According to a 23 January DARPA solicitation, the cost, complexity, and limited availability of military ranges mean that operational training and system test and evaluation will increasingly take place in completely virtual environments. Furthermore, the increasing adoption of artificial intelligence (AI) will require orders of magnitude of more test and training time.
Live Virtual Constructive training has really expanded the field of virtual reality. LVC involves “three distinct classes of military simulation: live simulation, virtual simulation, and constructive simulation”. Live simulation involves real people operating real systems, such as having a real soldier practice aiming a weapon mounted on an aircraft at targets below. Virtual simulation involves a real person operating a simulated system. A pilot using a flight simulator would be an example of virtual simulation. The previous two techniques have been used for years by the military but constructive simulation is relatively new due to advancements in virtual reality. Constructive simulation involves simulated people operating simulated systems like a simulated enemy attempting to thwart the plan of the mission.
LVC requires the secure, collection, assimilation, and exploitation of information and the timely transmittal of that information available to the user (often in disparate locations) in the form of intelligence and decision making-quality information.
A LVC event can happen without everyone being in the same room. Before, military training involved flying the team members to a specific training site to learn specific skills over a period of weeks. Now with LVC, everyone can participate in the exercise literally from the comfort of their own homes as long as they have the equipment.
DARPA sets goal for ‘virtual’ digital RF environment
The US Defense Advanced Research Projects Agency (DARPA) has revealed plans to develop a virtual environment able to emulate the characteristics of an increasingly complex radio frequency (RF) environment in real time.
Known as the Digital RF Battlespace Emulator (DRBE), the programme is being funded under Phase II of DARPA’s Electronics Resurgence Initiative (ERI). Representing an investment of more than USD1.5 billion over a five-year period, the ERI is a five-year activity intended to ensure far-reaching improvements in electronics performance well beyond the limits of traditional scaling. As a part of ERI Phase II, DRBE aims to apply the benefits of domain specific processing architectures to defence systems.
As a result, a new training paradigm – referred to as Live, Virtual, Constructive (LVC) – is being increasingly adopted for training modern fighter aircraft pilots, where simulator time is used to augment real-world aircraft flight hours. However, current LVC systems rely on conventional computing to create their virtual world and, as such, lack the veracity to accurately model the scale and high bandwidth of inputs and outputs to and from complex sensor systems.
Also, as critical systems – such as radars and electronic warfare (EW) equipment – grow their use of AI, the ability to train and test the systems themselves in a virtual environment will become more important. The challenge here is that today’s processors (general purpose processors, field-programmable gate arrays and graphics processing units) cannot deliver the high-performance computing required to effectively emulate the complex interactions of an operationally relevant number of RF systems.
To redress theses shortfall, DARPA’s DRBE programme seeks the creation of a new Real-Time High Performance Computing (RT-HPC) environment that effectively balances computational throughput with extremely low latency to create what it claims would be “the world’s first large-scale virtual RF test range”. It is estimated that the HPC must achieve roughly 20 PFlop/s with a latency of less than 2.5 micro-seconds.
DARPA intends that the DRBE system will enable numerous real RF systems (such as radar and EW systems) to interact with each other in a fully closed loop RF environment. Computationally, DRBE will be responsible for calculating the interactions of high bandwidth RF waveforms with emulated objects in a physical environment (such as ground clutter and radar target returns).
DARPA’s DRBE will develop application-specific integrated circuits (ASICs) that realise this low-latency, high compute capacity vision, and will assemble these ASICs into a multiprocessor high performance computing system. It will then design and integrate the necessary tools to demonstrate the use of the RT-HPC as a large-scale virtual RF test range.