Live, virtual, and constructive

Live, Virtual, and Constructive (LVC) Simulation is a broadly used taxonomy for classifying Models and Simulation (M&S). However, categorizing a simulation as a live, virtual, or constructive environment is problematic since there is no clear division between these categories. The degree of human participation in a simulation is infinitely variable, as is the degree of equipment realism. The categorization of simulations also lacks a category for simulated people working real equipment. The LVC categories are defined as follows:

• Live - M&S involving real people operating real systems, e.g. a pilot flying a jet.

• Virtual - M&S involving real people operating simulated systems. Virtual simulations inject a Human-in-the-Loop into a central role by exercising motor control skills (e.g., flying a simulated jet), decision making skills (e.g., committing fire control resources to action), or communication skills (e.g., as members of a C4I team).

• Constructive - M&S involving simulated people operating simulated systems. Real people stimulate (provide inputs) to such simulations, but are not involved in determining the outcomes. See Cognitive model.

Other associated terms are as follows:

• LVC Enterprise - The overall enterprise of resources in which LVC activities take place.
• LVC Integration - The process of linking LVC simulations through a suitable technology or protocol to exploit simulation interoperability within a federated simulation environment such as the high level architecture (simulation) or Eurosim.
• LVC Integrating Architecture (LVC-IA) - The aggregate representation of the foundational elements of a LVC Enterprise, including hardware, software, networks, databases and interfaces, policies, agreements, certifications/accreditations and business rules. LVC-IA is intrinsically an Enterprise Architecture, given the system-of-systems environment it must support.

LVC-IA bridges M&S technology to the people who need and use the information gained through simulation. To accomplish this a LVC-IA provides the following:

• Integration through simulation equipment, interoperability tools and support personnel. See also Enterprise integration and Enterprise architecture. Integration creates network-centric linkages to collect, retrieve and exchange data among live instrumentation, virtual simulators and constructive simulations as well as between the joint military and specific service command systems. Integration also bridges together data management, exercise management, exercise collaboration and updating training support systems.
• Interoperability through common protocols, specifications, standards and interfaces to standardize LVC components and tools for mission rehearsals and training, testing, acquisition, analysis, experimentation, and logistics planning.
• Composability through common and reusable components and tools such as enterprise after action review, adapters, correlated terrain databases, Multilevel security for multinational players and hardware/software requirements.

History

LVC Simulation Architectures Venn Diagram

Usage Frequency of Simulation Architectures

Prior to 1990, the field of M&S was marked by fragmentation and limited coordination between activities across key communities. In recognition of these deficiencies, Congress directed the Department of Defense (DoD) to “... establish an Office of the Secretary of Defense (OSD) level joint program office for simulation to coordinate simulation policy, to establish interoperability standards and protocols, to promote simulation within the military departments, and to establish guidelines and objectives for coordination [sic] of simulation, wargaming, and training.” (ref Senate Authorization Committee Report, FY91, DoD Appropriations Bill, SR101-521, pp. 154-155, October 11, 1990) Consistent with this direction, the Department of Modleing and Simulation Office (DMSO) was created, and shortly afterwards many DoD Components designated organizations and/or points of contact to facilitate coordination of M&S activities within and across their communities. For over a decade, the ultimate goal of the DoD in M&S is to create a LVC-IA to assembly models and simulations quickly, which create an operationally valid LVC environment to train, develop doctrine and tactics, formulate operational plans and assess warfighting situations. A common use of these LVC environments will promote closer interaction between operations and acquisition communities. These M&S environments will be constructed from composeable components interoperating through an integrated architecture. A robust M&S capability enables the DOD to meet operational and support objectives effectively across the diverse activities of the military services, combatant commands and agencies.

The number of available architectures have increased over time. M&S trends indicate that once a community of use develops around an architecture, that architecture is likely to be used regardless of new architectural developments. M&S trends also indicate that few, if any, architectures will be retired as new ones come online. When a new architecture is created to replace one or more of the existing set, the likely outcome is one more architecture will be added to the available set. As the number of mixed-architecture events increase over time, the inter-architecture communication problem increases as well.

M&S has made significant progress in enabling users to link critical resources through distributed architectures.

In the mid 1980s, SIMNET became the first successful implementation of a large-scale, real-time, man-in-the-loop simulator networking for team training and mission rehearsal in military operations. The earliest successes that came through the SIMNET program was the demonstration that geographically dispersed simulation systems could support distributed training by interacting with each other across network connections.

The Aggregate Level Simulation Protocol (ALSP) extended the benefits of distributed simulation to the force-level training community so that different aggregate-level simulations could cooperate to provide theater-level experiences for battle-staff training. The ALSP has supported an evolving “confederation of models” since 1992, consisting of a collection of infrastructure software and protocols for both inter-model communication through a common interface and time advance using a conservative Chandy-Misra based algorithm.

At about the same time, the SIMNET protocol evolved and matured into the Distributed Interactive Simulation (DIS) Standard. DIS allowed an increased number of simulation types to interact in distributed events, but was primarily focused on the platform-level training community. DIS provided an open network protocol standard for linking real-time platform-level wargaming simulations.

