CWNP Wireless Certification & Wireless Training - What's YOUR Definition of "Adjacent Channel?"

Your Account |

Search:

Free Stuff

What's YOUR Definition of "Adjacent Channel?"

Add to:

Written by Devin Akin

Monday, 07 July 2008

I've been having an interesting discussion with an instructor regarding the term "adjacent channel", and I would like to solicit comments from our constituents regarding this term. In the 2.4 GHz band, channels 1 and 2 are considered "overlapping" by the 802.11 standard, but the question is, "are they adjacent?" According to the standard, they are not. The instructor - a very expererienced one - argues that the most common use of the term "adjacent channel" has it meaning "directly next to" rather than defining it by a given frequency distance. For example, the standard defines channels 1 and 6 in an 802.11g system as adjacent because they are separated by at least 25 MHz. The reason this terminology is so important is because the term "adjacent channel" carries over into "adjacent channel interference." The defining of what adjacent channel interference *is* and *is not* is important.

The standard defines interference between channels 1 and 2 as co-channel interference caused by overlapping channels. Also, inteference between channels 1 and 6 is considered adjacent channel interference. What we have to decide here is whether we want to continue to follow the standard's definition or choose to follow what is currently perceived by some industry professionals (how many, we don't currently know) as different from that of the standard. Below is the definition/chart that The CWNP Program currently uses in its test terminology document:

The CWNP Program has aligned all of its terminology with the 802.11-2007 standard and The Wi-Fi Alliance. Before any significant changes are made, it makes sense to justify such changes through much feedback from our constituency. It's my opinion, that before we would want to make a change to say that "channels 1 and 2 are adjacent" that we would need to:

1) See if there is an industry-standard organization that has documentation that CONSISTENTLY defines and documents "adjacent channel" in a way that opposes the 802.11 standard's definition.

2) See if there are products in the market place, that function in the 2.4 GHz band (the only one where this discussion has meaning), that use the term "adjacent channel" in a way that contradicts the 802.11 standard

3) Show that those products are in-and-of themselves industry standards ("AirMagnet Survey Pro" for example) and have standardized terminology that contradict the 802.11 standard.

I would love to hear back from as many of you as possible on this topic, because definitions are a big deal - especially for test-takers. Thanks!

--------------------------------------------------------

As additional information, here are some good examples of where industry players go directly against the standard:

Craig J. Mathias (Farpoint Group) Whitepaper:

Cisco's WLAN Controller Configuration Guide 5.0 (new)

Neil Diener (of Cisco, former CTO of Cognio) whitepaper

These are well respected people and companies using this terminology in a way inconsistent with the 802.11 standard.

The standard's terminology is perfectly clear. For example, Section 18.4.8.3 says:

"In an adjacent channel (≥ 25 MHz separation as defined by the channel numbering)..."

So should we push harder to educate those that use the terminology in a manner inconsistent with the standard or should we accept an alternative use as 'the defacto standard' instead?

Comments (17) Add Comment

Subscribe to this comment's feed

...
written by Dan C., July 07, 2008

Well here's my two cents for what it's worth:

When I think of 'adjacent channels' regarding an 802.11 system I tend to follow the 802.11 standard. This is only because I have learned all my terminology directly from the CWNA and CWSP certifications.

However, before I forced myself to think the same way as the standard I most definitely considered channels 1 and 2 to be 'adjacent' channels. I think this is the normal way most people would think about channels until they are trying to get certified and force themselves to follow a more standardized terminology.

So I guess in summary:

My personal definition - Channel 1 and 2 are adjacent (this seems more logical to me)
When I am working - I follow the standard

...
written by Glenn De Haes, July 08, 2008

It would have helped a lot if the standards would have been comparable from the start. Both 11a and 11b started at the same time but still, 11b - and 11g - talk about all available channels non-overlapping channels whereas they have chosen with 11a to only talk about non-overlapping channels. It is then clear that with 11a you can only have adjacent channels and adjacent channel interefence. If we would only have channels 1,6 and 11 to choose from in 2.4GHz, there would be no co-channel interference. Up to today, I still come across people who use their system on channels 1, 2 and 3 for instance as they think they are different, i.e non-overlapping. I use personnally in our courses adjacent channels for non-overlapping channels and co-channel (interfence) for overlapping channels.
I feel this is less a problem of this naming convention but more the way we handle channels and will be threating them in the future. 11n will also only give non-overlapping channels and thus adjacent channels...
I feel it is time that we use 11m to get rid of these inconsequences and that we start with a solid base that is mutual to all standards.

