Problem Statement

Arguably one of the most powerful problem-solving techniques available today, simulation is playing an increasingly valuable role within the United States Department of Defense (DoD). The creation and prominence of the Defense Modeling and Simulation Office (DMSO) and the Executive Council on Modeling and Simulation (EXCIMS) bear witness to this. Simulation is being applied within every facet of DoD activity: from analysis to training to acquisition. But simulation is far from a defense-specific technique. Simulation has its roots in a variety of disciplines including Operations Research, Management Science, Computer Science, Statistics and Mathematics, and is widely utilized in industrial and academic settings as well as government and DoD settings. Each of these groups will influence the directions of simulation support technology.

But before we try to guess where simulation support technology is going, we would perhaps be wise to first consider where it has been. Of course, a proper examination of this topic would yield a recitation filling several volumes. Recognizing the virtue of brevity -- and admittedly oversimplifying -- we describe the history of simulation support as comprised of 4 periods: (1) the advent (circa 1950 - 1960), (2) the era of simulation programming languages (circa 1960 - 1980), (3) the era of simulation support environments (circa 1980 - 1990), and (4) the modern era (circa 1990 - today). The advent begins with the inception of digital computers and is marked by early theories of model representation and execution. For example, the fixed-time increment and variable-time increment time flow mechanisms (TFMS) were proposed during this period. Generation of random behavior and methods for statistical evaluation of simulation results were also formulated during the advent. Once the general principles and techniques had been identified, a desire to formulate special purpose languages for simulation emerged. The proliferation of simulation programming languages (SPLs) like Simula, Simscript , GPSS and MODSIM mark the second period in this history. The design and evolution of SPLs also helped refine the principles underlying simulation. The dominant conceptual frameworks (or world views ) for discrete event simulation -- event scheduling, process interaction, and activity scanning -- were defined during the second period largely as a result of SPL research. Also during this period the first cohesive theories of simulation modeling were formulated, e.g. Zeigler's DEVS [5] (discrete event system specification) and its basis in general systems theory. The third period in the history of simulation support is evidenced by a shift of focus from the development of the simulation program toward the broader life cycle of a simulation study -- extending software and methodological support to such activities as problem formulation, objectives identification and presentation of simulation results. This was the era of the integrated simulation support environment (ISSE). Environment research occurred throughout the primary simulation communities: within academia (CASM, VSE, JADE), within industry (SIMKIT, SES Workbench , COMNET III ) and within the government (ROSS, KBSim, IMDE). Also during this third period a great interest emerged within the DoD in interoperable, networked simulators. Efforts in this vein resulted in SIMNET and later the DIS architecture and protocols.

The immediacy of the modern era makes it difficult to assess, but emphasis and direction seem to vary across the primary simulation communities. In the commercial sector environments and languages remain a major focus. The objective seems to be maximized market share through specialization; for example, there are at least 5 commercial environments specializing in communications network simulation. Within academia environments are also a focus, but these environments appear to favor generality of purpose over specialization. The majority of modeling methodology activity remains in the academic sector and most of the work involving the execution of simulation models on parallel computers is also occurring in university laboratories. Within DoD the focus on architectures and infrastructures continues as shrinking budgets demand maximum utilization of expenditures. Interoperability and reuse are the watchwords of the day. Accompanying this change of focus has been a paradigm shift in computer-assisted training toward large-scale, distributed training environments. Notable in this regard have been the DIS-supported Synthetic Theater of War (STOW) experiments and the Aggregate Level Simulation Protocol (ALSP) and its primary application the Joint Training Confederation (JTC). While the objectives underlying DIS and ALSP are similar -- the interoperation of distributed simulations -- the mechanisms they employ to achieve their objectives are significantly dissimilar. The High Level Architecture (HLA) is a recent initiative undertaken by DMSO to unify these approaches and satisfy Objective 1-1 of the DoD M&S Master Plan which calls for ``a common high-level simulation architecture to facilitate the interoperability of all types of models and simulations among themselves and with C4I systems, as well as to facilitate the reuse of M&S components.''

Is the HLA the future? Perhaps. But there is another factor to consider.

Balci and Nance [1] characterize the evolution of simulation support as reflecting both a ``needs push'' and a ``technology pull.'' The former is the familiar catalyst echoed in the adage that necessity is the mother of invention. The latter phenomenon is, perhaps, less intuitive and makes the case that invention can be the mother of necessity -- that is, technological innovation can offer fundamentally different, previously inconceivable, forms of meeting an existing need. The rise of World-Wide Web (WWW) portends a significant technology pull for simulation support. The potential of the WWW as a computing infrastructure continues to expand as a vast collection of ancillary technologies is developed to enhance and exploit the Web. Client/server architectures have proliferated resulting in new standards such as the Common Object Request Broker Architecture (CORBA). Using a design principle originally conceived to engender portability of Pascal compilers in the 1970s, the Web-based programming language, Java offers a common representation potentially suitable to bridge the diverse computing platforms that utilize (and compose) the WWW.

The potential of the WWW as a technology pull is cited in a recent article that describes the application of the WWW to the manufacturing process [2]: ``Our initial experiments at putting engineering, design, and manufacturing services on the Web are so successful that we believe we should rethink the traditional approaches and tools for coordinating large, distributed teams.'' With respect to simulation, a similar revolution seems plausible. Web technology has the potential to significantly alter the ways in which simulation models are developed (collaboratively, by composition), documented (dynamically, using multimedia), analyzed (open, widespread investigation) and executed (using ``massive'' distribution). Of course, as in life, there are no guarantees for the WWW. Page 9 of the same journal in which the previously cited article appears refers to a particular market forecast that predicts a dissipation of interest in the Internet in 1996.