In the mid 1990s, the Defense Modeling and Simulation Office (DMSO) sponsored the High Level Architecture (HLA) initiative. Designed to support and supplant both DIS and ALSP, investigation efforts were started to prototype an infrastructure capable of supporting these two disparate applications. The intent was to combine the best features of DIS and ALSP into a single architecture that could also support uses in the analysis and acquisition communities while continuing to support training applications.

The DoD test community started development of alternate architectures based on their perception that HLA yielded unacceptable performance and included reliability limitations. The real-time test range community started development of the Test and Training Enabling Architecture (TENA) to provide low-latency, high-performance service in the hard-real-time application of integrating live assets in the test-range setting. TENA, through its common infrastructure, including the TENA Middleware and other complementary architecture components, such as the TENA Repository, Logical Range Archive, and other TENA utilities and tools, provides the architecture and software implementation and capabilities necessary to quickly and economically enable interoperability among range systems, facilities, and simulations.

Similarly, the U.S. Army started the development of the Common Training Instrumentation Architecture (CTIA) to link a large number of live assets requiring a relatively narrowly bounded set of data for purposes of providing After Action Reviews (AARs) on Army training ranges in the support of large-scale exercises.

Other efforts that make the LVC architecture space more complex include universal interoperability software packages such as OSAMS or CONDOR developed and distributed by commercial vendors.

As of 2010 all of the DoD architectures remain in service with the exception of SIMNET. Of the remaining architectures: CTIA, DIS, HLA, ALSP and TENA, some are in early and growing use (e.g., CTIA, TENA) while others have seen a user-base reduction (e.g., ALSP). Each of the architectures is providing an acceptable level of capability within the areas where they have been adopted. However, DIS, HLA, TENA, and CTIA-based federations are not inherently interoperable with each other. when simulations rely on different architectures, additional steps must be taken to ensure effective communication between all applications. These additional steps, typically involving interposing gateways or bridges between the various architectures, may introduce increased risk, complexity, cost, level of effort, and preparation time. Additional problems extend beyond the implementation of individual simulation events. As a single example, the ability to reuse supporting models, personnel (expertise), and applications across the different protocols is limited. The limited inherent interoperability between the different protocols introduces a significant and unnecessary barrier to the integration of live, virtual, and constructive simulations.

Challenges

The current status of LVC interoperability is fragile and subject to several reoccurring problems that must be resolved (often anew) whenever live, virtual or constructive simulation systems are to be components in a mixed-architecture simulation event. Some of the attendant problems stem from simulation system capability limitations and other system-to-system incompatibilities. Other types of problems arise from the general failure to provide a framework which achieves a more complete semantic-level interoperability between disparate systems. Interoperability, Integration and Composeablity have been identified as the the most technical challenging aspects of a LVC-IA since at least 1996. The Study on the Effectiveness of Modeling and Simulation in the Weapon System Acquisition Process identified cultural and managerial challenges as well. By definition a LVC-IA is a socialtechnical system, a technical system that interacts directly with people. The following table identifies the 1996 challenges associated with the technical, cultural and managerial aspects. In addition, the challenges or gaps found in a 2009 study are also included. The table shows there is little difference between the challenges of 1996 and the challenges of 2009.

Type	1996 Challenges	2009 Challenges
Technical	Interoperability Data Description Availability Data Security and Sensitivity Physics based M&S Hardware and Software Limitations Variable Resolution	Interoperability Data Discovery Security Representative, Composeable and Validated Models Fault Monitoring and Persistence Fidelity, Scale and Resolution filters
Cultural	Acquisition Process Incentives for M&S use M&S workforce (Training and Access) Acceptance of M&S	Process Tools Communities of Practice Workforce Training and Collaboration Infrastructure
Managerial	Office of Secretary Defense Guidance Ownership of Data and Models VV&A Funding Process Use of System Model	Governance, Standards Policies Data & Model Mediation VV&A Consistent Funding Efficient Use and Best Practices

Approaches to a Solution

Ziegler's Architecture for Modeling and Simulation

M&S in the JCID process

A virtual or constructive model usually focuses on the fidelity or accuracy of the element being represented. A live simulation, by definition represents the highest fidelity, since it is reality. But a simulation quickly becomes more difficult when it is created from various live, virtual and constructive elements, or sets of simulations with various network protocols, where each simulation consists of a set of live, virtual and constructive elements. The LVC simulations are socialtechical systems due to the interaction between people and technology in the simulation. The users represent stakeholders from across the acquisition, analysis, testing, training, planning and experimentation communities. M&S occurs across the entire Joint Capabilities Integration Development System (JCID) lifecycle. See the "M&S in the JCID Process" figure. A LVC-IA is also considered an Ultra Large Scale (ULS) system due to the use by a wide variety of stakeholders with conflicting needs and the continuously evolving construction from heterogeneous parts. By definition, people are not just users but elements of a LVC simulation.

