1. Page 1 of 8
Nogesi case study: LSA proof of concept
José-Ramón Martínez-Salio Jose-Maria Lopez-Rodriguez
NADS NADS
Madrid, Spain Madrid, Spain
jrmartinez@nads.es jmlopez@nads.es
INTRODUCTION
NADS key product, SimWare has reached a new milestone with the delivery of the NOGESI project. NOGESI (NOdo GEnerico de SImulación) is a project for the Spanish MoD where SimWare is the key element for the creation of a real time simulation laboratory. NOGESI puts together simulators and COTS in HLA and DIS approaching and probing the LSA concept. NOGESI project has meant the testing workbench and evolution of SimWare Framework and SimWare RTI towards a truly proved Open Layered simulation architecture. As a proof of concept, a new product called SimWare Lab has evolved. Next steps are also discussed; SimWare Lab in WAN and the project of interconnection between Spanish Lab and COE M&S lab in Rome.
ABOUT THE AUTHORS
Jose-Ramon Martinez-Salio has an MSc in Industrial Engineering (specialized in robotics and electronics) by the
Polytechnic University of Madrid, Spain. He is Technical Presales engineer director of NADS. Until 2005 he was focused in the development area of SimWare product and was responsible of management different projects in the simulation area. He has 15 years of experience in Simulation projects, DDS and high tech projects, as main engineer and project manager.
Jose-Maria Lopez-Rodriguez has a MSc in Industrial Engineering (specialized in robotics and electronics) by the
Polytechnic University of Madrid, Spain. He is Vice-president for Business Development and cofounder of NADS
(Nextel Aerospace, Defence & Security). Until 2006 he was managing the development of NADS’s R&D projects, mainly related to the development of simulation middleware and framework compliant with HLA and DDS. He has about 15 years of experience in Simulation, C2 and Virtual Reality projects, as principal engineer, project manager and business development manager. He has more than 10 papers in international congresses, mainly related to Simulation and Virtual Reality issues.

2. Page 2 of 8
THE QUEST FOR A SIMULATION PLATFORM FOR INTEROPERABILITY
Since late 90s, there have been many projects trying to offer a seamless solution to the interoperability between different simulators on a network. Simulation community have developed many standards in order to try to provide better interoperable solutions but interoperability is still a great pain for lead users of distributed simulation like US DoD and other NATO members. These pains are very well documented in reports like LVC Architecture Roadmap (LVCAR) (Henninger et al. (2008)), sponsored by US M&S Coordination Office.
In the last decade the main simulation procurement agencies have invested hundreds of millions of US dollars looking for a solution. Projects like MATREX or LVC-IA in USA, ITCS in France and others, have advanced a lot in the development of processes and tools to manage repositories and to guide the development of federations. All those examples are not fulfilling the main goal of these projects: improve reusability and runtime interoperability. E.g. runtime interoperability has been tried to be achieved by developing complex translation and gateways solutions; but performance, capabilities, scalability and maintenance of these kinds of solutions are still very poor.
To this day, agencies with large budgets, like Us DoD, have been able to achieve some successful results, investing much more money and resources than expected and suffering a lot of delays and finally having to accept down-graded solutions. In other NATO countries; agencies with bigger budgets like UK, France, Germany or Sweden have archived partial success. On the other hand, countries with smaller budgets, like Spain, have advanced very little in distributed simulation, since the required investment is high and the ROI is very uncertain.
Because of that, Spanish MoD has been requesting some strategic requirements for Simulation in its Technology and Innovation Strategy for Defense (ETID) (Plan Estatal de Investigación Científica y Técnica y de Innovación 2013-2016 (2013)).
In order to fulfill these requirements, the Spanish MoD has been working with SimWare platform since late 2010. In this past two years, the Spanish MoD and NADS have been working together in two related projects in order to evolve SimWare simulation framework as a platform for open and interoperable distributed simulation.
As a result, Spanish MoD has now a Distributed simulation lab, called NOGESI, whose main capabilities are:
 Interoperability between simulated and operational systems using DDSI open wire protocol
 Development and maintenance of Simulation assets using MDA (Model Driven Architecture) paradigm.
 Deployment of simulation assets as services into an HLA federation.
 HLA interoperability to a wire protocol level.
Figure 1. NOGESI architecture
NOGESI deployment demonstrates that using an open platform for interoperability allows also an easy integration of legacy simulators. In this case, a Forward Air Controller (FAC) virtual simulator, developed by Indra for the Spanish Air Force in 2010, have been integrated into NOGESI supporting some

