4.2. Interoperation Structures of Major MRM Experiments
The selected 22 studies were scrutinized in this review. These studies built various structures to conduct experiments for the MRM and LVC systems. Except for two studies that are basic research for the other studies, a total of 20 experiments were included. In these experiments, the actual simulation platforms were used and interconnected to explore issues related to LVC integration and resolution difference. The specific structures and distinctions of each experiment are summarized in this chapter, and the issues commonly found are discussed in the next chapter.
] conducted the first simulation integration, which was the integrated Eagle/BDS-D project. The Eagle system was a company- and battalion-level simulation developed by the US Army Training and Analysis Command (TRAC). The goal of the Eagle/BDS-D project was the integration of the Eagle and SIMNET simulations. The Institute for Simulation and Training (IST) has created a Computer Generated Forces (CGF) Testbed that generates and controls individual entities in the SIMNET world. The Los Alamos National Lab (LANL) developed the Simulation Integration Unit (SIU). The SIU takes information about individual entities from the SIMNET Protocol Data Units (PDUs), which is processed and integrated by the SIU before transmission to the Eagle. Figure 5
illustrates how those systems are interconnected.
] created the Eagle II prototype that is a distributed combat model composed of the Eagle, the SIMNET/Semi Automated Forces (SAF), the SIU, and a visualization program (NPSNET) as described in Figure 6
. Since the Eagle Ⅱ prototype could not use real the SIMNET simulator hardware, a method for providing three-dimensional visualizations in high-resolution areas that are typically available in these simulators was required. They used the term “MRM” for the first time, emphasizing its importance. This study is meaningful in that it used the SAF, not the CGF.
] conducted the Brigade/Battalion Battle Simulation (BBS)/DIS project to explore the issues surrounding the integration of constructive and virtual simulators. As shown in Figure 7
, the BBS was chosen as a constructive system. Since a DIS compatible simulation was not available at the time, the SIMNET-compatible Semi-Automated Forces (SAF) was selected for a virtual system. The BBS Advanced Interface Unit (AIU) was built to support communication with the BBS and the SIMNET/SAF system. The BBS AIU consists of three main components: the AIU, the SAF engine, and the Simulation Control (SIMCON). The AIU delivers interfaces between the time-stepped and real-time simulations. The SAF engine performs the modeling of disaggregated entities. On the other hand, SIMCON is a software that offers an interface into the BBS common area. The experiment related to aggregation and disaggregation was not included in this study.
] implemented the interconnection of the Virtual SOF Inter-Simulator Network (SOFNET) aircraft simulator and the Joint Conflict Model (JCM) theater-level constructive simulation. The interoperability is accomplished using the DIS PDU. Figure 8
shows the structure connecting the two systems that exist in different locations. They used the Network Interface Unit (NIU) to regulate the data intercommunicated. Terrain correlation issues due to the differences in the resolution were identified, and they attempted to solve the problems by creating a terrain-following algorithm. This interface helped with mission reviews and mission rehearsal training.
] implemented a dynamic aggregation and disaggregation process in the Joint Precision Strike Demonstration (JPSD) program. The purpose of this program was to conduct technical research that could contribute to the defense arena. The Corps Level Computer Generated Forces (CLCGF) system was created by integrating the Eagle constructive simulation with the Modular Semi-Automated Forces (ModSAF) entity-level simulation. The SIU manages the resolution changes. The ability of the CLCGF to make dynamic resolution changes efficiently is enabled by three key design decisions: to tightly couple the SIU to ModSAF (Figure 9
), to represent all of Eagle’s units in ModSAF as aggregates, and to require that Eagle units be simulated at the company (or battery) level. They addressed the issue that dynamic aggregation and disaggregation is one of the keys to supporting large-scale DIS exercises.
] conducted the aggregation and disaggregation research by developing a linkage between the UK Army’s Advanced Battlefield Computer Simulation (ABACUS) and the ModSAF. They investigated the benefits and problems of this connection, and techniques for achieving interoperability between different levels of the resolution systems. A software module known as the Aggregation Disaggregation Unit (ADU) was used to translate the data for the appropriate simulation. As shown in Figure 10
, the ADU consists of four logical elements: the simulation management interface (SMI), the pseudo-disaggregation process (PDP), the aggregate update process (AUP), and the full disaggregation process (FDP). The SMI analyzes the ABACUS messages to gain information on simulated units and requests updated messages at an appropriate time before ABACUS starts its time step update. The PDP and AUP were coded as one for efficiency. The PDP retains a list of entities that are being simulated and sends this information to shared memory for access by the FDP and AUP. The FDP initializes all required ModSAF libraries and creates a local copy of the first database created by the ModSAF.
