<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	>
<channel>
	<title>Comments on: A look at 802.11a, b, and g throughput with short preambles</title>
	<atom:link href="http://blog.matthewgast.com/2007/01/27/80211-short-preamble-throughput/feed/" rel="self" type="application/rss+xml" />
	<link>http://blog.matthewgast.com/2007/01/27/80211-short-preamble-throughput/</link>
	<description>A former physicist tries to make sense of technology</description>
	<pubDate>Wed, 19 Nov 2008 11:54:51 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.5.1</generator>
		<item>
		<title>By: matthew</title>
		<link>http://blog.matthewgast.com/2007/01/27/80211-short-preamble-throughput/#comment-7679</link>
		<dc:creator>matthew</dc:creator>
		<pubDate>Thu, 05 Jul 2007 17:28:59 +0000</pubDate>
		<guid isPermaLink="false">http://blog.matthewgast.com/2007/01/27/80211-short-preamble-throughput/#comment-7679</guid>
		<description>Rishabh,

802.11g doesn't tell you which preamble to use.

The most relevant piece of the standard (802.11-2007 clause 9.13; it was copied unchanged from clause 9.10 in 802.11g) says: "Protection mechanisms frames shall be sent using one of the mandatory Clause 15 or Clause 18 rates and using one of the mandatory Clause 15 or Clause 18 waveforms, so all STAs in the BSA will know the duration of the exchange even if they cannot detect the ERP-OFDM signals using their CCA function."

In English, what that means is that you need to use either the original 1 Mbps and 2 Mbps data rates with direct sequence modulations (clause 15), or the 5.5 Mbps and 11 Mbps 802.11b higher rate direct sequence.  Either is acceptable, according to the standard.

When I wrote the original article in 2003, it was somewhat common to find pre-11b gear that would only understand long preamble.  If you wanted protection to work, you needed to use long preamble.  In the past four years, long preamble-only equipment has essentially become extinct, so it is OK in practice to use the short preamble.  You may, very occasionally, run into an ancient pre-11b piece of equipment that needs long preamble, but it's going to be a vanishingly small number of cases.

In 2003, the company I work for shipped APs with the default as long preamble.  We made the decision for compatibility reasons, since there was a fair amount of pre-11b equipment available.  As it became less common, we changed the default to short preamble, in large part for the performance reasons discussed in this post.</description>
		<content:encoded><![CDATA[<p>Rishabh,</p>
<p>802.11g doesn&#8217;t tell you which preamble to use.</p>
<p>The most relevant piece of the standard (802.11-2007 clause 9.13; it was copied unchanged from clause 9.10 in 802.11g) says: &#8220;Protection mechanisms frames shall be sent using one of the mandatory Clause 15 or Clause 18 rates and using one of the mandatory Clause 15 or Clause 18 waveforms, so all STAs in the BSA will know the duration of the exchange even if they cannot detect the ERP-OFDM signals using their CCA function.&#8221;</p>
<p>In English, what that means is that you need to use either the original 1 Mbps and 2 Mbps data rates with direct sequence modulations (clause 15), or the 5.5 Mbps and 11 Mbps 802.11b higher rate direct sequence.  Either is acceptable, according to the standard.</p>
<p>When I wrote the original article in 2003, it was somewhat common to find pre-11b gear that would only understand long preamble.  If you wanted protection to work, you needed to use long preamble.  In the past four years, long preamble-only equipment has essentially become extinct, so it is OK in practice to use the short preamble.  You may, very occasionally, run into an ancient pre-11b piece of equipment that needs long preamble, but it&#8217;s going to be a vanishingly small number of cases.</p>
<p>In 2003, the company I work for shipped APs with the default as long preamble.  We made the decision for compatibility reasons, since there was a fair amount of pre-11b equipment available.  As it became less common, we changed the default to short preamble, in large part for the performance reasons discussed in this post.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Rishabh</title>
		<link>http://blog.matthewgast.com/2007/01/27/80211-short-preamble-throughput/#comment-7661</link>
		<dc:creator>Rishabh</dc:creator>
		<pubDate>Thu, 05 Jul 2007 11:50:58 +0000</pubDate>
		<guid isPermaLink="false">http://blog.matthewgast.com/2007/01/27/80211-short-preamble-throughput/#comment-7661</guid>
		<description>In your previous article, you mention that frames involved in protection operation (self-CTS or RTS/CTS) would compulsorily use the long preamble mode, which also gets reflected in your calculations as 192us is used.
Article: http://www.oreillynet.com/pub/a/wireless/2003/08/08/wireless_throughput.html?page=2
Under Protection 1: CTS-to-self, you write
Protection dramatically reduces the maximum theoretical throughput because the additional CTS transmission is required with its "long" 802.11b headers.

