Peter Carbone, chief architect at Nortel for nearly three years and the company's chief technology officer since late November, talked with Computerworld about what he called the "invisible network," technology innovations that will bring about next-generation networks and how Nortel's customers see the company. A Nortel veteran of 28 years, Carbone also said that security is an ever present and growing concern for companies as they move more of their business operations to the network.
Excerpts from that interview follow:
What are enterprise customers you encounter most concerned about regarding networking?
I've been focused on working with customers to understand where they need to be applying technology, and they've reached [out] to Nortel as a trusted adviser. The top worries of customers continue to be with cost and, increasingly, security, as they put more business operations onto the network. Also, network reach is a worry with the mobile workforce [who] work at home and work in other countries, so technologies like voice over Internet Protocol and how IP is evolving to multimedia and Session Initiation Protocol are on their minds.
I personally am a mobile worker and spend time traveling. With voice-over-IP implementations, my office is wherever I open my PC. I can use Wi-Fi and high-speed access and fire up a VPN client to access corporate information. If you phone me at my office number, it rings me wherever I am. I can support full voice- and videoconferencing capability and some of the presence capabilities, all of which actually change how you work. I can just drop into an important meeting and people will be surprised that I'm actually in Japan or England.
Since you mention voice and Wi-Fi, are you developing dual phone capability at Nortel?
We have dual-mode capability between 802.11 and GSM and CDMA [wide-area cellular technologies] in the trial phase to allow a user to jump between networks. At Nortel, we have Wi-Fi across the campus with visitor hot spots that allow visitors to get onto the Internet. So, in a sense, the network is becoming invisible to people.
Another example of that invisible network is that I was recently at a partner meeting in someone else's facility and fired up a client to send and receive voice over IP and data. It was based on Nortel MCS [Multimedia Communication Server]. It's becoming so easy to use this technology, and in some ways it's easier to use than a telephone, because if you're in another country and are not familiar with dialing there, you can rely on your own VOIP phone and its familiar interface. And regarding cost savings, we found out when we went to a VOIP system in our own environment that we were able to replace cell phone calls for a savings of US$15 million a year over the past two years.
What's the most exciting technology of late for enterprise customers?
We're pulling a few technologies together where we have sensor technology in Ultra Wideband to track individuals, and we've bound that to MCS and the Wi-Fi infrastructure. So we have put together a medical application that we are currently demonstrating which uses mostly generally available technology that is dynamically adapted.
For example, in a hospital, a doctor walks around with a Tablet PC that is Wi-Fi-equipped, and he also has a name badge that has a sensor. We can detect over UWB which patient he is seeing and transmit to the doctor that patient's current medical record. And the system automatically updates medication and other information so that all the personnel would all have common access to that info. Also, if someone tries to contact the doctor in the hospital, he gets an indication on the Tablet and uses a BlueTooth headset working through the PC. All the voice communication can also be transferred to the doctor's desktop phone set once he's in the office. So you could have a video medical image sent to the laptop, which could be transmitted to a big screen in the office. All of that is really happening, with data collected from the sensor network and processed and coordinated with GPS. We have that health care application in trial and almost all the products involved are from Nortel.
What are the important networking technologies coming in the next two years?
The technologies of the next two years in the future revolve around the changing cost structure of network facilities customers gain access to and, secondly, with personalization of services with broadband and virtualization.
In the cost area, quite a bit is going on with wireless, especially with OFDM [Orthogonal Frequency Division Multiplexing] and MIMO [multiple-input, multiple-output]. OFDM is a foundation for CDMA, UMTS and WiMax, and we're going to see various deployments of this technology on fixed and nomadic and full cell systems that will make mobile communications as capable as fixed communications for personal use. OFDM dramatically increases the amount of applications that will be built for this infrastructure, such as doing training or any desktop video streaming or videoconferencing application, which is hard to do across GPRS. We're working with Intel to get consumer devices to use OFDM and MIMO. MIMO is antenna technology that will increase capacity by 10 to 40 times of what we have now.
Pretty big changes are coming in the packet core. As we've added more layers of technology into the core of the network, it's starting to get more expensive rather than less. But optical technology is continuing to drive much better price and performance. For example, dispersion compensating optics gets rid of all repeaters used in a large network by a service provider or enterprise. It's starting to roll out. With an optical modem, it looks across a stretch of fiber and will do the predistortion correction and the signal goes through. As many enterprises start to deploy SANs and connect bigger sites, optical networks should be a cheaper alternative to lines available today, perhaps 40 percent to 70 percent [cheaper].
Also, carrier Ethernet is coming to provide quality of service and quality of experience, as well as security, for subscribers. It's 30 percent to 40 percent cheaper than the traditional way of extending packet services into the Metro Area Network. It means customers of carriers will be able to get cheaper VPNs. That's something we'll see in a two-year window.
Also coming, SIP capability and sensor capability will be merged, meaning location will become relevant in the use of devices. And that ability will change how people use the network, and change attention away from what device a person is using to just contacting the person. That, in turn, changes the jobs for users and IT who manage circuits or set up an audio or video bridge. Instead, you'll go into a voice or video session and add and delete sessions through a SIP-based infrastructure and simply engage the other parties.
It's nice to hear things might be simpler!
Absolutely, things have to get simpler, because if it's complicated, people won't use the technology. In another way, if you add security and force users to keep remembering passwords, it won't be used. The network has to be more invisible to work in.
How are customers reacting to the new Nortel when you go out and talk to them?
I can't address marketing, but from our customers' perspective, they know when we put something out that it will work. In fact, they are often frustrated with us that we didn't tell them aggressively that we had some of this new technology in the works. Few companies other than Nortel can preintegrate SIP and sensor technology and other things successfully on their own platforms.
Nortel got back into doing its own research in the mid-1990s, so how many R&D staffers do you have today?
And customers really aren't expressing worries over previous Nortel financial problems, management changes and the impact of such things on technology development?
They have no concerns about talking to us about our technology proficiency. Universities talk to us for this reason, as opposed to getting sales pitches. In fact, we're sought out by a significant part of the university community, including the MITs of the world.
How does it feel to be facing 2006 after the recession and other problems from the first half of this decade?
We're coming into a technology age where there's going to be a fair bit of change in fundamental things. The coming age of next-generation networks is an age of personalization where you aren't expected to adapt to the network anymore, and it's adapting to you. Security and passwords should be simpler and brokered by the network that you're using. Network managers won't really deal as much with single vendors, and we're going to be reaching and expanding to move back and forth across public and private networks easily. Networks will really become an afterthought. Fundamentally, they will be not as complicated. They'll be simpler so that they just work, which is what people want.