a simulation model based on orbital mechanics, and concludes with the recommendation to develop models for accurately dead reckoning spacecraft entities while conserving network bandwidth (a conclusion that was also reached by other authors. 7 ) Simulation of the space environment A second example is not related to the simulation of entities, but that of environmental aspects, i.e., processes and conditions in the space environment such as solar phenomena (sunspots, coronal mass ejections, solar flares), the interplanetary medium (solar wind), and phenomena of the earth s atmosphere (aurora, geomagnetic storms). 10 Accurate simulation of these 5

6 phenomena is required to be able to assess their influence on, e.g., propagation of communications signals, navigation systems, satellite performance, health hazards, etc. The paper also discusses the gathering of authoritative data from space monitoring services, existing space environment models, and data standards issues, and concludes with a call for more collaboration between the space, defence, and other communities on this subjects. 3.3 Future Work Based on the activities and current hot topics in the space domain that were described above, we can identify a number of steps that can be taken by the space community to take advantage of ongoing developments in the DIS/HLA arena. These steps involve both concrete technical R&D work, but also more policy oriented measures: Further development and standardisation of, e.g., dead reckoning algorithms for spacecraft entities. Some basic work has already been done for the DIS protocols DIS, and the need for additional work within the framework of the HLA has been identified. Active participation by the space community, both industry and the space agencies, in the ongoing HLA standardisation activities. This can be accomplished in various ways, e.g., by participation in the semi-annual Simulation Interoperability Workshops, but also by experimenting with a Space Federation for instance through the joint development of standard Federation Object Models and Simulation Object Models. The potential advantages of distributed simulation technologies, not limited to but including DIS and HLA, for a number of space applications are evident: increased flexibility; building on existing software and communications standards; maximisation of the use of existing simulation assets; reduced costs. Therefore, an active participation of the space community, either by using existing DIS and HLA standards and products, or by contributing to and influencing the ongoing HLA development activities is warranted. 4. DIS AND HLA IN THE CIVIL DOMAIN On the relationship between the civil and military domains in the area of distributes simulation technology, the following observation can be made. The DIS protocols were developed in the military domain, and later found various applications in the civil domain. For the development of the HLA standards, DMSO is actively seeking contributions from non-military users and technology developers. What we see here is that civil technologies are being incorporated in military standards and products, developers and end-users from the civil domain are co-steering the development of the HLA standards and the associated products, and that civil applications are much earlier adopting the HLA standards. This section describes examples of these issues, 4.1 Civil technology for military applications HLA software based on CORBA For the development of the first HLA Real Time Infrastructure (RTI) prototype, the development team employed a CORBA compliant Object Request Broker. 11 More specifically, the Orbix implementation of CORBA was used. The use of CORBA allowed the team to rapidly develop a fully functional RTI prototype, taking advantage of CORBA platform independent communication facilities. The early availability of the RTI prototype in turn allowed the concurrent development of prototype HLA federations and the verification of HLA interface specifications at an early stage. The downside of using CORBA turned out to be a performance degradation due to CORBA s communication overhead. CORBA is also being used for the development of HLA support software. The HLA Modelling and Simulation Resource Repository (MSRR) will 6

