• HOME
• NEWS & EVENTS
• THE GRID

THE GRID

Vinton G. Cerf: Security and the Internet
By Miriam Moss, DISA Corporate Communications

Vinton G. Cerf, vice president and chief Internet evangelist of Google, is responsible for identifying new technologies to support both Google's Internet products and services. Cerf, considered to be one of the "fathers of the Internet" because he developed the Internet's governing protocols, spoke to the conference audience about the security pitfalls of the Internet and a need to develop and identify solutions.

The reliability of the Internet is undermined by the both poor security and poor software design and is particularly so in the case with Internet browsers and operating systems, according to Cerf.

"The [Internet] browsers are probably the most open sore in the security universe, where you can just ingest software just driving by a Web site. The browsers have sole access to the resources of the operating system," said Cerf.

The problem is that because users have so much direct access to the functionality of the operating system itself, it makes it quite easy for the operating system to be exposed to viruses and malicious programs, according to Cerf.

"In the past, attacks were made directly against the operating systems. Now browsers serve as the access point for attacks, as it is easier," said Cerf.

On the battlefield, getting authentic information to the warfighter is essential to the warfighter's well-being. Work in this area is still needed according to Cerf. Solutions in this area will not be easy to identify. Cerf believes that currently, it takes so long to set up some of the security systems needed to transmit information to the warfighter, especially on the battlefield.

One area that engineers and researchers have examined or considered in an effort to resolve the security issues is transmission control protocol (TCP), which Cerf and Bob Kahn invented in 1973 under the auspice of the Department of Defense (DoD). TCP is core protocol of the Internet that manages the size of communications between Web servers and Web clients and makes those communications possible.

"[TCP] does not solve all of your [security] problems, and in particular it does not solve problems in the tactical environment. … TCP is fragile in a field environment. … It gets a lot of interference," said Cerf. "TCP just doesn't work. We have to do something about that. We can't just sit back and hope for the best. We have to deal with that problem."

Cerf believes that the issue of securely delivering information and capabilities to the warfighter in a tactical environment is further complicated by the timeliness of delivery. This is one of the main issues with TCP; there is a delay in delivering timely information that is secure, especially in the tactical environment. TCP attempts to optimize the accuracy of the information that is delivered and not the timeiness of the delivery. Delivery time can not be ignored or extended on the battlefield as lives are often at stake.

"I'm a keen believer in systems engineering. I'm a believer in trying things out, finding what doesn't work and changing it until it does. I'm a complete believer in having an overall architecture that you can work with, that fits together," said Cerf. He challenged the audience to rethink the procurement process to include testing until we get it right, but we have to understand the importance of timing. For instance, he asked how long will it take to deploy the solutions we identify in a secure fashion to the warfighter.

DoD's acquisition process is broken, and there is an over dependence on requirements specification according to Cerf.

"The way we live now, it takes so darn long to get things implemented that by the time they roll out, it's too late," said Cerf.

We are working on technology today that will not be available for deployment to the warfighter, who is currently in harm's way, until 2018. By 2018, the technology of the day will be dramatically different than what we see today, and there will be new threats that will need to be addressed that we have not yet even conceived of in our minds today, according to Cerf.

The technology we are designing today for 2018 will probably be so outdated when it is deployed that its usefulness will be minimal if useful at all. Cerf believes that whatever we are designing today is more than likely yesterday's technology as far as the warfighter is concerned. He challenged the audience to think about extending adaptability to address this issue.

"It's not just the [frequency wavelength agile] radio that has to be adaptable, it's the whole system. The protocols that run the system have to be able to adapt as well," said Cerf. "We need systems that can self-configure."

"It's okay to look at ideas and technology that didn't work before because the rules of the game have changed," said Cerf. In other words, another look needs to be given to previous technology as it can provide a foundation for future solutions.

The networks need to have assurance and integrity built in from the start, while taking into account rapid motion of data between various access points on the network.

Developers have to get to a place where they are developing systems that have resilient architecture, strong authentication capabilities, and further exploitation and exploration of delay and disruption tolerant networking (DTN) Phase II, said Cerf.

Our solutions need to change as quickly, if not more quickly, than the changes the warfighter is experiencing on the battlefield. Cerf suggests that on possible path to achieve this is by resurrecting the Federal Research Internet Coordinating Committee (FRICC). Still, the need goes beyond the FRICC. If and when select entities, agencies, and departments that possess the budgetary authority to create and deploy adaptable solutions come together for brainstorming sessions, the groundwork will be laid for the development and deployment of needed solutions to the warfighter, said Cerf.

 

Return to The Grid Homepage