During the development of various LVC-IA environments, attempts to understand the foundational elements of integration, composability and interoperability emerged. As of 2010, our understanding of these three elements are still evolving, just as software development continues to evolve. Consider software architecture; as a concept it was first identified in the research work of Edsger Dijkstra in 1968 and David Parnas in the early 1970s. The area of software architecture was only recently adopted in 2007 by ISO as ISO/IEC 42010:2007. Integration is routinely described using the methods of architectural and software patterns. The functional elements of integration can be understood due to universality of integration patterns, e.g. Mediation (intra-communication) and Federation (inter-communication); process, data synchronization and concurrency patterns.

A LVC-IA is dependent on the Interoperability and Composability attributes, not just the technical aspects, but the social or cultural aspects as well. There are sociotechnical challenges, as well as ULS system challenges associated with these features. An example of a cultural aspect is the problem of composition validity. In an ULS the ability to control all interfaces to ensure a valid composition is extremely difficult. The VV&A paradigms are challenged to identfy a level of acceptable validity.

Interoperability

The study of interoperability concerns methodologies to interoperate different systems distributed over a network system. Andreas Tolk introduced the Levels of Conceptual Interoperability Model (LCIM) which identified seven levels of interoperability among participating systems as a method to describe technical interoperability and the complexity of interoperations. Zeigler's Architecture for M&S extends on the three basic levels of interoperability; pragmatic, semantic, and syntactic. The pragmatic level focuses on the receiver’s interpretation of messages in the context of application relative to the sender’s intent. The semantic level concerns definitions and attributes of terms and how they are combined to provide shared meaning to messages. The syntactic level focuses on a structure of messages and adherence to the rules governing that structure. The linguistic interoperability concept supports simultaneous testing environment at multiple levels. The LCIM associate the lower layers with the problems of simulation interoperation while the upper layers relate to the problems of reuse and composition of models. They conclude “simulation systems are based on models and their assumptions and constraints. If two simulation systems are combined, these assumptions and constraints must be aligned accordingly to ensure meaningful results”. This suggests that levels of interoperability that have been identified in the area of M&S can serve as guidelines to discussion of information exchange in general. The Zeigler Architecture provides an architecture description language or conceptual model in which to discuss M&S. The LCIM provides a conceptual model as a means to discuss integration, interoperability and composability. The three linguistic elements relates the the LCIM to the Ziegler conceptual model. Architectural and structural complexity an area of research in systems theory to measure the cohesion and coupling and is based on the metrics commonly used in software development projects. Zeigler, Kim, and Praehofer presents a theory of modeling and simulation which provides a conceptual framework and an associated computational approach to methodological problems in M&S. The framework provides a set of entities and relations among the entities that, in effect, present a ontology of the M&S domain.

Composability

Petty and Weisel formulated the current working definition: "Composability is the capability to select and assemble simulation components in various combinations into simulation systems to satisfy specific user requirements. Again both a technical and user interaction is required indicative of a sociotechnical system is involved. The ability for a user to access data or access models is an important factor when considering composability metrics. If the user does not have visibility into a repository of models, the aggregation of models becomes problematic. In Improving the Composability of Department of Defense Models and Simulation, the factors associated with the ability to provide composability are as follows:

• The complexity of the system being modeled. The size (complexity) of the M&S environment.
• The difficulty of the objective for the context in which the composite M&S will be used. The flexibility of exploration, extensibility.
• The strength of underlying science and technology, including standards.
• Human considerations, such as the quality of management, having a common community of interest, and the skill and knowledge of the work force.

Economic Drivers

To produce the greatest impact from its investments, the DoD needs to manage its M&S programs utilizing an enterprise-type approach. This includes both identifying gaps in M&S capabilities that are common across the enterprise and providing seed moneys to fund projects that have widely-applicable payoffs, and conducting M&S investment across the Department in ways that are systematic and transparent. In particular, “Management processes for models, simulations, and data that … Facilitate the cost effective and efficient development of M&S systems and capabilities….” such as are cited in the vision statement require comprehensive Departmental M&S best-practice investment strategies and processes. M&S investment management requires metrics, both for quantifying the extent of potential investments and for identifying and understanding the full range of benefits resulting from these investments. There is at this time no consistent guidance for such practice.

LVC Continuum

The development & use costs associated with LVC can be summarized as follows:

• Live - Relatively high cost since it is very human resource/materiel intensive and not particularly repeatable.
• Virtual - Relatively medium cost since it is less human resource/materiel intensive, some reuse can occur, and repeatability is moderate.
• Constructive - Relatively low cost since it is the least human resource/materiel intensive, reuse is high, and repeatability is high.

In contrast, the fidelity of M&S is highest in Live, lower in Virtual, and lowest in Constructive. As such, DoD policy is a mixed use of LVC through the Military Acquisition life cycle, also known as the LVC Continuum. In the LVC Continuum figure to the right, the JCIDS process is related to the relative use of LVC through the Military Acquisition life cycle.

See also

• Simulation Based Acquisition
• Simulation & Modeling for Acquisition, Requirements, and Training