Future perfect

Wireless enterprises, changing business dynamics and integration of communication with backend applications are all an assured part of the future for Philip Edholm, CTO and VP of network architecture, enterprise at Nortel, as NME finds out.

  • E-Mail
By  Sathya Ashok Published  March 23, 2008

Wireless enterprises, changing business dynamics and integration of communication with backend applications are all an assured part of the future for Philip Edholm, CTO and VP of network architecture, enterprise at Nortel, as NME finds out.

Can you describe the law that you have postulated? Are we on track to meet its goals?

Be very careful - adoption curves are not positive or negative. They are a matter of comfort with technology and the curve tends to get steeper.

Back in 2002 we deduced that devices would eventually be connected to three networks. The idea is you have a sea of 1x bandwidth. You occasionally go to islands which are hotspots or hotel rooms and that is ten times as much bandwidth and then occasionally you go to a dock, which is your office and this is a 100x more bandwidth.

So what actually started this was a thought process on a very different area, which was if you got a bunch of things sending you information - like e-mail, video files, IM - what you really wanted was a system which could manage when you wanted them delivered - bigger and less important files when you are in a high bandwidth area and the other way around.

When we were trying to look at the bandwidth we noticed that all of this seemed to be growing at the same rate. If you put it on a chart that says time vs bandwidth, where there is a log scale, what you end up with are three lines that are growing steeply upwards. Then we began to realise that a few things came out of this thought process.

The first realisation was that as these lines went up, if you looked at a specific technology, it moved under the lines and a specific app would move from one to another, across the growing bandwidth.

The other part of this is that these lines in lowering bandwidth options will coincide with employees, employees and partners and employees, partners and customers. So as a company I can understand the devices and services that are going to move from being things that I do inside the office to outside the office, that was kind of one thought.

The other thought was there are limitations to requirements for bandwidth. You can argue that this is a bit controversial but our eyes can only see non-peripheral vision and the argument is there is another amount of bandwidth to fill the peripheral space. With probably around 100Mbit/s, you can replicate a visual field that is equivalent to your eyes.

So the argument goes that there is a point at which this bandwidth at the lowest point versus this bandwidth at the higher point, the apps that you can do up here are not really important and therefore the bandwidth at the lower scale is sufficient. I believe this will happen between 2010 and 2012. This is when you will also start to see a wireless enterprise.

Now what's really interesting is that there are two value sets to a wireless enterprise. The obvious one is that if you don't put in the wires you can save the cost of the wiring, you can potentially reduce the cost of some equipment and it makes rearrangements of buildings much easier. But the more interesting thing is that you can actually make the environment a mobile structure.

Instead of cubicles and offices, what if there were desks, we have an open space, we have partitions and when we want to work on a project, we just roll our desks together, put some partitions around us and create a virtual office and because we have wireless we don't have to plug in.

All of a sudden you have to think about the workplace becoming much more dynamic. Most people today that work in most environments are moving from the fixed space environment to a teeming collaboration thing. So this is actually pretty predictable today.

But if you follow this concept of bandwidth and when we project the line out, sometime around 2018-2020, the bandwidth you can get anywhere will be just about the same. And by then you have actually changed from being flexible in the office to flexible out of the office.

If you look at this from the enterprise point of view, the employee is no longer tied to going to work. We coined a phrase in Nortel about five years ago, work is something you do, not a place you go to.

But businesses are still thinking of doing business in our buildings. The moment we start to move that out of the office, with higher bandwidth and wireless, people can do business anywhere and I think there is a big re-thinking of business coming in the next five to eight years.

Instead of business being somewhere you go to you can do business anywhere. And it is not the business anywhere of the internet, but employees can go and do business where the customers are. That has a dynamic change in business.

The whole law of bandwidth is an observation. If you look at Moore's Law, it is an observation, but it does allow you to draw a framework to predict where things are going to go.

Are we on time to meet the targets that you have specified?

I think we are. The desktop line has gone to1Gbyte on the desktop and we are putting 10Gbyte on the risers; that line I am not sure if it is not peaking. I don't think we are going to put 10Gbyte for the desktop not because we can't but because there is really no demand for it.

The wireless LAN line, 802.11n combined with multimode antennas where you have not just a single antenna and frequency reuse, combined with some of the other technologies like jamming management, has the potential to grow over the next three to four years.

If you look at the wireless line, the technology is there, whether LTE (long term evolution) or WiMAX. I think the question there is less a technology question and more of an adoption question.

Will the carriers that have strong commitments to 3G migrate to 4G? That line continues technologically. From an adoption point of view, it may be adopted first outside of Western Europe and maybe US.

For wireless LAN and 802.11n, enterprises will need to change their access points and switches, so the investment made some years ago probably has to be changed to make that transition.

A lot of people are starting to see the wireless enterprises and are adopting technologies. They say, I am going to leave the wires that I have now but I am going to put in an infrastructure so that I don't have to worry about the wires in the future.

So, if your desktop has been wired, it's there and it works but if I rebuild part of the building I won't put in wires. I think that will start to happen around 2010-2011.

How do you see this trend affecting Middle East enterprises who are putting in place fresh wired infrastructure?

One of the things that a lot of this depends on is whether the developers are conservative or not. I don't know specifically but I think a lot of the developments here because they are so large tend to be more conservative on how they adopt things. If you look typically at technology adoption it starts first in North America, Scandinavia and then Western Europe and then moves to the rest of the world.

I am not sure how that adoption curve will look here but I think what's going to happen is that you are going to see some major deployments happening. 2010 is probably very early, I think then you will see it more in branches than in HQ. That is a very important distinction. A branch office where you have 40 people, those are the places that you will see it first.

I think what will happen is that it will take 12 to 18 months for people to get used to it. Since most of the construction happening here are large greenfield deployments, they will be in the second tier of adoption.

It's not because the technology is good or bad, it's the level of comfort. We saw the same thing with VoIP. I am not being negative. Be very careful - adoption curves are not positive or negative.

Add a Comment

Your display name This field is mandatory

Your e-mail address This field is mandatory (Your e-mail address won't be published)

Security code