3. Page 3 of 8
advanced simulation services like HLA´s ownership transfer. By using HLA Ownership services over DDSI wire protocol, FAC virtual simulator can interoperate in real-time with a simulation server in uses cases like Close air Support or Call for Fire. In this exercises, friendly forces simulated by VRForces or Stage can call for air support to a FAC that it is going to engage foes simulated by VBS2 with missiles or bombs that are being simulated by another HLA federate; the weapon server.
WHAT IS DIFFERENT IN NOGESI
In a system of systems simulation project, like MATREX or ITCS, the traditional approach has been to try to implement a simulation engineering methodology, like FEDEP or its evolution DSEEP, with a set of software tools and processes. In this approach new developments are focused in the system engineering not in the deployment of the simulation. For the deployment, these projects propose some kind of wrapper or middleware that provides an abstraction of different simulation protocols like HLA, DIS or TENA. Translation code is generated to do gateways and bridges between the different types of simulators. There are not innovations in the deployment of the simulations, using instead existing runtime infrastructures to run the exercises. The simulation services available for the federation are the basic services included in HLA or DIS, like entity and interaction management. Advanced simulation services like execution, ownership or time management are unavailable, even when some protocols like HLA support them.
In NOGESI, main innovations are not in the development process but in the simulation technical architecture. NOGESI is based on a pure distributed data-centric architecture made possible with the use of DDS. The use of DDS with its Quality of Services (QoS) allows avoiding the main pains in HLA, like vendor’s incompatibility and the difficulty of achieving scalability. Proposed architecture combines the use of traditional simulation standards like HLA and DIS with the use of DDS as the common communication standards.
ARCHITECTURE OF THE SOLUTION
NOGESI is a very good example of how SISO LSA architecture could be performed in reality. NOGESI architecture is based on LSA architectural concept, showed in the next figure.
Figure 2. NOGESI architecture as LSA proof of concept
The first and pivotal point of both SimWare architecture and LSA proposal is the use of DDS in conjunction with HLA and DIS standards. Since DIS is only oriented to pure communication and HLA evolves the concepts of DIS, the discussion can be centered in the possible differences and synergies between HLA and DDS.
Looking at the technologies from a pure technical comparison of features; (Lopez-Rodriguez et al. (2012) A Comparison of Simulation and Operational Architectures; Joshi et al. (2006). A Comparison and Mapping of Data Distribution Service and High-Level Architecture) we can draw next figure:

4. Page 4 of 8
Figure 3. Comparison between DDS and HLA
We can see that, in features, HLA is different to DDS mainly on the services related with simulation. This could be easily foretold due to the orientation and origin of both standards.
Keeping the comparison in terms of features and not in term of possible uses, two different approaches can be established starting from this comparison:
1. Substitution of HLA with DDS for the communication of the simulator: this is the approach used in SimWare Framework architecture; use DDS as the main base for all the communications related with the simulation. The sole use of DDS presents a number of advantages (Lopez-Rodriguez et al. (2011) Use of the OMG DDS standard in Simulation) In this approach, the services that HLA have and DDS doesn´t have (e.g. time management) can be easily implemented, if needed, over the DDS communication layer.
2. Complementation of HLA with DDS as the HLA RTI: this is the approach of SimWare RTI, building HLA standard over DDS communication layer. HLA standard allow the use of something like DDS standard as the communication layer
(Lopez-Rodriguez et al. (2011). How to Develop True Distributed Real Time Simulations? Mixing IEEE HLA and OMG DDS standards).
NOGESI and SimWare Lab (its offspring) present and take advantage of both approaches.
Second in importance is the communication with already existing HLA and DIS elements (including COTS). Since NOGESI has a pure DDS core, this communication is made with the use of data gateways.
These gateways are built with a multilayered approach that avoids traditional problems like reusability, maintenance and overhead. This approach is based on the pre-existence of two libraries of code:
 NcWare SIM: is the common library to connect with HLA and/or DDS. It makes transparent to the user the connection details of each vendor. A common layer allows connecting to DDS or HLA with the same code and capabilities.
 DataModel in DDS: This is the common data model of all the DDS-based simulation. The data model in DDS is not limited to the “objects” but includes all the publishers and subscribers (including callbacks) to use them in a DDS communication.
With these two elements in the base, the creation and maintenance of a data gateway is then limited to select and connect the chosen fields of the model leaving the communication complexities to the preexisting libraries.

