Modeler's discrete event simulation kernel, the core of the simulation environment, has been continuously enhanced for over 20 years to deliver faster simulations that scale efficiently with traffic growth.

Aside from raw performance, Modeler incorporates numerous features to accelerate simulations, including:

• A 64-bit kernel
• General parallel simulation
• Grid computing support

Modeler’s inherent parallel discrete event simulation kernel leverages multi-core processors or multi-processor machines to accelerate simulation run-time. Users can select which process models are executed in parallel to ensure optimal resource utilization. In grid computing environments, Modeler can distribute a series of simulations to multiple machines for simultaneous execution. This feature is especially useful for parametric studies where one or more model parameter values are varied to analyze their effect on overall network behavior. An important use of parametric studies is to validate simulation results. This is done by varying the random number seed across a family of simulations to ensure statistical validity. All distributed simulations are centrally controlled from the DES Execution Manager within Modeler. Important system information such as CPU utilization, memory usage, and event execution are managed from this convenient console.

In addition to discrete event simulation, Modeler offers two other simulation technologies that require less detail than that offered through discrete event simulation. The first, Flow Analysis™, offers full analytical modeling. It provides the fastest execution time relative to other approaches and scales to support large networks and traffic volumes. Flow Analysis is useful for iterative simulations used in network design and failure analysis. The second simulation technology, Hybrid Simulation, offers a mix of discrete and analytical traffic modeling for highly detailed and rapid simulations.

OPNET simulation technology detail vs. speed

Model debugging is performed directly within Modeler using an innovative graphical debugger. Users can visualize packets flowing through the model to quickly pinpoint errors and use the event-level debugger to inspect and control the simulation. Frequently-needed debugging information, such as event lists and packet content data is readily available and updated in real-time. Users can also leverage integrated source-level debugging to analyze C/C++ process model code. The built-in code and memory profiler allows users to further optimize model performance.

Modeler’s advanced Checkpoint/Restart feature simplifies simulation excursions and debugging. By saving or “checkpointing” the simulation state periodically during lengthy initializations, users can execute multiple excursions from a checkpoint rather than restarting the simulation for each excursion run. Similarly, checkpoints eliminate the need to restart a simulation in the event of an abrupt termination.

OPNET’s HLA capability supports building and running a federation of many simulators (federates), each modeling some aspect of a composite system exchanging messages and maintaining appropriate time synchronization. The OPNET HLA module has been validated for use with HLA RTIs 1.3 and IEEE 1516.

Traditional wargaming simulators may account correctly for battlefield strategy, but generally make simple assumptions about the availability of communications. Arguably, the availability and robustness of communications can significantly affect the outcome of a battle.

By federating Force-on-force (FoF) simulators with the Modeler Wireless Suite for Defense, wargaming simulations incorporate mission-critical network effects such as protocol performance, terrain masking, mobility, and jamming. This ensures that battlefield conditions and the reality of modern-day communications are accurately represented in the wargame.