However, in the above blog, you include calculations with short preamble also. I am confused as to whether we can use both preamble modes of 11b in protection mode of mixed 11b/g operation ? Please guide in this regard, also kindly tell what the 802.11g Std. says .

Thanks in advance !!</description>
		<content:encoded><![CDATA[<p>In your previous article, you mention that frames involved in protection operation (self-CTS or RTS/CTS) would compulsorily use the long preamble mode, which also gets reflected in your calculations as 192us is used.<br />
Article: <a href="http://www.oreillynet.com/pub/a/wireless/2003/08/08/wireless_throughput.html?page=2" rel="nofollow">http://www.oreillynet.com/pub/a/wireless/2003/08/08/wireless_throughput.html?page=2</a><br />
Under Protection 1: CTS-to-self, you write<br />
Protection dramatically reduces the maximum theoretical throughput because the additional CTS transmission is required with its &#8220;long&#8221; 802.11b headers.</p>
<p>However, in the above blog, you include calculations with short preamble also. I am confused as to whether we can use both preamble modes of 11b in protection mode of mixed 11b/g operation ? Please guide in this regard, also kindly tell what the 802.11g Std. says .</p>
<p>Thanks in advance !!</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: matthew</title>
		<link>http://blog.matthewgast.com/2007/01/27/80211-short-preamble-throughput/#comment-283</link>
		<dc:creator>matthew</dc:creator>
		<pubDate>Mon, 29 Jan 2007 21:31:08 +0000</pubDate>
		<guid isPermaLink="false">http://blog.matthewgast.com/2007/01/27/80211-short-preamble-throughput/#comment-283</guid>
		<description>The preamble does not enter into either of those cases.  802.11a doesn't have backwards compatibilty to worry about, since it was the first 802.11 PHY developed for the 5 GHz band.  What you refer to as the "native" mode in 802.11g is called ERP-OFDM by the standard.  ERP-OFDM looks almost exactly like 802.11a.  (In fact, large parts of the 802.11g specification refer to the clause for 802.11a.)  The only time that you need to worry about the long versus short preamble is on 802.11 frames transmitted at 1, 2, 5.5, or 11 Mbps.

This may be a bit repetitive to point out, but every 802.11g network that I have come across in the real world has been using protection.  There are still a lot of 802.11b-only devices out there.</description>
		<content:encoded><![CDATA[<p>The preamble does not enter into either of those cases.  802.11a doesn&#8217;t have backwards compatibilty to worry about, since it was the first 802.11 PHY developed for the 5 GHz band.  What you refer to as the &#8220;native&#8221; mode in 802.11g is called ERP-OFDM by the standard.  ERP-OFDM looks almost exactly like 802.11a.  (In fact, large parts of the 802.11g specification refer to the clause for 802.11a.)  The only time that you need to worry about the long versus short preamble is on 802.11 frames transmitted at 1, 2, 5.5, or 11 Mbps.</p>
<p>This may be a bit repetitive to point out, but every 802.11g network that I have come across in the real world has been using protection.  There are still a lot of 802.11b-only devices out there.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Chris</title>
		<link>http://blog.matthewgast.com/2007/01/27/80211-short-preamble-throughput/#comment-282</link>
		<dc:creator>Chris</dc:creator>
		<pubDate>Mon, 29 Jan 2007 20:39:27 +0000</pubDate>
		<guid isPermaLink="false">http://blog.matthewgast.com/2007/01/27/80211-short-preamble-throughput/#comment-282</guid>
		<description>So the preamble issue doesn't apply to Wireless G ("native" mode for lack of a better description) or Wireless A?</description>
		<content:encoded><![CDATA[<p>So the preamble issue doesn&#8217;t apply to Wireless G (&#8221;native&#8221; mode for lack of a better description) or Wireless A?</p>
]]></content:encoded>
	</item>
</channel>
</rss>
