Forum

Someone asked the following question up on the CWDP Study Forum today. Thought I'd put it up here, along with my reply, as it may prove be of interest:

Question:

"I am trying to understand how the client and the AP choose their transmit power.

On the 1st paragraph it states that a client transmits a frame at 54mbps with transmit power of 20mW. Then if it sends another frame at 36Mbps it has to change its transmit power to 50mW.

Do the APs behave the same way? I searched the data sheets of some Cisco APs and i cant find a table similar to the table 2.5 of p 45. I can only find which are the transmit power levels it can support. There is no connection with data rates and how it can choose the appropriate power level.

Can somebody explain me how does this work?"

Reply:

Firstly, it is very important to realise that this chapter deals mainly with client devices. The two tables presented ( Tables 2.4 and 2.5 ) relate to client devices.

We?ll get started with an analogy.

Imagine that you are at a political rally with an elderly relative. You have excellent hearing, but he has is a bit hard of hearing. You are standing at the back of the crowd, and the politician begins to speak. The crowd is well behaved at this point, and nobody is talking. You listen to what the politician is saying. Your relative looks puzzled. You ask him why. He tells you that he cannot hear the politician. He does not mean that he actually cannot hear anything coming from the man?s mouth, it simply means that although he can hear ?something?, he cannot make sense of the audio energy that is hitting his ears. You tell him that you?ll both move through the quiet crowd until he can hear better. After moving a few dozen yards, he tells you that he can hear the politician well. You, meanwhile, can hear the politician much better.

The politician keeps on speaking. But he says some things that the crowd do not like. They start to get noisy. Your relative once more tells you that he cannot understand what the politician is saying. You now move further towards the politician until your relative says once that he can hear him better now.

The politician then says something about massive tax increases. The crowd is an uproar. You now need to move much further toward the politician in order for your relative to hear properly. In fact, even you, with good hearing are having a hard time understanding him.

As the crowd gets noisier and noiser, you have to move your relative closer and closer towards the politician.

People in the crowd have started shouting questions to the politician. Some are shaking their heads at the back of the crowd, as they cannot hear the politician. He takes a megaphone and uses that to talk to the crowd. His voice can now carry much farther. The crowd has now gotten much larger. It?s diameter has increased. Another relative has joined the crowd. He has good hearing, but is way at the back. He tries to shout a question to the politician, but the politician shakes his head. He can?t hear your relative.
Your relative starts shouting louder. Although your relative can hear the politician fine, the politician cannot hear him.

Your relative cups his hands in front of his mouth and yells as loud as he can. Although it looks like the politician can hear him, he cannot understand what he is saying. Your relative?s voice is distorted. He has reached the limit of what his vocal chords can do.

We will analyze some of this analogy ( and like all analogies, it can only be carried so far ) in the next post.
Dave
Receive sensitivity refers to the ability of a Wi-Fi receiver to properly interpret a signal at a particular data rate with a normal noise environment ( i.e. without interference etc ). Receive sensistivity is defined in terms of dBm for a particular data rate. This information can usually be obtained from a manufacturer?s data sheet. Looking at Table 2.4 , the specification information for this particular radio ( under the 802.11g column ) says that if we receive a signal at a level of ? 71 dBm for example, that we would be able to demodulate ( some literature uses the term ?decode?, but that can cause a lot of confusion, as the term ?decode? can cover a range of things ) that signal up to a maximum of 54 Mbps.

It is important to note that these specifications assume a normal noise environment without any interference. The receiver is not just concerned with signal level ( RSSI ), but also with SNR ( Signal to Noise Ratio ).

If however, we received the signal at a level of ? 75 dBm, we would only be able to demodulate the signal up to a maximum of 48 Mbps. In reality, it is more complex than this, as we really mean ?able to demodulate the signal with a given packet error rate?, but we won?t get into that here.

The terminology used for receive sensitivity discussions can get a little complex. Going back to our analogy. You have ?good hearing? and your relative has ?poor hearing?. What does this mean ? It means that your ears are more sensitive than his. What does that mean ?

