Nigel Ballard on mesh, WiBro and everything else

I had occasion to look at the card stats for all the hotels we manage at Matrix Networks. Below are the percentages of the top four client cards used across ten US states and in almost 30,000 guest rooms:I had occasion to look at the card stats for all the hotels we manage at Matrix Networks. Below are the percentages of the top four client cards used across ten US states and in almost 30,000 guest rooms:

Intel Corporation 18.58%
Agere Systems 11.32%
Gemtek Technology Co., Ltd. 9.15%
Apple Computer, Inc. 6.33%

What we learn from this is that Intel’s $310 Million spent on getting Centrino onto as many vendors motherboards as possible, clearly worked. The old adage “If you throw enough money at a problem” clearly applies here. Well done the folks at Intel Marketing.

My day job has become much easier after deciding to take all our hot spots, hotels, marina’s, restaurants and convention centers over to the Nomadix hospitality gateway. With so many clients running IPsec VPN’s, you really need a solution that can seamlessly hand out routable IP’s as well as do all the other stuff effortlessly 24/7. Those Taiwanese knock-offs fall so far short of the Nomadix in this and other respects. Learn more: http://www.nomadix.com.

We now have the remote ability from our NOC to automatically throttle clients, authenticate passcodes, blacklist spammers and machines with viruses, as well as provide a nice monthly report of who used the network, how much traffic passed across the T-1, which MAC addresses were blocked for doing bad things, how many emails were SMTP redirected, etc. Such control is desirable in a Hilton, but expand the user base to a citywide municipal network and this level of granularity is essential.

My dog Bowlie ate my phone last week; to add insult to injury, he also ate the replacement phone that the UPS man left at my front door five days later. The nice people at Nokia must be adding peanut butter essence to those express-on covers!

Such is my desire to get the Motorola MPx 300 SmartPhone that I’m holding off getting locked into a year long contract with a new phone that has few of the features I really want.

So what do you REALLY want in a phone? I’m glad you asked. I want a phone that incorporates 802.11b as standard, no plug-in card, no workaround hack. A flip would be nice, a clamshell would be nicer, but a combined flip/clamshell would be nicest.

What else . . . a 1.3m pixel camera for the single image I’d probably snap a month. Bluetooth for wireless connectivity to a headset is desirable. I need a web browser to negotiate all those hot spot login screens. Full integration to my Microsoft Office desktop is an essential requirement as is quad-band GSM operation for true global roaming.

And short of a device to pick the wax out of my ears, the eagerly awaited Motorola MPx 300 has all this and much more.

It passed FCC approval in December, so it can’t be too far off. You can buy a seriously overpriced and buggy engineering unit off eBay for a King’s ransom today, or wait till Feb/March when I’m reliably informed the MPx 300 will officially appear via one or more US carriers.

At which point I’m expecting to see the user base at hot spots start to shift a little as coffee shops will no longer be only for those who are willing to lug a seven pound behemoth along just on the off chance that there is Wi-Fi up and running. With a true 802.11 client on your belt equipped with its color browser and a usable keyboard, I foresee more and more people logging in and surfing from more diverse locations.

On my personal web site at http://www.joejava.com I have a sneaky tracker that shows me who came to visit, where they came from and what search words they used to find me. Six months ago, the top search word was “WiMAX”. Interestingly enough it is now “mesh”.

Many mesh solutions today are based loosely upon early DARPA approaches designed to solve speedy battlefield deployments. DARPA solutions focused on small data loads with a high degree of jamming resistance. Consumers in the mesh arena are typically less mobile, are not being barrage jammed but do have substantially larger data amounts to transfer than is typical over a battle-net.

A number of vendors are pushing single-radio mesh units. If the single radio approach is based upon the older 802.11b standard, you are unlikely to have more than 5.1Mbps of throughput initially. You then need to slice that in half to manage your backhaul as your single radio is pulling double duty. Add a few hops and you aren’t left with terribly impressive throughput capabilities.

Trying more than three mesh-to-mesh hops is suicidal on a single radio system, so you need to backhaul it off the mesh. I’m not a great fan of using 802.11a for this as although the silicon is cheap, it was never designed for this purpose. Better to pull your backhaul out of the mesh box directly into a more it-for-purpose radio such as the Motorola Canopy products. If you can provide PoE onto your backhaul connector, then your utility pole installation becomes that much neater with less points of failure.

Motorola recently released 900 MHz Canopy devices. 900MHz has one huge advantage and three notable disadvantages, so choose your backhaul bands with caution.

Advantage: 900MHz will fire a signal right through wet leaves, over hills, through buildings and is as close to the nirvana of municipal frequency bands as you can get, namely 700MHz.

Disadvantage 1: If you are anywhere near paging towers, they will obliterate your signal. A partial fix is to put in 902-928MHz band pass (notch) filtering, but it doesn’t always work.

Disadvantage 2: The Canopy 900Mhz base maxes out at 2.3Mb (latest 6.1 firmware), and that’s not per client, that’s the whole base station. So, if you’re thinking of putting up one antenna and then competing head on with your local phone companies DSL offering, think again as splitting 2.2Mb amongst 50 users is just going to give your local DSL and cable Internet providers something to chuckle about.

