Embedding computing and the cloud

News

October 07, 2016

CHELMSFORD, MA. Tactical, or as some call it, combat cloud computing will need to leverage more than commercial data center technologies, but will also need embedded computing – hardware and software – to enable it to work in secure battlefield applications.

“With every platform becoming a sensor and consumer of information in the military, you will see combat clouds on aircraft carriers, fighter jets, and in ground vehicles,” says Richard Jaenicke, director, market development and strategic alliances at Mercury Systems in Chelmsford, Massachusetts.

The term “combat cloud” was coined by Lt. Gen. David A. Deptula, U.S. Air Force (Ret.), Dean, The Mitchell Institute for Aerospace Studies, in his paper titled “Evolving Technologies and Warfare in the 21st Century: Introducing the ‘Combat Cloud,’” he says.

In it, Deptula states that a “combat cloud” is essentially “an operating paradigm where information, data management, connectivity, and command and control (C2) are core mission priorities. The combat cloud treats every platform as a sensor, as well as an “effector,” and will require a C2 paradigm enabling automatic linking, seamless data transfer capabilities, while being reliable, secure, and jam-proof. The combat cloud inverts the paradigm of combined arms warfare – making information the focal point, not operational domains. This concept represents an evolution where individually networked platforms – in any domain – transform into a “system of systems” enterprise, integrated by domain and mission-agnostic linkages.”

Enabling a combat cloud is natural extension of embedded computing, say Mercury Systems engineers.

“The embedded computing industry uses using the same kind of architecture that the cloud industry uses for data centers, but high-performance embedded computing companies can provide the necessary high level signal processing and real-time software necessary for enabling clouds and big data centers in military applications,” says Shaun McQuaid, Director, Product Management at Mercury Systems. “For example, when you look at a warfighter dealing with multiple sensors coordinating with multiple platforms – air, ground, and sea – you see a big data problem as there is much imagery needed to be processed at high bandwidth and in real time. Commercial cloud applications do not have the signal processing hardware and deterministic software necessary to enable this for a combat warfighter scenario.”

You start with the basic cloud technology then apply the embedded technologies to it to make it trusted and secure enough for tactical applications, Jaenicke adds.

Design challenges

To understand the design challenges for enabling a combat cloud, you need to understand how these cloud applications may differ from commercial applications.

“Cloud security for the commercial world is much different than cloud security for military tactical applications,” McQuaid explains. “No one is likely to storm the doors of your run-of-the-mill commercial data center to go after their servers. But if a manned or unmanned platform gets taken down unexpectedly and falls into enemy hands it will need extra security such as encryption and methods to prevent reverse engineering. The military also needs to make the technology associated with a combat cloud and data center more deployable through rugged packaging and thermal management techniques, and also enable unrestricted bandwidth via high-speed fabrics such as Infiniband.

“On the bureaucratic side, you have to take a wider, more macro view, by taking into account military platforms communicate with other platforms on the battlefield such as what Gen. Deptula writes about,” he continues.

Hardware vs software and security

Which is more important when securing the cloud: protecting the software from cyberattacks or the hardware used in the data centers?

“I think you have to take both into account,” McQuaid says. “You can have all the hardware in world, but if you don’t have software that works for you then you can’t manage open interfaces and enable portability. Without that, tech refreshes become more difficult.”

From a cloud point of view, once you have the hardware piece developed, “then you start hardening from the software side, where the challenge then is how to make the software solution protected from cyberattacks, secured from reverse engineering, etc.,” Jaenicke says.

Mercury will be announcing a new line of secure, rackmount servers designed specifically to address combat cloud applications, McQuaid says. “The family also takes technology from the OpenVPX side and the ATCA side and applies it to rackmount architectures. In addition to the cloud applications, it can be used for any kind of mission computing and data exploitation application.” (See Figure.)

For more on embedded technology and cloud computing, read a blog by VITA Chairman of the Board Ray Alderman, titled “Cloud, fog, mist, fluid, blockchain, and other things that irritate me” at http://mil-embedded.com/guest-blogs/cloud-fog-mist-fluid-blockchain-and-other-things-that-irritate-me/.