Imagine that, in turn, you and your relative are put in a soundproof room and sat at a table. The power from a audio oscillator ( at a particular frequency ) is gradually increased until such time as you say ?OK, I can hear that?. It would be found that the power required from the oscillator would need to be much higher in order for your relative to be able to first hear that tone. In other words, your ears are more sensitive than his.

When we look at a decibel scale of power, we can have three important areas ( zero dBm, above zero dBm and below zero dBm ). Zero dBm represents one milliwatt of power. If we have a value of transmit power of plus 16 dBm, we can quickly mentally ?see? that that value is higher than say plus 10 dBm ( humans are more comfortable working with postive numbers than negative ones ).

If we now take a value of -61 dBm and another of ? 71 dBm, we can see that the value of -61 dBm is higher in power than that of ? 71 dBm. That thought process does not come as quickly for most people as in the ?positive case?, as if we imagine a ?y ?axis? with zero dBm, less than zero dBm and greater than zero dBm, we need to understand that less negative numbers ( -61 dBm ) are ?more powerful?, literally, than more negative numbers ( -71 dBm ), in the power scenario.

In terms of sensitivity, it becomes a bit ?back to front? at first glance. For example, if a receiver can demodulate a particular signal with a particular data rate at a level of -71 dBm, while another requires ? 69 dBm to accomplish this, we say that the first receiver is more sensitive than the second, even though the power level of the first is less ( more negative ) than the power level of the second.

Whay would one receiver be more sensitive than another ? The quality of the electronics is important here. For quality, read ?cost?. When we come to Chapter 6, we shall discuss amplifier operation. Amplifiers ( in both the transmit and receive direction ) not only amplify the actual signal we are concerned with, but also the noise. As well as that, they actually generate noise internally. We shall learn of a factor known as the noise figure, which is related to this.

Semiconductor purity is very important, as impurities can increase the amount of intrinisic noise:
http://www.youtube.com/watch?v=i8kxymmjdoM

Now we need to take a look at the concept of Dynamic Rate Shifting, or DRS. In general, higher data rates require more sophisticated forms of modulation and coding. This in turn, requires higher values of Signal to Noise Ratio, or SNR.

When we are close to an AP, we may be able to use the highest rate possible that our particular client is configured for.

When a client goes through an association process, the client tells the AP ( via an information element ) what rates it is capable of supporting. The AP ( which has already advertized it?s range of rates via a similar information element in a beacon ), looks at what the client has suggested, and then lets the client know what is acceptable.

When we start to move away from an AP, the RSSI drops and hence the SNR ( assuming no interference/multipath ) drops at the same rate. Eventually, a position is reached where the SNR has dropped sufficiently that the demodulator in the radio receiver can no longer properly demodulate a signal having that bit rate. A rate shift down to the next value supported by that particular PHY standard then occurs.

For example, if we look at Table 2.2 on PP 32 -33 of the CWDP study guide, we can see that in changing from a rate of 54Mbps to a rate of 48 Mbps, we still use 64-QAM as the modulation method, however, the coding rate ( error-detecting and correction coding rate ) has now dropped from a rate of three quarters to two thirds. At each step, as we get further away from the AP, rate shifts will occur when we reach a physical point where the demod can no longer support that rate. In going from 54 Mbps down to 6 Mbps, we have gone from 64 ? QAM at three quarters coding rate down to BPSK at one half coding rate.
In documentation, we often see nice concentric circles drawn to represent rate shifting boundaries. Real life is much different from that, however, with multipath and interference etc. Some of the very important reasons for doing site surveys.

Next, we?ll look at the transmit side of things, and get to answer the original question.

Imagine that we have an STA operating at such a distance from an AP that it can satisfactorily receive a 5.5 Mbps signal. Now imagine that without changing the distance between the AP and STA, we wish to receive an 11 Mbps signal. Without any changes to antenna sizes etc, the only practical way is to have the AP increase it?s transmit power, in order to increase the SNR at the STA such that the demodulator can reliably demodulate that signal. However, this creates a problem. With the exception of proprietary Transmit Power Control ( TPC ) schemes, an AP has a Tx power level that is generally set based upon a site survey. The ?worst?client in the BSS usually is set to that same physical power value as per what the Antenna Reciprocity Theorem describes ( per a previous post ). Adjusting power levels on the fly as STA s come closer and move further away was not considered in the original IEEE specs.

