Wi-Fi

Enterprise 802.11n: The fog may be clearing

There appears to be some semblance of a master plan rising out of the 802.11n implementation fog. Enterprise equipment developers have done their research and are incorporating additional technologies into 802.11n equipment that should garner some impressive wireless network performance characteristics.

There appears to be some semblance of a master plan rising out of the 802.11n implementation fog. Enterprise equipment developers have done their research and are incorporating additional technologies into 802.11n equipment and by doing so should garner some impressive wireless network performance characteristics. I’d like to point out the improvements in each of these areas, so we can all gain an appreciation of what the future Wi-Fi network will look like.

802.11n firmware enhancements

The proposed 802.11n standard has some very impressive improvements being made to the physical hardware and firmware. Most of these enhancements are already in place on pre-release equipment and are much of the reason why 802.11n equipment is performing better than expected.

  • The physical data rate selection algorithm (whew) has the tough job of determining what data transfer rate should be used based on the measured signal strength. 802.11a/g uses 12 steps from 1Mbps to 54Mbps. Whereas 802.11n has a total of 88 incremental data rate steps, which provides a more granular drop off when the signal strength weakens.
  • 802.11a/g uses transmit diversity which is useful and logical as the device transmits from the antenna that displayed the best reception characteristics during the last receive cycle. 802.11n uses spatial multiplexing, a technique that divides the information to be transmitted into independent and separately encoded data signals called streams. Each data stream is then transmitted from an independent antenna. The 802.11n standard allows up to four transmit streams. This is an interesting concept that increases data transmission capacity by multiplexing or reusing the space dimension multiple times. Ahh, the space-time continuum.
  • In order to be able to transmit multiple data streams, designers had to rework OFDM, which is the digital multi-carrier modulation scheme used by 802.11a/g. MIMO-OFDM used by 802.11n devices is the results. I would like to point out that this development is quite extraordinary as evidenced by the fact that it took almost ten years and a great deal of innovative work to perfect. MIMO-OFDM is viewed by many as the most singularly significant development of 802.11n.
  • 802.11a/g networks do not work well in multipath environments as I discussed in an earlier post. Basically having multiple—slightly different in phase or timing due to the environment—copies of the same transmitted RF signal arrive at the receiving antenna, drives the receiver nuts to put it politely. Since the real world is mostly a multipath environment, 802.11n was developed to make use of the slight differences exhibited by the arriving RF signals to distinguish the different data streams being transmitted. I discuss this really deep and intensely complicated subject in this post that quotes Dr. Greg Raleigh, who was instrumental in multipath research.
  • Channel size is one determinant of how much data can be passed over a wireless link. The 802.11a/g standard uses 20 MHz channels and history has proven that amount of bandwidth to be a limiting factor. In the past few years equipment developers have tried to improve the physical transfer rate of data by using proprietary technology which combined adjacent channels to support greater data rates. The 802.11n standard describes how to use the much wider 40 MHz channels—that are easy to implement, cost effective, and only require moderate increases in digital signal processing. If properly implemented, 40-MHz channels can provide greater than two times the usable channel bandwidth.
  • By design, TCP/IP traffic requires error-free transmission of data and one of the controls used to regulate the processing of traffic is the ACK bit. The ACK bit is sent by the receiver to acknowledge receipt of each frame, which turns out to be significant management overhead just for receipt verification. One way to improve throughput would be to devise a way to acknowledge the receipt but more efficiently. That is exactly what 802.11n does with the Block-ACK. By removing the need for one acknowledgment frame for every data frame, the amount of overhead required for the ACK frames, as well as preamble and framing, is reduced.
  • No stone was unturned when the developers were looking at ways to improve throughput and efficiency. Even the lowly guard interval was tweaked. The guard interval is used to prevent data loss from propagation anomalies as well as interference created if the following transmission starts too soon. Can that interval be reduced? It would help throughput, even if just slightly. 802.11n specifies two guard intervals 400ns and 800ns. If optimal conditions exist the 802.11n device will drop down to the 400ns guard interval to reduce what is considered unnecessary idle time. I guess if you have a bunch of 400ns intervals, it eventually adds up to something significant.
Final thoughts

I hope that I have been able to effectively convey some sense of all the subtle improvements that exist in the 802.11n standard. For the most part these enhancements are already designed into the pre-release hardware. Still 802.11n equipment developers are not satisfied with just these improvements. Smart antennas, multiple radios, and mesh technology are some heavy duty technologies that are being added to enterprise 802.11n appliances, which will allow 802.11n gear to evermore approach wired network parameters. Please stay tuned, as there are some pretty amazing things going on with "smart antenna systems" and "mesh networking."

About

Information is my field...Writing is my passion...Coupling the two is my mission.