So where is the technology of simulation support headed? It is difficult to say. The degree to which the WWW can, or should, play a role in this future is uncertain. But clearly the time is right for a meaningful, focused investigation of the subject. As Fishwick observes[3]:

We are in a kind of twenty-first century gold rush since everyone is rushing to put multimedia information on the web, but there is great uncertainty about how things should look and feel. How will simulation-related societies change and deliver their publications? Will companies expand their sales by offering simulation models of their products? One thing is for certain. We will never know any of these answers unless we experiment and take risks if necessary. The ``right way to do simulation'' will naturally emerge and evolve as we try various web-based simulation approaches. The worst possible approach is to sit back and adopt a ``let's wait'' attitude: simulation will be worse off as a discipline unless we move forward now to incorporate the web-based technologies.

Objectives

The strategic objective of this research is to:

• Influence the directions of the WWW technology pull through participation in the Object Management Group (OMG) and Simulation Interoperability Standards Organization (SISO) standards activities and anticipation of DoD needs beyond HLA.

The technical objectives of this research are to:

• Identify the WWW technologies that could be leveraged in a simulation support role and classify their use. For example, a preliminary taxonomy might include:
  • WWW as a publishing medium. The Hypertext Markup Language (HTML) provides both document formatting and direct linkage to other documents. These capabilities may yield significant improvements in the processes of information acquisition and presentation to both the model developer and model user.
  • WWW as a heterogeneous document store. The WWW has adopted the concept of hypermedia . A ``document'' on the WWW may contain graphical images, video, audio -- potentially, any conceivable form of information -- in addition to text. Using so-called ``helper applications'' and ``plug-ins'' the myriad formats and encoding schemes can be reconciled without end-user intervention. Clearly, the power of hypermedia could profoundly improve traditional approaches to model development and model use. Evidence of this exists in several of the recent simulation support environment efforts, as well as the few existing web-based simulation efforts.
  • WWW as a user interface. To varying degrees, most simulation support environments attempt to insulate the modeler from programming-language-level descriptions. Many of these environments adopt a dialogue-driven approach that dates to the first program generator, Programming by Questionnaire (PBQ), developed at RAND in 1966 [4]. The forms capability of HTML provides a simple mechanism to construct a graphical user interface (GUI) that could be used to support the ``extraction'' of a model specification from a modeler. GUIs to support model configuration and execution are similarly supported.
  • WWW as a software delivery mechanism and distributed computation engine. Perhaps of most relevance to this research effort, we observe that the Internet is (among other things) the world's largest ``distributed computer.'' Efforts like CORBA are enhancing the mechanisms through which a user can invoke and utilize services -- computer programs -- resident on remote machines. However, simultaneously satisfying a large (potentially unbounded) number of requests can overburden even the most computationally powerful servers. Recent efforts like Java and Safe-Tcl enable safe and efficient execution of service software on the client machine. Current indications are that CORBA, Java and web browsers like Netscape will evolve toward a unified, distributed object environment.
• Using the taxonomy resulting from the effort above, formulate a ``vision'' for web-based simulation support. How could simulation model definition, specification, analysis, use and evaluation be effected using web-based technologies? How should these technologies be applied? In all likelihood, each aspect of the taxonomy above could be realized within a simulation support environment to various degrees of success. Although the resulting design space is infinite, a set of meaningful design alternatives should be articulated.
• Using the requirements for simulation support provided by real-world FAA and DoD examples (e.g. collaborative air traffic management, and rapid response mission rehearsal) select a suitable design from the alternatives identified above.
• Identify the methodological requirements to support the FAA and DoD applications.
• Identify the services necessary to support distributed simulation. These services may be exactly those defined for the HLA, or modifications and extensions of the HLA services.
• Prototype as necessary and demonstrate the feasibility of applying web-based technology to the technical challenges faced by FAA and DoD.

References

1

Balci, O. and Nance, R.E. (1988). Simulation Model Development: The Multidimensionality of the Technology Pull, Technical Report SRC-88-012, Systems Research Center, Virginia Tech, Blacksburg, VA, November.

2

Erkes, J.W. et al. (1996). Implementing Shared Manufacturing Services on the World-Wide Web, Communications of the ACM, 39(2), pp. 34-45, February.

3

Fishwick, P.F. (1996). Web-Based Simulation: Some Personal Observations, to appear in: Proceedings of the 1996 Winter Simulation Conference, Coronado, CA, 8-11 December.

4

Oldfather, P.M., Ginsberg, P.L. and Markowitz, H.M. (1966). Programming by Questionnaire: How to Construct a Program Generator, RAND Report RM-5129-PR, November.

5

Zeigler, B.P. (1976). Theory of Modelling and Simulation, John Wiley and Sons, New York.

Endnotes

...periods:

Since the focus of our investigation is computer-based support for simulation, the history presented is limited to the age of digital computers. The use of simulators predates the advent of digital computers, with the first flight simulator being credited to Edward A. Link in 1927.

...include:

This taxonomy is used in [2]:

Ernest H. Page 21 October 1996