Disadvantage 3: The 900MHz CPEs cost more than their 2 and 5GHz cousins. Not sure why. I always thought the higher you go in the RF spectrum, the more precise component selection and placement had to be. Bueller Bueller?

I’ve recently done some consultancy for a city going the mesh route, additionally I’m one of the directors of Portland Oregon’s Citywide wireless steering Committee, and let me tell you, that whole mesh subject? She’s not going away anytime soon.

Despite the lack of throughput and the problems with running VoIP over it, mesh is the choice of America’s municipalities as the deployment time is short and can be realized very quickly without fifty planning and zoning hearings. Also the cost is easy to calculate and looks way more attractive on paper than either trenching 100 miles of city streets or installing access points and associated backhaul and power on every third building.

Utility poles are typically controlled by the city, leeching power off them is easy and quick, and the 30 feet of the ground that mesh likes, is easy achieved.

OK, enough of this mesh stuff already – what of WiMAX?

A subject close to my heart is WiMAX and I’m rightly accused of being a big WiMAX evangelist. Some of the media gave me a hard time after my “WiMAX, Why Not?” ISPCON keynote in San Jose.

A personal crusade is providing equal access to the Google interface to all. The commercial vendors in the US charge so much for individual domicile net access that wireless has to be the savior here. Wi-Fi doesn’t scale too well, but WiMAX does, since after all, is exactly what it was designed for.

I’ve witnessed 802.16-2004 equipment in trials and having seen what is really being achieved in terms of concurrent clients served and throughputs achieved I stand by my assertion that WiMAX is going to shake up the municipal access model like nothing has before.

Having done my time as a technology journalist, I know they need to write something to put food on the table. If they can’t hold it, touch it and feel it, then it is typical to take all the hype that surrounds a new technology and decry it – heck, it fills the column, right? Have faith brothers and sisters! WiMAX delivers and will change things forever and for the better.

Broadband, what exactly does that word mean to you? If you live in the US and have some money then it probably means having access to nice downloads speeds in the range of 1.5Mb for DSL or a 3-6Mb for Cable, but with lackluster upload speeds coming in at around $35 to $45 a month. If you live in Japan or Korea and you are reading this then you are rightly feeling suitably smug as broadband to you probably means $15 for 50Mb speeds, now that’s broadband! It appears these countries have bridged the digital divide with affordable access, so why can’t we?

Korea has a much more aggressive approach to broadband, especially the portable variety, with mobile phone handsets that are usually about 2 years more advanced than anything we can get stateside. Their data speeds are higher and consumers are pushing for more. They’ve developed WiBro and plan to deploy it aggressively. According to the Korea Information Strategy Development Institute (KISDI), the WiBro service is predicted to attract as many as 9.3 million subscribers by 2011.

WiBro was first conceived as a Korean technology standard called HPi (for high-speed Portable Internet). However, as a Korean standard rivaling WiMAX, it was a proprietary standard. As a result, Korea brought an end to HPi earlier this year, agreeing to use the 802.16e standard for WiBro.

WiBro, formerly known as 2.3GHz portable Internet, enables mobile users to remain connected to the Internet at the speed of current fixed-line broadband. The upcoming service promises a downlink transmission speed of around 1 Mbps with mobile reception at up to 60 kilometers per hour.

I recently bought a Canary Wireless Digital Hotspotter. And while it is actually fatter than it appears and lacks a backlit LCD, apart from that, this is a really superb piece of equipment. Unless you are comfortable whipping out your ThinkPad every time you think there might be an open access point (I know I’m not), then this is $49 well spent.

It identifies the presence of Wi-Fi B/G, tells you the full SSID, the channel, signal strength and whether it is encrypted or not. It even shows Access Points with their SSID broadcast disabled by displaying “cloaked”. It typically takes two seconds to present you this information; can you do that with your laptop as quickly? For more information go: http://www.canarywireless.com/.

And what about MIMO, did you notice that normally uninspiring Belkin released a pre-certification radio and client card ahead of the pack? Their unit is nice and impresses people in trials. Now everyone is following suit at speed. This Pre-N stuff only really flies as advertised if you use the matching Pre-N client cards it should be noted.

Should you Wi-Fi your entire city in Pre-N equipment today? Only if the CEO of the hardware company gives you a written undertaking that if the existing hardware cannot be firmware upgraded to 802.11n when the IEEE finalizes the specification, then they’ll swap out all your hardware for free.

I keep getting asked to recommend professional wireless survey tools, no, not Netstumbler, but dedicated RF mapping programs. I have one I use and there is this Ekahau product I’d like to trial. If I get hold of a copy, my next missive will incorporate a thorough rundown of what’s available, what they can do and why they might make a good investment for anyone who wants to take anything from a small school to an entire city wireless.

You can contact me at nigel[at]joejava.com or via http://www.joejava.com.

Comments

  1. Excellent Artical, just wondered…since ‘Symbol’ have gone all ‘Moto’, will you be looking at:
    Motorola LAN & MESH Planner
    Symbol RF Manager
    v
    AirMagnet Survey Pro & Survey Planner