Technology Today

2013 Issue 1

Raytheon Air and Missile Simulation: A Selectable Fidelity Tool for Radar and Systems Analysis

Raytheon Air and Missile Simulation:
A Selectable Fidelity Tool for Radar and Systems Analysis

Designing systems and confirming that they can defend against coordinated raids of ballistic missiles, aircraft and cruise missiles is challenging. How, for example, can enough fidelity be provided to test components and algorithms? Likewise, how can complex interactions between system elements be examined before the system design is complete? Raytheon Air and Missile Simulation (RAMS) addresses these challenges by using collections of threat modeling, radar processing, battle management and interceptor modules with a plug-and-play foundation. RAMS can represent the system elements at low-, medium- or high-fidelity levels, and it supports a range of applications from standalone radar analyses to mission-level performance assessments of the integrated operation of multiple radars with battle management and interceptors.

RAMS has been used for detailed system performance analysis in multimission environments, for radar feature analysis, and for characterizing the performance of advanced discrimination algorithms for new or evolved ballistic missile defense (BMD) radars. Each application that uses RAMS gains efficiencies by leveraging improvements that have been made by prior applications. RAMS supports the incorporation of the new efficiencies by using a common code base for all applications and allowing alternative representations of functionality to be selected using configuration files that are processed at program initialization.


RAMS is a modular, plug-and-play framework and set of routines that support development and evaluation of approaches for BMD and air defense, and that can readily be extended to support other missions such as surface or electronic warfare. RAMS can simulate detailed radar data to thoroughly test signal processing and data processing algorithms, or it can rapidly generate lower-fidelity data to support system-level end-to-end performance assessments. In this framework, modules coordinate using well defined interfaces, allowing alternative realizations, i.e., different combinations of modules, to be instantiated at run-time.

Figure 1. Plug-and-play framework allows evolution of fidelity and capabilities from functional models to detailed implementations as the program progresses.Fidelity levels are selectable and range from functional models that represent typical or idealized performance to detailed models that capture every step and dependency of a process. Figure 1 illustrates three fidelity levels for representing scene and radar components, each of which is appropriate for particular analysis questions. During the concept definition stage, operational analysis using simple models (left column) supports system tradeoffs and guides top-level capability definitions. Incrementally, the scene fidelity, signal processor, tracker and discrimination models are replaced with higher-fidelity ones to enable assessments of candidate approaches, prototype designs and even real-time implementations.Because common interfaces are used for these functions, different combinations of signal processing, tracking and planning can be instantiated together to match the analysis goals and design maturity.

Most of RAMS is implemented in Java™, which readily supports modular interfaces and provides mechanisms to quickly identify code and algorithm defects. Some detailed signal processing routines have been recoded into C to provide more speed, and a few modules have been implemented in Compute Unified Device Architecture (CUDA®) to utilize graphical processing units (GPUs) for additional speed. For cases in which existing algorithms are coded in C, C++ or Fortran, the code can be reused as is, by applying a Java wrapper around the legacy code so that the routine can properly interface with the RAMS simulation infrastructure. Multithreading, where multiple modules or portions of modules are executed simultaneously on a multiprocessor computing system, is also used to increase execution speed.

Figure 2. An analyst at his desktop uses RAMS to understand the different waveform resolutions that are produced for the same object. High resolution shows details in range, medium resolution provides gross size information, and low resolution indicates the presence of an object. See Figure 3 for a close-up of the screen.

For detailed BMD analyses, RAMS represents ballistic missile complexes in high fidelity, with operating frequency and angle-dependent, range-resolved signatures of principal objects such as re-entry vehicles (RVs), stages, decoys and debris. All key phenomena that may be exploited by BMD radars have been represented to ensure that algorithms designed for signal processing, tracking and discrimination can be thoroughly tested. Figure 2 shows an example of the range dependence of an object’s RCS (radar cross section) as it is observed with low-, medium- and high-resolution waveforms, each of which is used to perform distinct radar functions. Simulating multiple simultaneous ballistic missile complexes provides stressing conditions to assess and refine radar planning and pulse scheduling algorithms.

To evaluate future BMD radar capabilities, RAMS was configured to represent a potential future upgrade of an existing radar containing advanced discrimination algorithms. Although a specialized high-fidelity simulation of this radar existed, the flexible RAMS architecture made it easier to assess potential system improvements since RAMS only required development of new configuration files vs. having to modify specialized simulation routines.

Aircraft and cruise missiles are currently modeled using frequency and angle-dependent signature models. Environmental effects such as attenuation from the atmosphere and rain are modeled to ensure correct radar sensitivity calculations vs. range. Medium- or high-fidelity representations of the terrain and atmospheric refraction effects can be included to enable visibility calculations that determine when low-flying objects, such as terrain-following aircraft, are observable or occluded by the terrain. RAMS currently uses a statistical detection model for air vehicles, together with a functional model of track accuracy based on observed SNR (signal to noise ratio), number of looks, time in track, and time since maneuvers. This level of fidelity has been selected to refine and assess performance of multimission radar resource management algorithms. Higher-fidelity signature models and detailed signal processing and tracking algorithms for air vehicles can be implemented to support more detailed air defense system design and analysis.

RAMS currently has a functional model of battle management and interceptor flyout, using a simple set of rules or a script to determine which targets are engaged and when they are engaged; this information defines when firm tracks need to be established and discrimination needs to be complete. The framework is sufficiently flexible to allow higher-fidelity battle management and interceptor flyout models to be incorporated. To demonstrate the potential for sharing information among multiple radars, RAMS was configured with two radars jointly observing a long-range ballistic missile event from different locations, with interceptors launched from a third site. Data from these sensors were fused to highlight the potential benefits of coordinated, multisensor operation.

Figure 3. RAMS screen close-up

By providing a simulated radar system prototype that has the fidelity to represent its operation in detailed scenes, RAMS supports development and refinement of concepts of operations, software architectures and software requirements. RAMS has produced simulated data for studying phenomena, exploring algorithm concepts and maturing algorithm implementations, including:

  • Signal processing algorithms for multipulse processing, super-resolution and detection characterization.
  • Advanced tracking techniques that provide improved accuracy and continuity in dense scenes.
  • Advanced BMD discrimination techniques that identify lethal and non-lethal objects.
  • Multimission pulse scheduling algorithms that develop schedules for all radar transmit and receive actions.

As a development program progresses, the RAMS representation of the radar can be verified, validated and accredited (VV&A), and then used to demonstrate that the system requirements are achieved. The VV&A version of RAMS can also provide pre-mission predictions of flight tests.

RAMS provides a set of capabilities that can be leveraged and further extended to address the needs of future programs. Selectable fidelity via configuration files readily supports spiral evolution of a design, in which functional modules are incrementally replaced with more detailed modules and performance predictions are updated iteratively. Its flexible framework maximizes the use of common modules while still allowing system tailoring, via specialized plug-in modules, to accurately represent specific radar systems.

David Cebula, Ph.D.

Share This Story

Top of Page