Guest opinion: Solid state security - A potential threat

Story

July 14, 2010

Mark Downey

White Electronic Designs Corporation

History has proven that an effective strategy for protecting the security of information is one of the most critical factors in wartime success. When it comes to solid state information storage, the biggest threat to security might reside in the technology and manufacture of the storage device itself. Reliability is also imperative in ensuring information integrity - and ensuring warfighter safety.

Defense and aerospace applications such as rugged computing and mobile data acquisition systems were among the first to utilize solid state storage technology. This adoption leadership was realized because designers in these markets understood the harsh environmental survivability benefits of solid state technology when compared to mechanical hard drives – and decided solid state was worth the price premium. Solid State Drive (SSD) developers then began creating innovate SSD technologies in response to the defense market’s values, requirements, and priorities. Thus, the door for implementation of solid state storage technology was opened wide, and myriad benefits ensued.

As time marched on, development cycles continued to turn out SSD technologies that reflected the defense and aerospace market’s desires. These SSD wares also positioned the U.S. military with the unsurpassed storage systems that we have come to appreciate and value. However, the recent explosion of SSD technologies into more commoditized, commercial applications has disrupted this trend, stealing the focus away from the defense and aerospace markets in exchange for the handsome prospects of larger markets. The pricetag: A downturn in SSD security and reliability that could put our nation’s information – and warfighters – at risk.

The way it was

Throughout the ’90s, design engineers developed system technologies for defense and aerospace applications and garnered the full attention of the many SSD developers. By being the most significant adopters of the technology, defense industry needs came first and helped to mold and define development goals. Most SSD providers were relatively small companies and would readily go the extra mile to customize or design harsh-environment, ruggedized products needing to fulfill specific security and reliability requirements. However, over the past few years, SSD production has begun to explode in both the enterprise and notebook markets, in turn forever changing the SSD provider marketplace.

The way it is

Today, military systems designers are faced with a different paradigm. No longer do they define the storage requirements they would like to see from the SSD providers, but rather they must settle for what might or might not be available. They find themselves wading through a large number of SSD technologies, most of which hold hidden risks or subtleties that might be difficult to discern:

How is the SSD actually tested?
Will the device meet the specification over the entire life of the product?
Will it survive extended temperature operation and if so, for how long?
How will obsolescence or life-cycle changes be managed?

Program requirements continue to push cost savings and, therefore, the use of Commercial Off-the-Shelf (COTS) products, further complicating the already cumbersome decision process. Today, more than ever, directives are often based on what is available rather than what is possible. Designers find themselves working to solve problems associated with using commercial-environment products in a harsher and more security-critical defense and aerospace environment where reliability is key (Figure 1). Typical questions might include:

How does one make this Serial Advanced Technology Attachment (SATA) connector work in a high-vibration and high-shock environment?
Is this reliability information really valid at industrial temperatures?
Can one get an SSD built with high-grade flash devices that are burnt-in to ensure a higher level of fielded reliability?

Figure 1: The potential consequences of an SSD malfunction in military applications can be disastrous.

(Click graphic to zoom by 1.9x)

The bottom line is that it is becoming increasingly more difficult to identify an SSD manufacturing partner willing to address these real technical challenges, never mind one whose original design intention embodies the reliability and security priorities that mission-critical environments demand.

Security threat

One conflict evolving with this recent trend relates to the topic of security. As a primary requirement for successful adoption, the larger and more dominant SSD markets continue to pressure providers for lower costs. To meet these reduced costs, trade-offs are made that can threaten overall security in military applications. Price pressures might force SSD manufacturers to utilize non-U.S. designers and labor, resulting in potential security risks for U.S. DoD and U.S. military applications that might go unrecognized.

SSDs are constructed around a controlling management device, often a processor, which also contains firmware. This construction makes them susceptible to certain types of attacks if proper precautions are not taken. These attacks might be subtle, only preventing the erasure of critical data in times of duress – or they might allow some master key to be used, permitting an enemy to extract encrypted data without the original key. Thus, maintaining control over security in tomorrow’s warfighting systems becomes more challenging, yet increasingly urgent and essential. Data is moved and stored throughout networks like never before, and the storage element of the overall system is one of the most vulnerable areas. The question of securing this storage effectively must remain in the crosshairs of design priorities and cannot be ignored or traded to save a few dollars. Security remains of paramount importance and must not be diminished.

How does one accomplish world-class security with products that are sold globally? Is it even possible? Isn’t “standardized security” akin to “fat-free oil”? Standardization is important, as long as we find a way to maintain security; where conflicts arise, one must choose security. Pentagon spokesman Bryan Whitman in a recent press conference warned, “As you develop those [technologies], you have to be mindful of how the enemy can counteract any technology you have. That’s why you always have a constant review process in place to not only improve that capability, but address any vulnerability it may have.” His statement preceded a discussion of a security breach of the unmanned Predator drones used in Afghanistan and Iraq. Shi’ite fighters in Iraq used open market software costing less than $26 to intercept and monitor video feeds from U.S. drones.

Such incidents demand attention and require consideration as a threat to warfighter network security. How does one place a value on, or define a cost target for, security? Designers must be aware of these risks and consider that SSD choices will continue to be influenced by the requirements of the larger enterprise and/or notebook markets.

Reliability threat

Along with security, a second area of defense industry concern regarding SSD technologies is reliability. While many have had a PDA or laptop freeze up at an inopportune time, the potential consequences of that kind of malfunction can be disastrous for the warfighter. The soldier under fire in the mountains of Afghanistan would not want to know that the SSD on which he is relying has components designed for use in an air-conditioned computer lab – or that design trade-offs were made such that it could meet the $150 price point required for sale into the soon-to-be explosive Netbook markets of India. The implications of using commercial-grade SSDs in military applications designed for such disparate markets could be tragic (Figure 2).

Figure 2: There are problems associated with commercially designed SSDs used in security-critical defense and aerospace environments, where reliability is key.

(Click graphic to zoom by 1.9x)

A call to action

So what is the answer to this perplexing paradigm? Government agencies and the private sector must work together to establish clear and definable safeguards against such evolving SSD security and reliability risks. Price pressures will continue to dominate the decision process until something tragic triggers a reactive response in the supply channel.

With the shift of attention away from military/aerospace requirements comes the risk of settling for mediocrity in solution sets. Both reliability and security are at risk of being watered down or lost with the disruptive, commercially driven changes in SSD development. The warfighter in the field will no longer be getting the best of what one can design for them, but rather the closest fit of what has been designed for other more dominant markets. Although this might appear as a subtle difference today, given sufficient time and product design cycles, the technology gap between would have been and what is will certainly expand.

References:

Malicious Firmware Could Sabotage Military, Security Systems, By Stew Magnuson, National Defense magazine, Feb. 2010, http://www.nationaldefensemagazine.org/ARCHIVE/2010/FEBRUARY/Pages/MaliciousFirmwareCouldSabotageMilitary,SecuritySystems.aspx

Mark Downey is former Director of Defense Technology at White Electronic Designs. He has 18 years’ experience in the electronics industry in the areas of solid state storage technology, high-speed memory bus architecture, electrical simulation, thermal management, complex microelectronic packaging, and process development. He co-holds a U.S. patent in the area of dense memory module packaging. Mark also holds a BSEE from the University of Massachusetts at Lowell.

White Electronic Designs (now Microsemi) 602-437-1520 www.wedc.com