...
written by Howard, July 08, 2008

Until a few weeks ago, I too was confused by the term "adjacent channel". Perhaps because I had not needed to read the specification that closely before. Afterall, doesn't adjacent mean "next-to"? All that changed when I started using an Anritsu WLAN tester, and had to really get into the guts of the specs to create a comprehensive set of test scripts. Like many endeavors, poorly chosen terminology, can cause everlasting confusion. I think the only way around it would be to change the spec (fat chance) or make a greater effort to educate people.

...
written by Christopher, July 08, 2008

It has been my experience that most people correctly use the terms 'overlapping' and 'adjacent' in ways consistent with the standard. My feeling is that the terms are correctly defined and folks in the industry (at least those that I know) use them in that way. It is nice to see this topic highlighted to call attention to an important topic not frequently discussed.

...
written by Brett, July 08, 2008

Dan C. hit the nail on the head on this one for me.

Overall I don't think the CWNP program should alter what it considers adjacent channel. Follow the standard, not the "industry".

...
written by Solairaj, July 12, 2008

As per my Study,
Channel-1 & Channel-2 are Co-Channels.
Channel-1 & Channel-6 are adjacent channels.

...
written by BryanH CWNE, July 12, 2008

The standard may use the term one way, but the industry has been and will continue to call adjacent channels those that are directly next to each other. Just like the standard has called 802.11a, b, and g other things but the industry keeps using a/b/g. The same is true for PoE, Standard 802.3-2005 clause 33....industry 802.3af.
To quote from the movie "The Man Who Shot Liberty Valence", when the legend is bigger than the truth...you print the legend. The industry uses terms incorrectly every day like hacker when they should be saying cracker. If the program wants to sort this out that is great. Let's just all pick one way or the other and remove the confusion before it is too late. A quick fix would be to stick strictly to what the standard says in all of the teaching and documentation. This may take a lot of re-education though.

...
written by David Boyd, July 13, 2008

This is an interesting issue. How a channel is determined to be adjacent can be measured in multiple ways. First there is the channel number 1,2,3... and there is the primary frequency 2412, 2417,2422... These two options are for all intents and purposes the same. Now comes the the frequency band that they use. If we look at channels 1 and 2, they can not be adjacent because they're occupying the same 17MHz frequency space . The definition for adjacent is 'having a common end point or immediately preceding or following'. We don't meet that condition until channel 6. So, for simplicity, I'm guessing that the industry players listed above are using the channel numbers for their end-users - dumbing down for the least common denominator, whereas the standard was written for technical personnel.

Though I've been using and supporting wireless for years I am fairly new to the technical side. I think that the standard continues to work and that vendors need to be 'educated' and follow the standard.

...
written by BryanH CWNE, July 17, 2008

Upon re-reading the standard and the 802.11 Designers Handbook 2nd ed. page 258, I have found three types of neighboring channels listed: 1. Co-Channel (on the same channel with overlapping cells) 2. Adjcent Overlapping channels (channels that have overlapping side bands) and 3. Non-Overlapping Adjcent Channels ( channels that are close to each other but do not have overlapping side bands i.e. 1,6, and 11).

I find that chanels that overlap and those that do not overlap can both be called adjcent. The question is are they interfering with each other or not. I would be happy calling the channels Adjcent Overlapping and Adjcent Non-Overlapping if we can clear this up and have the industry follow our lead.

...
written by Devin Akin, July 17, 2008

Hi Bryan,

The standard doesn't define it this way, but apparently Bob and Al did. I like your idea alot. I think this is a very good middle ground. :) Excellent work my friend!

Devinator

...
written by Michael Finneran, July 21, 2008

Channels 1 and 6 are "adjacent" by every definition of "adjacent" provided above. What's confusing everyone is what the standard means by "Channel 1" and "Channel 6". The original 802.11 standards defined channels with center-frequency separation of 5 MHz. Unfortunately, the original 802.11 and 802.11b radio links require a channel bandwidth of 22 MHz, which takes 5 of those 5 MHz channels. So the signal on "Channel 1" is actually occupying channels 1 through 5! The span is 2.4095 to 2.4345 GHz (Center Frequency 2.422 GHz, the center frequency of Channel 3). What they call "Channel 6" is actually channels 6 through 10, 2.4345 to 2.4595 GHz (Center Frequency 2.447 GHz, the center frequency of Channel 8). The top end of "Channel 1" (i.e. Channels 1-5) butts up against the low-end of "Channel 6" (i.e. Channels 6 - 10). That's"adjacent". It probably would have been clearer if they called them Channels 3, 8, and 13, but that's life.