5. Page 5 of 8
Figure 4. Gateway structure
Another important point that NOGESI demonstrates is the capability of reuse simulation models between different simulations. This is closely related with the capability of fast and easy change of simulation scenarios. The simulation models became services that can be deployed “as is” in any simulation scenario. Most important, these models are independent. So, any model can be used to run against already existing models.
This architecture has all the potential to grow (e.g. by adding new models) and to be adapted, if needed, to new COTS or tools. NOGESI comes integrated with a Weapon Server, which is deploying a set of simulation services that can behave as another HLA federate or as DDS simulation elements. The sole dependency that these models have is the data model.
LOOKING INSIDE THE ARCHITECTURE
Layered architecture and its approach with SimWare is better understood if you look into the detail of the weapon server architecture. Dynamical simulation of the weapon services is managed by SimWare runtime infrastructure, which is a fully compliant DDS distributed infrastructure and is composed by four main components
 Scheduler. Is a real time scheduler which manages the execution of the simulation and the state machine during runtime. The entire configuration is based on a XML configuration file.
 SimEngine. Manages cyclical execution of the simulation assets during runtime. In a fully distributed architecture, like SimWare, a SimEngine instance is required in each node that it is running any number of simulation assets.
Figure 5. Dynamic simulation architecture
 Simulation Assets. Dynamical models of simulated entities. In NOGESI deployment there are simulation models for infrared missile, sensor, aircrafts and LGB
 ACS. Tool to manage weapon server and its instances. Allows interacting with instances of the simulation services during runtime.
One important feature of SimWare runtime infrastructure is that every data used to control the

6. Page 6 of 8
simulation is managed using a DDS topic, using a DDS compliant control datamodel to manage the simulation.
Figure 6. NOGESI and HLA federates
Having an open datamodel to manage simulation assets is currently a topic of discussion in LSA forums, because it has been considered essential to improve the interoperability to an asset level. One natural evolution of this discussion has been an approach to ANDEM study group.
Even though the main innovation in NOGESI is its technical architecture, a quite innovative approach has been also made in order to improve interoperability and reusability of simulation assets (Jimenez et al. (2012). The Spanish MoD M&S Infrastructure: a new perspective for the development and reusability of simulations assets). NOGESI, therefore, is using a fully open layered architecture as LSA proposes
PROMOTING LSA SOLUTIONS WITH NOGESI
SimWare Framework and SimWare RTI have been successfully tested with the NOGESI project. The result is incorporated to the latest releases of SimWare Framework and SimWare RTI.
As a proof of technology, the result and sum of it is a new product called SimWare Lab.
SimWare Lab appears as a demonstrator of technology of NADS. SimWare Lab has been built with multiples different objectives in mind:
 Composability of models and simulators: we want to join together different models and simulators compatibles with HLA1516 without changing them.
 Integration area: we want to integrate new technologies and old legacy models and simulators.
 Battle lab: SimWare Lab is a fully operational battle lab with simulation models and scenarios.
 Demonstrator of technology: SimWare Lab is a demonstrator of NADS products and technologies
 Interoperability hub: SimWare Lab mixes DIS, HLA and DDS in the same simulation.
 Test bed: SimWare Lab is the perfect simulation test bench.
SIMWARE LAB AS SIMULATION BLACK BOX
SimWare Lab can be considered a “simulation in a box” COTS. The Lab can be used to connect with any HLA/DIS compatible element using different HLA vendors without changing any pre-existing element. The Lab can add new simulation elements to any scenario (fully interactive) or just communicate data between different HLA vendors or implementations. Fully interaction between HLA federations is possible with the transition of ownership transfer (in all the possible ways) through the data gateways.

7. Page 7 of 8
NOGESI AS PROOF OF CONCEPT OF LSA
NOGESI has been also the proof of concept and test bed of the LSA architecture. The final architecture of the project demonstrates some of the main elements of LSA proposal and also opens some interesting questions to discuss in a possible future standard.
 It is feasible to join DDS and HLA/DIS using data gateways without incurring in high latency costs.
 Data model openness is an open question: using the “fixed” datamodel system of HLA has some advantages (e.g. the lab can be a real black box that you can always use in any scenario) but losses some of the flexibility of an open datamodel.
 Data model of control is possible in DDS. It is needed to define it carefully (like a meta-FOM) going beyond the pure control model (e.g. including the ownership transfer).
 DDS has proven that can be the HLA RTI.
 DDS qualities of service are too complex for the simulation approach. Maybe it could be useful to have predefined QoS patterns for different simulation scenarios (e.g. for WAN).
NEXT CHALLENGES FOR NOGESI
Next challenge for NOGESI and SimWare Lab is the interconnection of regional laboratories in a pan- European network. A three staged plan has been established to achieve this result:
1. NOGESI project installed and deployed in the Spanish Mod and in NADS laboratories. This has been reached in June of 2013
2. SimWare Lab installed in NATO COE of Rome. Made on July 2013.
3. Interconnection using WAN of both facilities. To be done during summer-fall 2013.
This is now a work in progress that has some interesting challenges pending:
 Wan interconnection using DDS. The QoS file has to be customized to establish and keep the connection. It is necessary to establish a net service for changing dynamically the QoS?
 Compensation of network latencies between simulation participants. Is that really necessary? In witch extension? Can DDS without help compensate the latencies involved?
 Network requirements and limitations. Is really necessary to have a dedicated network? Here we are, of course, not considering security issues yet.
 Security requirements. Can they be covered by DDS security standard? Is anything else needed?
SUMMARY
NOGESI project with the Spanish MoD have demonstrated that one of the proposals made at LVCAR for improve interoperability in LVC simulations, converging existing architectures with a new one, it is possible to be achieved with a low budget and low technical risk by the simulation community by following an open innovation model like SimWare. Taking advantage of new data exchange technologies, like DDS, successfully applied in many others domains, a new architecture, structured by layers and based on open interfaces, can be a solution to the many pains that exist right now in the simulation industry.