] presented an architecture for connecting multiple aggregate-level wargame simulations to multiple virtual components. They linked the Eagle aggregate simulation with the ModSAF and ITEMS using a DIS network, as shown in Figure 11
. ITEMS is a UNIX process that can run on Silicon Graphics workstations. The Eagle Constructive Simulation Interface Unit (CSIU) provides the functionality of the Aggregate Simulation Interface (ASI). The SIU communicated with Eagle via UNIX RPCs, but the CSIU can be configured to use TCP/IP or UDP network packets. The Virtual Simulation Interface Unit (VSIU) provides the functionality of the Virtual Simulation Interface (VSI) for the ModSAF and ITEMS, and communicates with ModSAF using the persistent object (PO) protocol. This attempt is viewed as a new and powerful tool for conducting analytical research, but also a way to bring legacy simulations into the virtual world.
] performed the DARPA’s Dynamic Multi-user Information Fusion (DMIF) project related to the analysis of sensor data collected in large-scale combat. To achieve the goal of this project, the STOW SAF sensor models are integrated with Eagle and ModSAF, as shown in Figure 12
. The STOW SAF is made up of the ArmySAF, AirSAF and MarineSAF. The pseudo-disaggregation was implemented within the Eagle/ModSAF linkage system that was developed initially for the CLCGF project. Through the experiments, they found that the ModSAF suite had an optimum capacity of between 2500 and 3500 pseudo-entities. They also suggested the following further pseudo-disaggregation ideas: enhancing pseudo-disaggregation realism and increasing system capacity.
] describes the Swedish Defense Research Establishment project that is working on the problem of applying the HLA concept to interoperating simulation models on different levels of resolution. Three legacy models were used to build a federation: ARTEVA, FBSIM and TYR. The ARTEVA is a stochastic model for technical-level battle. The FBSIM is a unit-level simulation that models a higher aggregation level unit than the ARTEVA. TYR is a simulation engine to support the command and decision-making training of senior officers and staff. The aggregation and disaggregation strategies are illustrated in Figure 13
. They pointed out the data inconsistency and fidelity problems that occur when using models on different resolutions.
] implemented a study to build the battlespace federation using HLA and the federation is described in Figure 14
. The aggregate-level simulation, Advanced Regional Exploratory System (ARES), was used as a campaign-level model to interoperate with the ModSAF. The Virtual Command Center (VCC) provides a point-and-click interface to interact with the unit as well as showing the commander’s view of the combat. Some issues related to the aggregation and disaggregation were reported. The author pointed out that the unit model needs to maintain more information about its entity status for data consistency.
] built a federation to link the Joint Theater Level Simulation (JTLS) with the Joint Conflict and Tactical Simulation (JCATS) using the HLA. The JTLS is a unit-level simulation used for supporting the analysis of operational plans. The JCATS is an entity-level simulation developed for training. JTLS-JCATS federation architecture is described in Figure 15
. They used a shared object ownership method to avoid repeated changes to the number of units passed from JTLS to JCATS. They also emphasized the importance of using early FEDEP work.
] presented the Army Constructive Training Federation–Multi-Resolution Modeling (ACTF-MRM) architecture. The Program Executive Office (PEO) creates the ACTF-MRM for simulation, training and instrumentation (STRI). As shown in Figure 16
, the federation consists of the Corps Battles Simulation (CBS), the Combat Service Support Training Simulation System (CSSTSS), and the JCATS. These were connected using four interfaces: HLA, DIS, CBS Master Interface (MI), and point-to-point. Both TACSIM and the runtime manager (RTM) receive data for the aggregate units through the MI. They create an algorithm referred to as FLUD, which is a mechanism that converts CBS aggregates into representations of platforms and squads. This ACTF-MRM can solve the issues concerning the inconsistent representation of individual vehicles and squads in the Command, Control, Communication, Computers, Intelligence, Surveillance, and Reconnaissance (C4ISR) systems.
