Home / Companies / People /  Internet’s architecture doesn’t handle mobility as well as it should: Vinton Cerf

Widely known as the “Father of the Internet", Vinton G. Cerf co-designed with Robert Kahn the protocols and basic architecture of the Internet. Currently, he serves as chief Internet evangelist at Google Inc. He was in New Delhi on Tuesday to talk at an event organized by industry body Federation of Indian Chambers of Commerce and Industry on innovation, jobs, the Internet and the Digital India programme, and met telecom minister Ravi Shankar Prasad before the event . In an interview on the sidelines of the event, he talked about the evolving architecture of the Internet, future trends, and security and privacy concerns that have plagued the Internet. Edited excerpts:

On the need to overhaul Internet’s architecture

The first thing to observe is that the basic architecture of the Internet has remained relatively stable over the 40 years. On the other hand, it has been evolving, the details have been changing. Moving from IPv4 to IPv6 (Internet Protocol version 4 to version 6) is different address formats. Another thing which wasn’t a part of original architecture and was added later was the firewall. Third thing was network address translation, a way of trying to share the IPv4 address space out because there wasn’t enough of it, which is actually harmful. We did not know in 1973 that 4.3 billion terminations were not going to be enough for Internet. By 1996, the standard’s group managed to figure out we needed standardized IPv6. So those kinds of things have all evolved.

Domain name system security is another example where we started out with very trusting and safe environment and now of course it is a part of a global and public system, and we have bad guys out there causing trouble. So we use cryptography for the domain name system for protection and confidentiality as we do it at Google.

There is still more to be done. We don’t handle mobility as well as we should. There is not any broadcast mechanism right now, there is a multicast, but no explicit use of radio broadcast capability. We should use the capability of radio, which has the ability for one transmitter to reach multiple receivers, and our protocols do not explicitly do that, they could.

There are a bunch of other things that could be done, and I want students, professors, researchers and private sector to feel free to explore new ways of enhancing Internet service. The most recent evolution is called open flow, which was also developed at Stanford University where I did my original work. The open-flow design is an innovative idea, where basically they said why should we be constrained by using just the bits in the address space for packets to figure out how to route them. What about other bits in the package? Could we use those to help us decide which way this packet should go where—so they broke that limitation, then they created a separation of a data point and control point which increased reliability and security in the system.

Google recognized this open flow, put it in our backbone network, and we are operating that way. This is another example of the continued ability of the Internet architecture to ingest new ideas and allow people to experiment. Over the course of years, a whole bunch of real-time protocols have been developed... New applications have been implemented, so there is no end in sight. There is an opportunity to keep inventing and reinventing this architecture and the way in which it works.

On security and privacy

Certainly one of the things that have to happen is that we need to develop better operating systems and browsers that are more resistant to some of these problems. Google is doing that. There are Android and Chrome, for example, and others are pursuing that. We are using cryptography for stronger authentication to keep people from highjacking your network and resources. We need to teach people how to be more aware of what their risks are, just like we tell children to not get into strangers’ cars. We need software that will help us detect problems in our configuration. We need browsers that are more conscious of the possibility of the malware.

On future trends

A lot of devices that are programmable and have capability are becoming part of the Internet environment. The first thing is the device can keep ingesting and gain new functionality, which means without changing the physical device, you increase its utility by making it do more things. A really good example is Tesla cars. Frankly, it is all software. They keep downloading new software every night to do new things—and that is the perfect example in the context of the Internet of Things. It keeps adding to its functionality based on programmers’ new ideas. So that is going to continue as a trend.

The other thing is increased use of mobile operation, increased use of radio, increased sharing of radio spectrum, different kinds of ways to allow multiple modulation schemes to coexist in the same frequency band—which you need to get more utility out of the radio spectrum.

The system is going off the planet, there is interplanetary extension already in operation, that will continue. There is even a project right now to design a spacecraft to go to a star in 100 years’ time, and my problem is how do you communicate over the light year distance—haven’t solved that problem yet (laughs)…so what I foresee is a continuing evolution of both network technologies and application space.

You can see what happened to mobiles. As soon as they become smartphones, they become a platform for new applications to be built at. All of this is very refreshing and it isn’t over yet.

Subscribe to Mint Newsletters
* Enter a valid email
* Thank you for subscribing to our newsletter.

Never miss a story! Stay connected and informed with Mint. Download our App Now!!

Edit Profile
Get alerts on WhatsApp
My ReadsRedeem a Gift CardLogout