Proxim has come a long way. It was probably the best-known wireless LAN player at once stage. Now it'sfacing de-listing from the Nasdaq stock eschange.

Chief executive Pankaj Manglik has a plan. He's emphasising point-to-point links, and planning non-standard WiMax, and he thinkgs it's still too early to get into 802.11n Wi-Fi. And he also thinks Intel might be preparing to back down and accept LTE as an alternative to WiMax. 

We spoke to him earlier this summer.

So how did you get to Proxim?

I left Aruba to start another company and sort of started another company. I was looking at different things, and Proxim came along. Everyone knows about Proxim, and there was a turnarond situation. I looked at all the assets that they had

When you are a start up you have no brand name - and Proxim is a good brand name. When you are a startup you have no channels - and Proxim has great channels. And when you are a startup you have no products - and Proxim has good products.

I thought, I can do this - it should be a walk in the park. Which it certainly isn't!

We have focussed on product development in a big way. If you take the last 18 months and the next six months, we will have upgraded our entire product line, across all product segments, with new products.

We divide our products into three areas. The first one is wireless LAN access, and that includes both indoor and outdoor wireless, and mesh. The second part of the business is point-to-multipoint (P2MP) which is fundamentally WiMax. The third part of the business is point-to-point, which includes our products for carrier backhaul. It also includes our millimetre wave products, that can deliver 1Gbit/s over 1km.

Among those, I'd say about half the revenue comes from the WiMax, and the other two are roughly even at 20 points each, and there's about ten points for service and that kind of stuff. About fifty percent of our business is in the US, 25 points in Europe, I'd say 15 points in Asia, and ten in Latin America.

Overall our business is fairly diversified - so we get to take a step back and look at general trends in the marketplace, and where we should invest.

What are the big trends?

On a high level, there are probably two big trends that we see today. The first trend is in the emerging economies, for people to want better connectivity, In all of these economies, given the price of raw materials, people do not install copper and they do not install fibre, because they will be stolen, So people take our equipment and install it in a locked room, and do the same thing the other end, and there you have a wireless link. Wireless gives them security.

In those markets we end up selling our first two product segments - the wireless LAN piece and the WiMax piece - WiMax provides the backhaul and Wi-Fi provides the access.

The other trend we see, in the US and Western European countries, is around security and surveillance. An example is where we will put up video cameras to serve a particular site. Terrorism has been a big factor here.

For this trend, we tend to sell primarily the WiMax equipment because that services the backhaul - and then we'll either sell with wireless LAN equipment if someone wants to provide Internet access, or we'll provide mesh capability for law enforcement, or point-to-point equipment to provide a bigger pipe to carry the traffic back.

The point is, there is a trend towards sales that go across multiple segments. This is important to Proxim because we are a wireless pure-play that has more than just one single segment. So we didn't plan the trend, but the trend is coming to us. We have products across the three segments, wireless LANs, WiMax and point-to-point. Nobody else has the entire portfolio that is also a wireless pure-play. - our biggest competitor is Motorola.

What does this synergy allow you to do, that other people can't?

Wired networks are interoperable. Take a Cisco router and a 3Com switch, and they work together just fine. Wireless networks very quickly tend to single-vendor networks. You're not going to find many deployments that have WiMax Proxim on one end, and WiMax Motorola on the other. It could work in theory, but in practice, an all-Proxim network or an all-Motorola network will work better than a Proxim-Motorola network.

So service providers tend to go with a single vendor (and you also see this on the DSL side) - because there are so many things that can go wrong. So we have the opportunity to go in and provide the network management platform that will manage the APs, the WiMax and the point-to-point links, and they will all work together.

The other thread we see is that 802.11n will start a trend were enterprises will go to all-wireless. Today, the enterprise is probably the one stronghold for wired networks. Everything else is pretty much going wireless - especially outdoors. If enterprises go wireless, you are going to see a trend to networks that are wireless end-to-end. Access, backhaul and distribution are all wireless.