This stuff isn't too tough if you understand the basics.

...
written by John Gleichweit, N6FOG, MSCE, CCNA, July 21, 2008

Mr Finneran is almost correct. The center frequency of Channel 1 is 2412, with 11 MHz on either side of that. This covers from 2401 to 2423. Channel 6 is from 2426 to 2448 with a center freq of 2437. This gives a gap of 3MHz between non-overlapping adjacent channels. The center frequency of Channel 3 is 1MHz below the top end of Channel 1, giving an overlap of 10MHz.

In the LMR world, "Adjacent Channels" are those that are next to each other, and do not have any overlapping in their modulation bandwidths. A frequency of 450.000 and 450.025 are adjacent channels when used with "16k0F3E" modulation; 450.000 and 450.0125 are overlapping when using the same modulation scheme, but are adjacent when using "8k0F3E" modulation, which take up only half the bandwidth.

...
written by Criss Hyde, July 22, 2008

Popular ERP channel identifiers include 1, 6, 11 -- five units apart.

Popular OFDM channel identifiers include 36, 40, 44 -- four units apart.

In both cases the "unit" is 5 MHz -- a distance between channel centers.

Adjacent familiar ERP channel centers are 25 MHz apart.
Adjacent familiar OFDM channel centers are 20 MHz apart.

What happens to ERP channels 2, 3, 4, 5?
What happens to OFDM channels 37, 38, and 39?

The extra ERP channel identifiers are used for standard channel widths but skewed in such a way as to maximize use of an overall narrower frequency space than that used by OFDM.

The extra OFDM channel identifiers are used for different channel widths in an overall wider frequency space than that used by ERP.

We have never been told to think of OFDM "channels" 36 and 37 as being adjacent. But in fact they are when using the Japanese 5 MHz channel centers, and "channels" 36 and 38 are adjacent when using the Japanese 10 MHz channel centers.

In both ERP and standard OFDM, "adjacent channel centers" are just far enough apart for electromagnetic comfort -- four 5 MHz units for OFDM and five 5 MHz units for ERP.

The standard is a beautiful thing.

I hope this helps. Thanks. /criss

...
written by chike Okoli, July 25, 2008

in the interest of consistency that will thus play down intellectuall rascality; the standard concepts of defining adjacent channels must be upheld. there is a standard as set in 802.11 wireless standard, thus i think we should pursue consistency. all the same, it was a good job

...
written by RNicholas, July 26, 2008

This appears not to be a problem of technical definitions, but one of semantics, as Dan C pointed out in the beginning.

The standard seems clear. In cases like these, the inconcrete nature of the discussion--where everything must describe because it cannot be witnessed or observed directly--I find it helpful to describe the situation using analogies to things more concrete. To a student I might say, for example, that ajacent rooms in a house do not overlap; or two cars in a garage are adjacent, but do not overlap. This would drive home the semantic difference (if not the technical one). The technical details are extremely important, but the two terms must be separate and distinct in the student's mind before a deatailed anaylsis may begin.

Of course, as most analogies break down at some point, this one does also: cars CANNOT overlap (at least not without a call to the insurance company).

...
written by chike okoli, July 29, 2008

Its true that it aint a case of technicallity. i think i will agree with u on that. we should discuss. i thought of that a shot while ago. may be we should put heads and find a non arbitrary school of thought in respect to adjact channel. i will come up with a carefull argument

...
written by CJ, August 26, 2008

The easiest way to describe this to anyone new into wireless is to draw it out. Some CWNA training guides will show a visual picture of the wireless spectrum with circles depicting each of the channels.

With it done this way it is very clear that channel 6 is "adjacent" to both 1 and 11 and also which channels overlap.

Being as how these terms are involved in the tests for certification this way, I would never advise even confusing the students with any other manner of thinking. A clear view of this will lead to professionals that understand the reasons why we cant have channel ovelap and leave much less of a mess for me to work around in the future.

Write comment