« Simplify. Secure. Roam. Repeat. | Main | Anal About Antennas »

June 26, 2011

A Polarizing Problem


Bunch-of-antennas-on-AP "iEverything" (wireless only) devices that don’t have an Ethernet connection will soon outnumber those that do. Unlike laptops, these devices enable truly mobile network computing.

This is a profound and fundamental shift that changes everything for computer networks - placing a BIG spotlight on the imperative to improve Wi-Fi communications.


We'll say this a couple times: with wireless, good speed is ALL about good signal. While bigger, stronger antennas have a significant impact on better Wi-Fi performance, other things related to antennas matter too. Arguably just as important to bigness is antenna orientation and the polarity of RF signals. 

An antenna provides three things to a radio: gain, direction and polarity. Gain is the amount of increase in energy that an antenna adds to the RF signal. Direction refers to the shape of the transmission, which describes the coverage area. Polarity is the orientation of the electric field (transmission) from the antenna. 

Understanding Polarity

Wave your hand up and down like you do when you put your hand out of the car window. That up and down movement of waves is called vertical polarity (VPOL). The opposite of vertical polarity is horizontal polarity (HPOL) which is like a snake slithering in the desert; waves that move from side to side. Polarization For an antenna to transmit waves vertically the antenna will be vertical and if an antenna is horizontal, the polarity is also horizontal. 

Here’s the scary part. Horizontal and vertical signals are so different that they aren’t compatible. If you have a perfectly vertical signal hit a perfectly horizontal antenna, that horizontal antenna doesn’t hear anything. Many times, in indoor environments, signals bounce off walls and things (called multipath by geeks) This can change the polarity of signals.


That said, in a linearly polarized system, a misalignment of polarization of 45 degrees can degrade the signal up to 3dB and if misaligned and 90 degrees the attenuation can be more than 20dB. That just sucks. Don't believe us?  Read some "real world" comments from Joe McBreen, who runs IT for a huge school district, St. Vrain Valley School District, in Colorado.


Which Way is Up?

Polarization-diagram Until fairly recently most wireless devices didn't move much. Laptops typically sit in one spot with the screen up (the antennas are usually behind the screen). And, until recently, the orientation of cellular phones has also been straight up and down because there were only used for talking, not computing per se.

Today, things couldn't be more different. Devices such as iPhones and iPads are so versatile they are moved around in almost every imaginable position. In the Wi-Fi world this is like someone on your roof moving your old free-to-air VHS antenna around with your picture fading in and out. Basically, every time you change the orientation of the device you are also changing the orientation (read: polarity) of the antenna of that device - and most of today's Wi-Fi APs can't do anything to deal with this.  

The Big Rub

Now-what Nearly every (non-Ruckus) Wi-Fi access point sold today utilizes omni-directional, "dipole" antennas that are vertically polarized. These have been accepted and considered “normal” for quite some time and for good reason. Prior to the mobile Internet boom, most devices were also vertically polarized and everything worked fine. But here's the rub. Any time your client device isn’t in the perfect orientation, the signal from the omni-directional AP is diminished which results in range, throughput and reliability all suffering. 

Adapt or Die

Contrary to an omni-directional antenna, adaptive antenna arrays (found in Ruckus APs) are designed with both horizontal and vertical antennas. This is much more difficult to do than it may seem. Some of you have already thought, “Well, I’ll just change the orientation of a few omni antennas and fix that polarity problem!” Unfortunately, that won’t work. It gets pretty complicated but what you would be doing is changing the coverage pattern of those antenna(s) but not others. This would seriously mess up your throughput and range of an 802.11n system. 

Giving It to You Straight (or Sideways)

Anytime a Ruckus AP transmits, it matches the polarity of the receiving device. It can use one or both polarities to ensure that the receiver hears as well as is possible. Conversely, every time a Ruckus AP hears a transmission from a client device we adjust how we will talk to them. It’s like playing Marco Polo in the pool. We know where people are just by listening; we don’t need to see them. 

Just Listen!

MRC Since APs can’t control the orientation of the client antenna, it's important to listen on all polarities. This is done with Polarization Diversity Maximal Ratio Combining (PD-MRC). MRC (click on diagram) is an 802.11n standard way of being able to combine multiple multipath transmissions into one good signal. PD-MRC is a way to do the same thing, except with the ability to combine multiple signals of varying polarities. This allows a Ruckus AP to accurately listen better to a client device, no matter how it is oriented. 

From the onset, Ruckus has been (and continues to be) focused on ways to solve the major problem that plagues Wi-Fi networks: optimizing Wi-Fi signaling. To really improve Wi-Fi performance you need to actually improve how Wi-Fi signals are transmitted and received. This will have (by far) the single biggest impact on Wi-Fi performance. At the end of the day, no matter what Wi-Fi equipment suppliers tell you, the more signal you can deliver to your clients, the faster they will be able to send and receive information. 

Ironically, the vast majority of Wi-Fi vendors focus on how to improve Wi-Fi AFTER clients are connected and doing nothing to really make Wi-Fi better. So are they really Wi-Fi vendors? Hmmmmm.

Comments

Devinator

Some thoughts:

Good blog.  Kudos.  Liked it.

Airtime scheduling matters for maximizing a cell's throughput when client data rate connections are mixed.  Just an additional thought.

Changing polarization and shaping the pattern on reception takes being predictive, and at high-density, doing both with extreme accuracy seems unlikely, regardless of marketing spew.  Doing otherwise means using multiple sets of omni antennas (one set of 3 per polarization) unless I'm out in left field somewhere, which I submit is possible.  That then means better reception due to polarization, but not directionality, which still could be very good I'll give you.

Being so unilaterally focused on antenna technology, unless you have the funds of Cisco, means that you have to give up something somewhere else, and I submit that's system intelligence in Ruckus's case.  Not a bad thing all told - it is what it is - every vendor picks their markets and focuses if they want to be successful.

Ruckus has good antenna tech - granted.  I'd never say different, but...if you're a one-trick pony, you pigeon-hole yourself unless you find alot of venues for your trick.

Something to think about...

Marcus Burton

Pigeon-hole?

In the spirit of argument with the Devinator, I completely disagree with the take that Ruckus' antenna focus negatively impacts their market opportunities. Ruckus has one of the least pigeon-holed Wi-Fi solutions out there today. What other WLAN vendor has equivalent outdoor, carrier, and indoor enterprise presence?

I see their market diversity as an extremely good thing and attribute it largely to RF performance.

Post a comment

Comments are moderated, and will not appear on this weblog until the author has approved them.

If you have a TypeKey or TypePad account, please Sign In.