The first question most people ask is whether this all-wireless network will be a secure network. When I was at Aruba, our whole goal was to make sure that the network did not go through the four walls of the building. But in an all-wireless network, your goal is exactly the opposite. Instead of trying to limit how far the network goes, your are trying to make sure it goes as far out as possible.

If you want the network to be omnipresent, then security becomes a very critical network requirement. Security has been talked about a lot in the wireless LAN space, but not talked about so much in the point-to-point space. We have released a new product line called the HS (high security). It is an overlay to our existing product lines. The first one was a WiMax HS product: it's a FIPS 140 Level 2 product. Our goal is to go after banking and the medical industry - areas that will mandate FIPS 140 [FIPS 140 includes 256-bit encryption and other features].

Now that we have that for WiMax, we're going to take the concept and push high security into point-to-point and into the access points, so we'll have a parallel product line in all our segments. 

I understand you plan to push the performance of WiMax. Aren't you limited by the standards?

If people start using video, and if 802.11n starts showing up on the access end, then the backhaul is going to have to get bigger, because the mismatch between access and backhaul will be huge.  If you are going to replace wired with wireless, the experience must be the same.

With WiMax, you get about 20Mbit/s per radio. Our goal is to work on products that will push that frontier substantially. We'll have products that are extensions to the standard.

Our belief is that the WiMax spectrum is going to be dominated by the Tier 1 vendors, because they have the cash to buy the spectrum. That won't stop the Tier 2 and Tier 3 carriers - they will deploy WiMax in the unlicensed space - and right now, WiMax is bigger in the unlicensed space.  Tier 2 and 3 vendors care more about getting a return on their investment than having a piece of paper that says this is certified WiMax.

Without getting into the technology details, we think there is a big market for non-standard WiMax products - whether it is an extension of WiMax or something completely different.  There's a market there, and we're trying to push the limits.

What about the prospects for mobile WiMax?

I think WiMax will succeed - but it won't be as big as some expected.

WiMax was created for a specific problem: fixed point to multipoint backhaul. It is probably the best technology available on the planet today for that problem. We plan to use WiMax for that fixed point-to-multipoint backhaul - both licensed and unlicensed, both standard and non-standard.

But will WiMax become an access technology? So far there is only one SKU [product] for a mobile WiMax client - the Nokia N810. If people are going to start deploying mobile WiMax networks, either someone is going to have to pony up the cash to build networks, so the devices make sense, or someone is going to have to pony up the cash to build the devices so the networks make sense. One of those needs to happen - and neither of them is happening.

There's a conversation about whether access will be by wireless LAN, LTE or Mobile WiMax. From our point of view, we don't care because we are more about the backhaul.

We've not jumped into Mobile WiMax. It's an area we could do - but it would make no sense.  It's a GSM-type deployment, so if you are a Tier 1 carrier, you are going to need someone with resources.

I think it is hard for one company like Intel to carry the load. People are banding together for LTE, just like they did for wireless LAN. Intel is a savvy company. If it turns out to be LTE, they'll just put LTE in all the devices. I don't think Intel cares.

Do you think  we are seeing the first phases of a climb-down at Intel?

Intel is an extremely data-driven company. I used to work at Intel years ago, and Andy Grove drilled it into us that you go where the data takes you. Do you remember the whole Pentium flaw thing?  They said it's only one in however many million Pentiums that is going to fail, and they said that is not a good enough reason to do a recall.

That was mishandled - they missed the whole emotional aspect of it - but the point is, if the data shows LTE is succeeding, they will go with LTE. I don't think they get emotional - that's what makes that company truly great.

LTE might actually work better in mobile devices - the uplink uses less power. WiMax had a theoretical lead to market, but has lost that. And finally, WiMax equipment was supposed to be cheaper because of lower intellectual property costs, but LTE vendors have seen what they have to beat.

Given all that, you've not jumped into Mobile WiMax, and people like Nortel are jumping out of it. Does it actually matter to you whether the industry goes with Mobile WiMax or LTE?