] performed an experiment that develops an HLA federation using the Joint Warfare System (JWARS) and the Joint Semi-Automated Forces (JSAF), as depicted in Figure 17
. The JWARS was used as the campaign-level low-resolution simulation, and the JSAF was utilized as the mission-level high-resolution simulation. The operating states were divided into three: fully aggregated (FA), disaggregated (DA) and pause state. The pause state is used for ensuring simulations stay synchronized at a reasonable level. The adaptability of the JWARS and JSAF models to handle other federation object models (FOMs) can gradually enhance the federation by adding a variety of systems.
] performed the NATO training federation (NTF) project that developed an HLA federation using the JTLS and JCATS, as shown in Figure 18
. The NTF was successfully used in a major NATO exercise for the first time in 2008. The Virtual Battle Simulation (VBS2) provides intelligence, such as aerial photography and video streams of unmanned aerial vehicles (UAVs), to help the training audience gain better insight into decision making.
] implemented a dynamic MRM using an HLA protocol. They used the COTS simulations to build a federation. The MASA Sword, a simulation platform for military training from the platoon- to battalion-level, was used as the aggregated simulation, and VR-Forces, a CGF platform for entity and unit model simulation, was utilized as the entity-level simulation. Figure 19
shows the aggregate unit and disaggregated entities at the same time. Both are COTS simulations. This study found that the latest simulated COTS products built using existing interoperability standards can communicate efficiently and realistically.
] presented an experimental LVC simulation framework. The SIMbox from SIMIGON was employed as a virtual flight and an anti-air missile simulator. VR-Forces was used as a constructive simulation. The SIMbox is a high-resolution simulator for the training of F-16 pilots. The AddSIM, which is a component-based simulation developed by the Agency of Defense Development in South Korea, participated in the simulation framework as a federate, as illustrated in Figure 20
. Furthermore, they utilized a tablet PC as the live component for the experimental LVC simulation framework. This work helps one understand the design concepts of the LVC simulation system.
] conducted the MRM experiment using COTS simulations. The Battle Command was used for executing a unit-level simulation, and VR-Forces was utilized as an entity-level simulation. Primarily, they emphasized the importance of geographic data consistency when building the MRM federation. All geospatial data and map feature data are coordinated through the terrain generation tool to ensure that geographic information for entities is always suitable during the simulation, as shown in Figure 21
] developed the integrated system for LVC and conducted Verification and Validation (V&V) study of the aircraft weapon system. They linked the Flexible Analysis, Modeling and Exercise System (FLAMES), the Extended Air Defense Simulation (EADSIM), the Reconfigurable Flight Simulator (RFS) and the Aircraft Combat Maneuvering Instrumentation (ACMI), as described in Figure 22
. The FLAMES and EADSIM are constructive simulations, the FRS is a virtual simulator, and the ACMI is the live training system. They utilized an unmanned model aircraft as the ACMI substitute. Three different scenarios and V&V test were conducted. The “MRM” was not mentioned in this study because differences in the resolution were not compared. This study is meaningful in that the LVC integration is successfully implemented.
] built an MRM federation using the VR-Forces and SIMbox using an HLA architecture. VR-Forces was used as the LRM, and the SIMbox was used as the HRM. The federation configuration is shown in Figure 23
. Two experiments were performed to analyze the phenomenon associated with the resolution difference. The engagement scenario was developed to investigate the interaction between objects created in each system. They found that the database gap could limit the representation of entities, and that all federation systems should have a common understanding to communicate with each other.
] performed the Virtual-Constructive (VC) system interlock test using the Air Force fighter simulators and War Game model. The F-15K, KF-16 and FA-50 simulators were connected to a unit-level Changgong model, as depicted in Figure 24
. This study is valuable in that the platforms currently being used were connected to communicate with each other, and the feasibility of the Air Force LVC training system was confirmed through this. The term “MRM” was not mentioned in this research, but some limitations that have been identified, such as object visualization issues, are strongly related to the MRM issues. Furthermore, to make a larger battlefield, resolution conversion is necessary, and the need for the MRM must be emphasized accordingly.