It worked the other way around, as described previously. In other words, we would have to move closer to the AP in order to get an increase in data rate.

In the Wi-Fi circuitry in both an AP and an STA, we have a transmit amplifier. The purpose of that amplifier is to boost the relatively low powered signal from the modulator, and in turn send that signal to the antenna. We will read more about amplifiers in Chapter 6. We can look at a graph that shows the output power of an amplifier versus the input power. We would imagine that it would simply be a straight line from the origin at an angle of 45 degrees ( assuming equal scaling on the X and Y axes ). Indeed it is. However, as we go up in power, a strange thing happens. The straight line starts to become curved, and then levels off.

Now, when we are in the linear region of the graph ( Page 284, Figure 6.28 ), a one dB increase in input power will cause a one dB increase in output power. However, when we reach the non-linear ( curved ) portion, that is no longer true. In the non-linear region, other impairments occur, which can cause severe degradation of the output signal. The reasons for this can be quite complex:

http://www.google.com/url?sa=t&rct=j&q=amplifier%20non-linearities&source=web&cd=3&ved=0CDkQFjAC&url=http%3A%2F%2Fciteseerx.ist.psu.edu%2Fviewdoc%2Fdownload%3Fdoi%3D10.1.1.23.6084%26rep%3Drep1%26type%3Dpdf&ei=ItYUT-SkHISEtgen_szsAQ&usg=AFQjCNHbYcXlQV-p4zeVvsc5sRfY9zTjrw

However, we run into another problem with OFDM as opposed to the .11 and .11b technologies. The OFDM signal can have large ?peaks? compared to the average level of the signal ( more about this in Chapter 6 ). This means that the OFDM signals must be ?backed off? more than other previous technologies. This means often means that we cannot use the maximum possible transmit power of an AP for OFDM type signals, whereas we may have been able to us them with e.g. 1,2, 5.5 and 11 Mbps signals.

Lets now take a look at some spec sheets. First, let?s start with a client adaptor:

Cisco Aironet 802.11A/B/G Wireless CardBus Adapter:

http://www.cisco.com/en/US/prod/collateral/wireless/ps6442/ps4555/ps5818/product_data_sheet09186a00801ebc29.html

And now some APs:

Cisco 1130 AP:

http://www.cisco.com/en/US/prod/collateral/wireless/ps5678/ps6087/product_data_sheet0900aecd801b9058.html

Cisco 1140 AP:

http://www.cisco.com/en/US/prod/collateral/wireless/ps5678/ps10092/datasheet_c78-502793.html

Cisco 1250 AP:

http://www.cisco.com/en/US/prod/collateral/wireless/ps5678/ps6973/ps8382/product_data_sheet0900aecd806b7c5c.pdf

Cisco 1300 AP:

http://www.cisco.com/en/US/prod/collateral/wireless/ps5679/ps5861/product_data_sheet09186a00802252e1.html

As we can see, as mentioned by Theo, we can get information on data rates along with
the maximum power levels on the client adapter data sheet:

802.11b/g:
? 20 dBm (100 mW) @ 1, 2, 5.5, and 11 Mbps
? 18 dBm (63 mW) @ 1, 2, 5.5, 6, 9, 11, 12, 18, and 24 Mbps
? 17 dBm (50 mW) @ 1, 2, 5.5, 6, 9, 11, 12, 18, 24, and 36 Mbps
? 15 dBm (30 mW) @ 1, 2, 5.5, 6, 9, 11, 12, 18, 24, 36, and 48 Mbps
? 13 dBm (20 mW) @ 1, 2, 5.5, 6, 9, 11, 12, 18, 24, 36, 48, and 54 Mbps
? 10 dBm (10 mW) @ 1, 2, 5.5, 6, 9, 11, 12, 18, 24, 36, 48, and 54 Mbps