8. Page 8 of 8
REFERENCES
Henninger, A., et al. (2008). LVCAR Final Report.
Institute for Defence Analyses.
Jimenez, P., & Lopez-Rodriguez, J.-M. (2012). The Spanish MoD M&S Infrastructure: a new perspective for the development and reusability of simulations assets. Simulation Interoperability Workshop Fall 2012.
Joshi, R., & Castellote, G.-P. (2006). A Comparison and Mapping of Data Distribution Service and High- Level Architecture. Retrieved from RTI: http://www.google.com/url?sa=t&rct=j&q=%22A+Comparison+and+Mapping+of+Data+Distribution+Service+and+HighLevel+Architecture%22&source=web&cd=1&ved=0CE4QFjAA&url=http%3A%2F%2Fwww.rti. com%2Fwhitepapers%2FComparison_and_Mapping_of_DDS_and_HLA.pdf&ei=pMbhT_bA
Live, Virtual, Constructive-Integrating Architecture (LVC-IA) http://www.peostri.army.mil/PRODUCTS/LVCIA/
López-Rodriguez, J. M., Seisdedos, I., & del Barrio, P. (2008). Using Open software architectures for developing simulation systems: a Way to the interoperability of simulations at subsystem level. 12TH C4ISR SEMINAR.
Lopez-Rodriguez, J.-M. (2011). Use of the OMG DDS standard in Simulation. A new Way for developing Real Time Distributed Simulations. OMG's Workshop on Real-time, Embedded and Enterprise-Scale Time-
Critical Systems.
Lopez-Rodriguez, J.-M., Gregory, D., & Martinez- Salio, J.-R. (2012). A Comparison of Simulation and Operational Architectures. Simulation Interoperability Workshop Fall 2012.
Lopez-Rodriguez, J.-M., Gregory, D., & Martinez- Salio, J.-R., Corsaro A. (2013). A New Approach for Converging LVC Simulation Architectures. Simulation Interoperability Workshop Spring 2013.
Lopez-Rodriguez, J.-M., Martin, R., & Jimenez, P. (2011). How to Develop True Distributed Real Time Simulations? Mixing IEEE HLA and OMG DDS standards. Simulation Interoperability Workshop Spring Conference 2011.
Modeling Architecture for Technology, Research, and EXperimentation (MATREX), https://www.matrex.rdecom.army.mil/
Naval Open Architecture and DoD Open Systems Architecture. (2012). Retrieved from https://acc.dau.mil/oa
OMG. (2009, January). DDS Interoperability Protocol (DDS-RTPS) v2.1. Retrieved from http://www.omg.org/spec/DDS-RTPS/2.1/
P1730.1 -Recommended Practice for Distributed Simulation Engineering and Execution Process (DSEEP) Multi-Architecture Overlay (DMAO). (n.d.). Retrieved from IEEE: http://standards.ieee.org/develop/project/1730.1.html
Plan Estatal de Investigación Científica y Técnica y de Innovación 2013-2016. Retrieved from http://www.idi.mineco.gob.es/portal/site/MICINN/ menuitem.7eeac5cd345b4f34f09dfd1001432ea0/?vgnextoid=83b192b9036c2210VgnVCM1000001d04140aRCRD
Saunders, R. (2010). Live Virtual Constructive Architecture Roadmap (LVCAR) Convergence approach. Proceedings of Interservice/Industry Training, Simulation, and Education Conference (I/
ITSEC) '10. Florida.