[A very long pause] … No.

If LTE actually beats Mobile WiMax, it might actually be good for you. Mobile WiMax kit can also handle fixed WiMax, so you as a fixed WiMax vendor won't have a lot of people who have succeeded in Mobile WiMax getting into your space.

Yes, That's why I thought so long before answering.

If Mobile WiMax does take off, you might be in a position to get into that space. If it doesn't then fewer people are in a position to get into your space. A double edged sword.

Sometime in the long Proxim saga [Symbol lawsuit, entering WiMax, bankruptcy and sale to Moseley, snatched by Terabeam] the company said it was betting on municipal Wi-Fi. That fitted with Wi-Fi and WiMax backhaul. But municipal Wi-Fi didn't take off. A lot of the businesses you talk about, like surveillance, are just the wreckage of that, aren't they?

In municipal Wi-Fi there were two markets. The first was the Internet access market. That went under because it was a free market - services were supposed to be free. It always went to an RFP, and there was always a digital divide angle to it. That never panned out.

The other market, the public safety market, is still available today, and that is a big market for us. We've always focussed on that - and toned down the mesh aspect.  Video surveillance is huge.

So, free Wi-Fi was a bubble that burst, but there's still plenty going on in that area?

What about your products?

In the next six months, we will have refreshed our entire product line. 

In point-to-point, we recently launched a product in our Lynx range. We're taking the lessons we learnt in the wireless LAN space and applying them to point-to-point.  Carriers are using T1s and E1s and backhauling over wireless, but they say they will want to go with Ethernet a year or two from now, so the new product has T1 and E1 interfaces, and Ethernet interfaces already. So when the carriers switch from T1s and E1s to Ethernet, all you've got to do is unplug it and plug it into the Ethernet.

It also has software selectable radios. You can do 2.4 GHz or 5.8GHz in the same box. That is taken for granted in the wireless LAN world, but we are the only point-to-point product that combines these two features.

In WiMax, we've done high security, and we're also going to have high performance products.

In Wi-Fi , we're going to introduce 802.11n access point products over the next six months.

Many moons ago, Proxim was going to do a Wi-Fi switch. Any plans there?

We do not intend to go down the wireless LAN switching route. There's enough competition there, and it's dominated by one single player - Cisco has 70 percent. The second player has less than ten points of market share.

For us, it's much better to be in markets that are more fragmented. That's true of the WiMax market and especially the point-to-point market.

But you need to keep your access point fresh.

We've got to do 11n - it's part of the DNA, and it doesn't take a lot to do it. There is no alternative to it.  Also, 11n is going to drive a resurgence of the fat AP.

Centralised switching made sense at a  time when the wireless speeds were slow, so you didn't need big switches. Now, you need 80Gig switches, and they cost real money.The other reason for wireless LAN switches was, the standards weren't baked in terms of actual deployment. Encryption and security hadn't been finalised. Companies like Aruba and Airespace said "We don't care which you are going to use - we've got it all in one box."

But now it's six years later. People have decided how they are going to deploy wireless.

You're not in a hurry to do 11n then? Everyone else has it already.

We are in a hurry for 11n - but the total market for 802.11abg far outstrips the 11n market today, and will do so for the rest of the year. After September, shipments of 11n will slow down, because the schools are done [education is leading the 11n charge, and school and installations happen during the summer holidays]. We'll go back to a slow adoption, with testing in the enterprises.

So it is not in our interest to overhang our current product set. We're going to announce 11n the day before we actually ship it.

802.11n is going to be massive, but the next six months is nothing - it's more important to get it right. This isn't a six month window, it's a six year game. If you don't get it right, you're dead.

What about Trapeze as part of Belden? Can it remain a player, and will it keep its contracts with Nortel and 3Com?

The big challenge for Belden is to make sure they keep the talent.  If they can stay on the cutting edge - then yes. But if not, it will become apparent very quickly. Nortel and 3Com have the reach to get the big deals. If they can't get them, they aren't going to stick around.