However, if we take a sample AP, say the Cisco 1130, we get the following:

802.11b:

CCK:

20 dBm (100 mW)
17 dBm (50 mW)
14 dBm (25 mW)
11 dBm (12 mW)
8 dBm (6 mW)
5 dBm (3 mW)
2 dBm (2 mW)
-1 dBm (1 mW)

As we can see, no mention of data rates. It is a similar situation with the other data
sheets. I dug through what seemed like a million documents, but couldn?t find any more
information. On a whim, I looked through the configuration guides for a few of the APs
and found a section on configuring Transmit Power:

?Set the transmit power for the 802.11g, 2.4-GHz radio to one of the power levels allowed in your regulatory domain. Settings are in dBm.
On the 2.4-GHz, 802.11g radio, you can set Orthogonal Frequency Division Multiplexing (OFDM) power levels and Complementary Code Keying (CCK) power levels. CCK modulation is supported by 802.11b and 802.11g devices. OFDM modulation is supported by 802.11g and 802.11a devices.
NoteSee the hardware installation guide for your access point to determine the power settings for your regulatory domain.
NoteThe 802.11g radio transmits at up to 100 mW for the 1, 2, 5.5, and 11Mbps data rates. However, for the 6, 9, 12, 18, 24, 36, 48, and 54Mbps data rates, the maximum transmit power for the 802.11g radio is 30 mW.?

That was more like it. At least now, we can match some power levels to data rates.

I started looking for the hardware installation guides. After a lot of dead links etc, I finally came across the following documents:

http://www.cisco.com/en/US/docs/wireless/access_point/channels/lwapp/reference/guide/lwap_chp.pdf

http://www.cisco.com/en/US/docs/wireless/access_point/1130/installation/guide/1130hig_book.pdf

The first document gives a lot of information about power levels and data rates. An example would be Table 10.3

Of course, the whole business of regulatory domains/antenna gain is another issue completely.

Dave

"Thanks Dave for your extended and very detailed reply!

So if i understand things right, the AP first sends a beacon with an IE advertising the data rates it can support. Then a client that wishes to associate with the AP sends its supported data rates. The demodulator of the AP based on the RSSI and SNR values of the frames it received from the client, chooses a Data rate and advertises it back to the client. Now the client is associated to the AP and both of them are using 48Mbps data rate.

The client moves away from the AP. After a few meters the RSSI of the frames coming from the AP are below the threshold of its receiver sensitivity for 48Mbps data rate, so the demodulator of the client receiver can't interpreter the signals it receives from the AP, so it switches the data rate down to 36Mbps (maybe it will also have to increase the output power a little bit). Same thing happens to the AP. When the client has moved away from it, the demodulator of the AP can't interpreter the signals coming from the client and switches to a lower Data rate.

What i am trying to say is that the decision about the DRS of the client is handled by its demodulator, based on RSSI information contained in frames received from the AP.
On the other hand same thing happens for the AP.
Generally speaking each device's demodulator takes its own decisions.

Is it possible for an AP to send frames at, lets say, 48Mbps and the client send frames at a lower data rate, for example 24Mbps? Or do they communicate with each other so that they will use the same data rate?

I hope that my understanding is correct

Finally i would really be grateful if somebody could point me the section that relates to all these in the 802.11-2007 document."

You?re welcome.

Unfortunately, one thing that you will find in the IEEE specs, is that very often when you wish to find out something, instead of the information all being in one area, it is often scattered throughout the spec. In fact, the IEEE gives very little guidance on this, leaving most of it up to the vendors. There are a few things that it does specify, which we shall talk about in a bit. However, we?ll take a look at some important areas in the next post or two.
One of the first things that happens is that the client will look at the receive signal strength of the beacons coming from the AP. From that value, it can ?back calculate? what data rate could be supported by a signal having that value of RSSI/SNR. What happens next is very vendor specific ( even if many of them won?t admit that ). The first thing that the client does is to make an assumption that both directions of the link are equal in terms of path loss and lack of interference. In other words, that the same conditions exist going from the STA to the AP and from the AP to the STA. The client will say ?Hey, if I am getting receive signal X with an RSSI of Y and an SNR of Z, then that should support a data rate of R?. ?Assuming link power balancing ( discussed in a previous post, assuming TX powers are equal at the AP and STA ) and there is no local interference at the AP, when I send data to the AP, he should be able to receive me with the same parameters that I receive him on.? We will discuss what happens when there is unequal power/interference etc later on, but there has to be a starting point and some assumptions at the beginning. Corrections can then be made later.