7 provide access to consistent, authoritative, and validated representation of Modelling and Simulation representations, e.g., real world operational entities, their actions and interactions, environmental aspects, etc. However, all of these resources store their data in different formats. In order to provide uniform access to these heterogeneous databases, a CORBA based framework, known as Catalyst, is being developed to support federation execution development. 12 Catalyst offers the simulation developer various means to access data, modify objects and relations, translate between data formats, etc. Services (ACTS) programme. 14 COVEN aims to develop a software platform for teleworking and virtual presence applications. Networking trials are performed between the COVEN partners, making use of the ISDN and ATM communication infrastructure. In the framework of COVEN, TNO-FEL participates in a study into the feasibility of using DIS and HLA concepts for the development of the COVEN platform. The results of the COVEN project will be relevant for a number of distributed simulation applications, including the ones mentioned in section 3.1. Use of open commercial standards for DIS/HLA applications In addition to being used for the development of distributed simulation technology, open commercial standards are also becoming increasingly important for the development of DIS/HLA applications. A recent paper reports on the use of a standard Windows/Intel platform, low cost graphics acceleration hardware, and an industrial standard Application Programmer s Interface (Microsoft s Direct3D), to produce a stealth viewing station (or Virtual reality Scene Generator). 13 In addition to the open commercial standards, the product also relies on various military modelling and simulation standards (DIS, CTDB terrain databases, the ModSAF computer generated forces (CGF) software, etc.), and this has resulted in a number of technical solutions for problems related to the combination of DIS/HLA interoperability standards and the industrial standard software. 4.2 Civil applications of DIS and HLA Simulation based design and manufacturing Design concept evaluation, an early stage in the system acquisition process, is an important application area for distributed simulation in the military domain. The use of parallel and distributed simulation, usually under the nomenclature of High Performance Computing and Networking (HPCN), have also been used for quite some time now in various civil markets. Problems related to the integration of simulations have led developers in the consumer electronics and heavy industry markets to choose an approach based on HLA. 15 There, distributed simulation in general, and DIS/HLA based simulations in particular are expected to become applicable for such diverse applications as conceptual design, assembly modelling, computational analysis, and training. After evaluating the RTI prototype implementation, which is based on CORBA (see section 4.1), the authors decided to develop their own version of the RTI, designed to meet their strict performance requirements. Collaborative Virtual Environments Currently the most advanced 3D user interfaces to collaborative distributed simulation applications are immersive virtual environments. Immersive VEs will probably pose severe communication requirements due to the nature of the generated network traffic. Additionally, there are a number of critical human factors that need to be assessed before this technology can be successfully applied. TNO-FEL is a partner in the COVEN project, carried out in the framework of the EU Advanced Communication Technology and Transport and logistics In the civil domain, transport and logistics are key elements in the overall effort to sustain a business or industrial operation. In today s modern markets, precise planning, reduced cycle times require precise planning of operations, and the trend towards global collaboration means an increase in interoperability requirements between information systems of a world-wide operating virtual company. Collaborative decision making amongst geographically dispersed business operators requires a shared situation awareness, built from 7

8 consistent, up-to-date, reliable and authoritative information. Distributed simulation facilities will play a key role in such collaborative decision aids, and DIS and HLA technology are prime candidates as supporting technologies for such applications. 16 Driving simulation The automotive industry generally recognises the potential value of driving simulation for safety enhancement, training, and as a research tool. To date however, driving simulators have mainly been employed in the last area. This is due to the reason that for almost all applications of driving simulation, vehicle control is the main focus, which puts several extremely high performance requirements on the simulator. This in turn tends to make the development (and running!) of driving simulators very expensive. In spite of this, the potential benefits of networking several driving simulators into one overall synthetic environment is being recognised. 17 Before such a distributed driving simulation can be implemented successfully, several technical issues need to be addressed. Two important issues are the development of dead-reckoning algorithms that can meet the accuracy and performance requirements of driving simulators, and the definition of scenarios that are specifically tailored to the driving task at hand. Civil Aviation Driven by its large installed base of simulators and the trend towards international collaboration, and even globalisation of the aviation industry, the civil aviation community has taken an active interest in simulator interoperability issues. The US Federal Aviation Administration (FAA) has, within the framework of its National Simulation Capability (NSC) program, developed a Formal Analysis Process (FAP, which is an analysis environment, based on HLA concepts, but adapted to meet a number of civil aviation requirements, to support R&D efforts aimed at improving traffic flow, enhancing safety, increasing capacity, and improving overall system performance. 18 One important element in the overall FAP, is the assessment of fidelity of individual simulators that are to interoperate in a particular scenario. The FAA has already developed an assessment technique that allows simulation designers to perform cost-fidelity trade-off studies. 19 In Europe, DIS has also been identified as an extremely useful technology to support the development of future Air Traffic Control concepts CONCLUSIONS We have presented an overview of the current state of the DIS standard and the development of its successor, the HLA. We have given a concise description of the major concepts behind both, and compared their strengths and weaknesses. The situation with respect to simulation in the space domain has been compared to that in the military field. In spite of there being a number of differences, several basic similarities have also been noted, including trends towards more international collaboration, the need for interoperability of heterogeneous simulators, and the drive toward cost reduction through reuse of existing simulation assets. All of these suggest an approach based on distributed simulation. Moreover, we have mentioned a number of applications in the space domain that exhibit technical characteristics and requirements that can potentially be addressed by standards for distributed simulation in some form. These similarities justify in our view a careful study of both the requirements and the capabilities of existing standards (DIS), and emerging architectures (HLA) for distributed simulation. We have also reviewed the spin-in and spin-off between the military and civil domains. The use of civil technology in the military domain and the use of DIS and HLA technology for civil applications have been illustrated by means of several examples that are currently being developed. Because of a common tradition of operating at the cutting edge of technology, and of similar requirements in applications such as team training, collaborative decision support, and project test beds, the transfer of DIS/HLA technology from the defence to the space domain seems a logical step, either through spin-in of military simulation technology or 8