10 comments
kathy
kathy

thanks for the information... __________________ cutie_tech123 Did you know there is a new cool and intimate new sushi place in Rome which offers high quality Japanese foods for eating or take-away, and offers great hand-made cakes and free wifi to all customers? http://naoko-sushi-roma.blogspot.com/

seanferd
seanferd

This is exactly what I found early on in my reading of your article. I was thinking that you had created a concise and easy-to-understand list of changes in 802.11n. Thanks for the explanations.

Michael Kassner
Michael Kassner

This post covered the sophisticated firmware incorporated in 802.11n enterprise equipment. Is there any additional interest in learning how "smart antennas" and "multiple radio mesh networking" will add even more weight to 802.1n networks as the eventual replacement for wired networks?

anne.powel
anne.powel

In two or three paragraphs, you have summarized about 6 much longer articles that took me 12 times as long to read (and made less sense in the long run) Thank you!

JimTeach
JimTeach

Some good information. It seems that 802.11a/g used half duplex transmission by the 1:1 Message/ACK ratio. Is the Block ACK in 802.11n the same as sliding windows except at a lower layer?

Timbo Zimbabwe
Timbo Zimbabwe

... but I can tell you of my experience. I get considerable throughput on my 802.11n network, to the point that it is faster than the connection to my service provider, so I don't feel that I'm being "held back" by my own equipment.

Michael Kassner
Michael Kassner

I believe you are correct in that the intent of both processes is the same. How it is achieved and the sophistication is where they appear to be different. First, I hope it is OK to maybe clear up a small point. All radio communications use half duplex transmission unless different and non-interfering frequencies are used. It is the nature of the beast so to speak; you can?t send and receive at the same time. The problem with wireless is that the data throughput rate normally associated with half duplex gets worse in the 802.11 wireless realm with the addition of management overhead. That is not directly considered a duplex problem, but the real world results are the same in that data throughput is less than what could be expected in a half duplex wired link. I found in my notes some information from Cisco about the TCP sliding window: ?A TCP sliding window provides more efficient use of network bandwidth than PAR because it enables hosts to send multiple bytes or packets before waiting for an acknowledgment. In TCP, the receiver specifies the current window size in every packet. Because TCP provides a byte-stream connection, window sizes are expressed in bytes. This means that a window is the number of data bytes that the sender is allowed to send before waiting for an acknowledgment. Initial window sizes are indicated at connection setup, but might vary throughout the data transfer to provide flow control. A window size of zero, for instance, means "Send no data." In a TCP sliding-window operation, for example, the sender might have a sequence of bytes to send (numbered 1 to 10) to a receiver who has a window size of five. The sender then would place a window around the first five bytes and transmit them together. It would then wait for an acknowledgment. The receiver would respond with an ACK = 6, indicating that it has received bytes 1 to 5 and is expecting byte 6 next. In the same packet, the receiver would indicate that its window size is 5. The sender then would move the sliding window five bytes to the right and transmit bytes 6 to 10. The receiver would respond with an ACK = 11, indicating that it is expecting sequenced byte 11 next. In this packet, the receiver might indicate that its window size is 0 (because, for example, its internal buffers are full). At this point, the sender cannot send any more bytes until the receiver sends another packet with a window size greater than 0.? As for the Block-ACK that originated with 802.11e (QoS protocol), it was one of many enhancements to try and minimize several problems of delay, jitter, and trying to guarantee bandwidth for real-time applications being sent over a wireless link. The Block-ACK is similar to the TCP sliding window approach, but as far as I know it is not as sophisticated. I am not sure, but I do not believe that it has the ability to adjust the amount of data frames being sent per burst like the TCP sliding window. I have linked two very well written articles about 802.11e, the CommsDesign article goes into a great depth explaining the actual process. As a small aside, if at all interested CommDesign and their associated websites are a treasure-trove of information. http://www.newlogic.com/products/802_11_wireless_abg/mandating_qos_in_wireless_lans.pdf http://www.commsdesign.com/article/printableArticle.jhtml;jsessionid=IKELPRA11KUPGQSNDLOSKH0CJUNN2JVN?articleID=17000388

Michael Kassner
Michael Kassner

Cisco instructor and protocol analyst. I am very glad to meet you. Now I know where to go with my questions. Reason being is that I also am employed as a network field engineer for Orange Business Services. If you have any other questions I would be more than happy to try and help. Also if you or your students have any topics that they would like to see discussed, please let me know as well.

JimTeach
JimTeach

Thanks for the great explanation, clarification, and articles. Sorry it took so long to respond - Spring Break from everything! Since I am a Cisco Networking Academy instructor, and a retired protocol analyst and specialist, I like to keep ahead of my students. I know they will ask me this same question when we get to wireless. You saved my neck and give me support. Trying to locate some of the minute details about protocol changes gets harder and much more time consumming. Keep up the great information so we can keep our students and engineers up-to-date.

Editor's Picks