Let?s assume that the client figures out from the beacon information ( even though the beacons are generally sent at one of the lower basic rates so that ?everybody? can hear it, we are only looking at received power levels here?.more about basic rates later ) that a rate of 24 Mbps could be supported ( distance between AP and STA being a major factor ). It sends a frame with that data rate. If there is no local interference ( i.e. interference at the AP) etc, then the AP should be able to demodulate that frame correctly and in turn, send an ACK message back to the STA. If however, there is local interference ( or the STA is not transmitting at the same power as the AP etc ?.ignoring TPC here etc ), then perhaps the AP cannot properly demodulate the frame. In that case, it says ?Hey that frame is not good, I?m not going to reply?.

Depending on vendor ( and they keep a lot of this ?secret? ), the STA may try a few more times at that rate ( assuming that perhaps there was bit of interference on the link ). After a certain number of retries, it will say to itself ?Maybe I?m transmitting at too high a rate?..let me drop that rate down a bit until we can get an ACK from the AP?.

One thing to note is that in general, the STA is operating at a fixed output power ( ignoring proprietary power control schemes ), as is the AP.

Now, if there is local interference at the AP, but not at the STA, it is entirely possible to have the client ending up sending frames at a rate of 24 Mbps and the AP ending up sending frames at 48 Mbps ( say ). Based on whatever signal levels are being received upon first ?contact?, both sides will try to send at the maximum of the group of supported rates that both ends have jointly agreed upon.

So, to summarize: the AP ( assuming that there are no vendor specific ?adjustments? ) cannot tell the STA to transmit at a certain rate, directly ( again, no TPC ?.Transmit Power Control ). It can do so indirectly by not sending an ACK frame when it is unable to properly demodulate the frame. Eventually, the STA ?gets the message? and transmits at a reduced rate, perhaps having to repeat this process, until such time as the AP says ?Ah, now I can demodulate that frame I sent you??as a prize, have an ACK frame !!?

We have looked at the situation where we have just one STA. What about the situation where we have multiple STAs all capable of different maximum rates ?

IEEE 802.11 is nothing, if not fair. It basically says ?OK, I understand that there are some STAs out there who want to transmit at really high rates and using sophisticated modulation schemes, but we have to give everybody a fair chance?. This means that certain control frames such as CTS must not be sent at a rate higher than the initial frame to which it is replying ( RTS ). This is to avoid the situation where a ?slow? client is unable to understand a control signal, which would lead to all sorts of problems. We can also have a situation such as with 802.11g/b co-existence, where we have to use ?protection mechanisms? in order to allow both slow and fast STAs to be able to co-exist.

We?ll go into this in more detail; in the next post

Dave

Let?s now take a look at the terms ?Basic Rates? and ?Supported Rates?. After rummaging through the CWDP book, I found I good explanation on P 518. You can refer to that as we go through this:

WLANs are often a ?luck of the draw? as far as who is trying to connect to an AP. For example, suppose that you are in an office, and are operating on 802.11g. The spec says that the possible list of rates that may be supported includes:

6,9,12,18,24,36,48 and 54 Mbps. Great. We have a large range to work with. But what if someone has a laptop fitted with an 802.11b card ? Shouldn?t they be allowed to play as well ? Again, the IEEE 802.11 specs are nothing, if not fair ( as well as fairly long ). The spec says that if you want your equipment to operate at .11a OFDM rates, you must still be capable of allowing .11b clients to connect, and .11 clients.