9 through the development of dual-use technology. In a later stage, the exploitation of DIS/HLA technology in the space domain can serve as an example for transfer of the technology to other civil domains, and thus creating a spin-off situation. REFERENCES 1. DIS Steering Committee, The DIS Vision, A Map to the Future of Distributed Simulation, Version 1, Institute for Simulation & Training, May 1994, Orlando, Florida, USA, IST-SP Duncan C. Miller, The DOD High Level Architecture And The Next Generation Of DIS, 14th DIS Workshop on The Standards for the Interoperability of Distributed Simulation, Orlando, FL, March Weatherly, D. Seidel, and J. Weismann, Aggregate Level Simulation Protocol, Proc Summer Comp. Sim. Conf., pp , Baltimore, July Defence Modeling and Simulation Office, HLA Management Plan High Level Architecture for Modeling and Simulation, Version 1.6, 17 July DMSO homepage: 5. Melton and J.E. Miró, The role of simulation in space system verification, European Space Research and Technology Centre ESTEC, Noordwijk, The Netherlands. 6. Hans Jense and Nico Kuijpers, DIS - The Next Generation: A perspective for European space programmes, Proc. 4 th ESA Workshop on Simulators for European Space Programmes, Noordwijk, October Vankov et.al., Distributed Interactive Simulation of Rendezvous and Docking with International Space Station, Proc Fall Simulator Interoperability Workshop, Orlando, FL, September Nico Kuijpers and Hans Jense, Collaborative Engineering in Distributed Virtual Environments, Proceedings of the 1997 Spring Simulator Interoperability Workshop, Orlando, FL, March Martin R. Stytz, Dead reckoning of spacecraft entities in earth orbit, 14th DIS Workshop on The Standards for the Interoperability of Distributed Simulation, Orlando, FL, March James H. Hammond et.al., Modeling and Simulation of the Space Environment - An Overview, Proceedings of the 1997 Spring Simulator Interoperability Workshop, Orlando, FL, March McGarry et.al., Design Issues for the High Level Architecture (HLA) Run Time Infrastructure (RTI) Prototype version 0.2, Proc. 14 th DIS Workshop, Orlando, FL, pp , Bruce R. McQueary et.al., A CORBA based framework supporting the development of complex distributed simulations, Proceedings of the 1997 Spring Simulator Interoperability Workshop, Orlando, FL, March Garth Smith, A Virtual Reality Scene Generator through open Commercial Standards, Proceedings of the 1997 Spring Simulator Interoperability Workshop, Orlando, FL, March Véronique Normand and Jolanda Tromp, Collaborative Virtual Environments: the COVEN project, Proc. Framework for Immersive Virtual Environments (FIVE) 96, Pisa, Italy, December Masakazu Furuichi et.al., Design and Implementation of Experimental HLA-RTI without employing CORBA, Proc. 15 th DIS Workshop, Orlando, FL. 16. David Payne, Interoperable Systems for 21 st Century Logistics: Training, Planning, and Execution, Proceedings of the 1997 Spring Simulator Interoperability Workshop, Orlando, FL, March Reginald T. Welles and Darrell R. Turpin, When worlds collide - Considerations for integrating a driving simulator into a distributed interactive simulation network (DIS), Proc. 14 th DIS Workshop, Orlando, FL, pp , Paula Nouragas et.al., A formal analysis process based on the High Level Architecture, Proceedings of the 1997 Spring Simulator Interoperability Workshop, Orlando, FL, March Paula Nouragas et.al., Fidelity assessment in aviation related simulation applications, Proceedings 9

10 of the 1997 Fall Simulator Interoperability Workshop, Orlando, FL, September Peter Anthony Pennie, Distributed Interactive Simulation, Proc. ITEC 97, The 8th International Training and Education Conference, Lausanne, Switzerland,