Now, there are a number of problems ( especially at 1 and 2 Mbps ) with ?slow clients?. Firstly, if a device ( few and far between these days ) that only operated at 1 or 2 Mbps ( original 802.11 ) was ?fairly close? to an AP, it would indeed transmit at 2 Mbps, then as it moved away from the AP, it would reach a point where it had to drop down to 1 Mbps. This causes a number of problems. Firstly, 1 Mbps is slow as mollases. All the other ?fast? stations would have to wait until that station finished transmitting. The concept of Airtime Fairness helps address this issue ( but usually at ?higher, slow rates? ).

The next problem is that a 1 Mbps signal can be ?heard? from quite a distance away. This now means that the client may be wandering into another area and causing co-channel interference with say, another AP.

Even with a .11a client, which maybe started off at 54 Mbps, and wandered off elsewhere, it could drop down to 48, 36, 24, 18, 12, 11, 9, 6, 5.5, 2 and then 1 Mbps. Again, we could end up with the problems mentioned before.

In order to help around some of this problems, vendors utilize the concept of Basic Rates and Supported Rates.

Let?s look at Supported Rates first. Supported Rates are just the rates that the AP or STA can physically be capable of operating under. For example, if card is capable of having Supported Rates for all the .11a rates, then it would support 1, 2, 5.5, 6, 9, 11, 12, 18, 24, 36, 48 and 54 Mbps.

However, the IEEE says that you MUST support 6, 12 and 24 Mbps for 802.11a conformance. You must also be able to support 1,2, 5.5 and 11 Mbps. It does not force you to support 9, 18, 36, 48 and 54 Mbps, but the number of manufacturers who would not support those latter rates must be few and far between.

When management frames are transmitted, as well as some control frames, ( and we?ll see how Cisco does this in the next post ), a low common rate must be used so that if we have a mix of .11, 11b and .11a clients, everybody can hear the frames ( which are very important for system operation ). Beacon frames are management frames. Vitally important. They are the ?heartbeat?of the system, pumping WLAN blood at the rate of about 10 per second. Heed ye well anyone who tries to alter the beacon transmission rate ( which you can do in most cases )?..many pitfalls may occur.

So, Basic Rates are rates which everyone in the BSS must support as a common baseline. If we have .11 clients ( again rare ) in our network, then 1 and 2 Mbps will have to be listed as Basic Rates. It is a subset that represents the basic minimum that every client must support.

In order to help get around this, many manufacturers allow you to select which basic rates are allowed, by means of config screens, GUIs etc. For example, if you wanted to ?shut out? any .11 clients from even being allowed to associate in the first place ( ?You are not allowed to play, even though I am ?following? the IEEE specs of being ABLE to allow you to play?..? ). This would also mean that if a client dropped down from 54 Mbps to 6 Mbps to 5.5 Mbps, once he got out of range?.tough luck. He would have to either come back in range or do without ( assuming a single AP environment here).
Dave

So, now we know that management and certain control frames need to be sent at a basic rate. What about multicast/broadcast frames ? The IEEE docs tell us that these frames do not need to be acknowledged. There is a good reason for that. The Wi-Fi medium is a congested one, and 802.11 STAs operate in half duplex mode. You can imagine the impact on performance if all the STAs in the group had to ACK those frames, almost simultaneously. At the same time, we want to make all efforts to try and ensure that the multicast/broadcast frames have a good chance of reaching their intended recipients and being demodulated properly. Multicast/broadcast frames are usually sent at a basic rate.

Let?s now take a look at some of the things that Cisco does in respect to all this:

http://www.cisco.com/en/US/docs/routers/access/1800/wireless/configuration/guide/s37rf.html

If you open up the pdf document ( way to the right side of the html frame ), it makes it easier to read.

We can now go down to the section labelled ?Configuring Radio Data Rates?. Quoting:

?You use the data rate settings to choose the data rates the wireless device uses for data transmission. The
rates are expressed in megabits per second. The wireless device always attempts to transmit at the highest
data rate set to Basic, also called Require on the browser-based interface. If there are obstacles or
interference, the wireless device steps down to the highest rate that allows data transmission. You can
set each data rate to one of three states:

? Basic (the GUI labels Basic rates as Required)?Allows transmission at this rate for all packets, both
unicast and multicast. At least one of the wireless device's data rates must be set to Basic.

? Enabled?The wireless device transmits only unicast packets at this rate; multicast packets are sent
at one of the data rates set to Basic.

? Disabled?The wireless device does not transmit data at this rate.

Note At least one data rate must be set to basic.

You can use the Data Rate settings to set an access point to serve client devices operating at specific data
rates. For example, to set the 2.4-GHz radio for 11 megabits per second (Mbps) service only, set the
11-Mbps rate to Basic and set the other data rates to Disabled. To set the wireless device to serve only
client devices operating at 1 and 2 Mbps, set 1 and 2 to Basic and set the rest of the data rates to
Disabled. To set the 2.4-GHz, 802.11g radio to serve only 802.11g client devices, set any Orthogonal
Frequency Division Multiplexing (OFDM) data rate (6, 9, 12, 18, 24, 36, 48, 54) to Basic. To set the
5-GHz radio for 54 Mbps service only, set the 54-Mbps rate to Basic and set the other data rates to
Disabled.?

Also, in terms of multicast/broadcast:

?Access points running recent Cisco IOS versions are transmitting multicast and management frames at
the highest configured basic rate, and is a situation that could causes reliability problems.
Access points running LWAPP or autonomous IOS should transmit multicast and management frames at
the lowest configured basic rate. This is necessary in order to provide for good coverage at the cell's
edge, especially for unacknowledged multicast transmissions where multicast wireless transmissions
may fail to be received.
Since multicast frames are not retransmitted at the MAC layer, stations at the edge of the cell may fail
to receive them successfully. If reliable reception is a goal, then multicasts should be transmitted at a
low data rate. If support for high data rate multicasts is required, then it may be useful to shrink the cell
size and to disable all lower data rates.
Depending on your specific requirements, you can take the following action:
? If you need to transmit the multicast data with the greatest reliability and if there is no need for great
multicast bandwidth, then configure a single basic rate, one that is low enough to reach the edges of
the wireless cells.
? If you need to transmit the multicast data at a certain data rate in order to achieve a certain
throughput, then configure that rate as the highest basic rate. You can also set a lower basic rate for
coverage of non-multicast clients.?

I think we can see that what appears to be a simple job: draw a few concentric circles, switch on your APs and clients and off you go??.is not quite as simple as it seems.

In the next post, we'll take a look at some more terminology used by vendors, as well as a quick look at what IEEE 802.11-2007 says

Note: previous was a complilation of individual posts from today

Dave

Bit tricky finding definitions for Basic Rate in IEEE 802.11 - 2007.

Eventually found it ?hidden? under BSSBasicRateSet ( P 319 ):

?The set of data rates that must be supported
by all STAs that desire to join
this BSS. The STAs must be able to
receive and transmit at each of the data
rates listed in the set.?

Supported Rates ( P330 ):

?The set of data rates (in units of
500 kb/s) that are supported by
AP, including indication of which
rates are part of the BSSBasic-
RateSet (according to 7.3.2.2).?

On page 171 of the Cisco configuration guide in the following link, we find even more terminology:

http://www.cisco.com/en/US/docs/wireless/controller/7.0/configuration/guide/c70cg.pdf

?Step 6 Specify the rates at which data can be transmitted between the controller and the client by entering this command:
config {802.11a | 802.11b} rate {disabled | mandatory | supported} rate
where
?disabled?Clients specify the data rates used for communication.
?mandatory?Clients support this data rate in order to associate to an access point on the controller.
?supported?Any associated clients that support this data rate may communicate with the access point using that rate. However, the clients are not required to be able to use this rate in order to associate.
?rate?The rate at which data is transmitted:
?6, 9, 12, 18, 24, 36, 48, and 54 Mbps (802.11a)
?1, 2, 5.5, 6, 9, 11, 12, 18, 24, 36, 48, or 54 Mbps (802.11b/g)?

The following link gives some more information on Basic/Supported rates from another manufacturer:

http://www.moonblinkwifi.com/differences_in_80211b_and_802.cfm

Dave