With MIMO based on spatial division multiplexing (SDM) in 802.11n, one device transmits multiple data streams to another device. With the inclusion of downlink multi-user MIMO (DL MU-MIMO) in 802.11ac, an access point (AP) simultaneously transmits independent data streams to multiple client devices. Consider the example illustrated in Figure 14.1, with an AP and two client devices. In the example the AP has four antennas, one client device has two antennas (device A in Figure 14.1), and the other client device has one antenna (device B in Figure 14.1). An AP can simultaneously transmit two data streams to device A and one data stream to device B.

The primary advantage of DL MU-MIMO is that client devices with limited capability (few or one antenna) do not degrade the network capacity by occupying too much time on air due to their lower data rates. Consequenty, downlink capacity, which is based on the aggregate throughput of the clients that receive the simultaneous transmission using DL MU-MIMO, is improved over that of 802.11n. However, the benefits of DL MU-MIMO do come with increased cost and complexity.

Link adaptation is the process by which the transmitter selects the optimal MCS with which to send data to a particular receiver. Link adaptation algorithms are implementation specific, however, they are generally based on the measured packet error rate (PER). Most algorithms monitor the PER and adjust the MCS to track an optimal long term average that balances the reduced overhead from sending shorter packets with a higher MCS with the increased overhead from retransmissions due to the increased PER from the higher MCS.

Determining the PER by necessity means monitoring packet errors over a period that is long in comparison with the duration of a packet. For example, to very roughly measure a 10% PER requires that the transmitter send ten packets of which one is in error. Because of this, link adaptation based on PER adapts slowly to changing channel conditions. In many environments the channel is changing with time as the stations move or with changes in the environment itself, such as the 50 Hz or 60 Hz ionizing cycle in fluorescent bulbs, the movement of objects in the environment, or changes in external noise sources. These changing conditions may occur on time scales faster than PER can be measured. As a result the link adaptation algorithm is choosing an MCS that is the long term optimal MCS and not the instantaneously optimal MCS.

The 802.11n standard is the high throughput amendment to the 802.11 standard. This section describes all the aspects of the physical layer which increase the data rate. MIMO/SDM is a key feature of 802.11n, which is discussed in Chapters 3 and 4. The other significant increase in data rate is derived from the new 40 MHz channel width. This section also discusses short guard interval, Greenfield preamble, and other modifications to the 20 MHz waveform.

40 MHz channel

In the last several years, regulatory domains have made much more spectrum available for unlicensed operation in the 5.47–5.725 GHz band for wireless local area networking. The addition of the new spectrum has more than doubled the number of available 20 MHz channels in the USA and Europe. Table 5.9 and Table 5.10 in Appendix 5.1 describe the current allocation in the USA and Europe, and the corresponding 802.11 channel number. Even with doubling the channel width to create 40 MHz channels, there are still more channels available for frequency re-use than in the early days of 802.11a. Furthermore, products currently in the market with proprietary 40 MHz modes have demonstrated that the cost for 40 MHz products is roughly the same as for 20 MHz products. Therefore, with free spectrum and relatively no increase in hardware cost, doubling the channel bandwidth is the simplest and most cost effective way to increase data rate.

The major new PHY features and enhancements for Very High Throughput (VHT) in 802.11ac include wider bandwidth, modulation/coding/spatial streams, and downlink multi-user (MU) MIMO. The channel bandwidth in 802.11ac was expanded to include 80 MHz and 160 MHz waveforms. With 80 MHz and four spatial streams the maximum data rate is 1733.3 Mbps and with 160 MHz and four spatial streams the maximum data rate is 3466.7 Mbps using the newly defined 256-QAM modulation. A new non-contiguous 80+80 MHz waveform design is also included in VHT to better fit a 160 MHz transmission into the available spectrum. The maximum number of spatial streams increases from four in 802.11n to eight. With 160 MHz and eight spatial streams, the maximum VHT data rate is 6933.3 Mbps. VHT includes additional smaller increases to data rate with 256-QAM. With the addition of downlink multi-user (MU) MIMO to the 802.11 standard, we now have a single user (SU) and MU packet structure. As we will see, these two packet structures have a common VHT preamble format. This is different from 802.11n, which has both a MF and GF preamble format. This chapter will review the new packet structure and waveform for VHT designed in 802.11ac to support the new features for single user transmission. MU transmission will be addressed in Chapter 14. We will also review minor modifications to features borrowed from 802.11n, including STBC and LDPC.

Channelization

The channelization for 802.11ac is the same as 802.11n for 20 and 40 MHz, with the addition of channel 144 as illustrated in Figure 7.1. Originally, the 802.11 channel numbers ended at 140 in the 5470–5725 MHz band to leave guard band between it and the 5725–5850 band. However, in recent discussions with the FCC, they indicated that overlap between 5150–5250 MHz and 5250–5350 MHz bands and 5470–5725 MHz and 5725–5850 bands is permitted. Note that regulations regarding maximum transmit power and radar detection must be observed in the respective bands. This new guideline allowed for channel 144 and also for 160 MHz channels.

Early on in the 802.11n standardization process it was recognized that even with significantly higher data rates in the PHY the fixed overhead in the MAC protocol was such that little of that gain would be experienced above the MAC. It was clear, as this chapter will show, that without throughput enhancements in the MAC the end user would benefit little from the improved PHY performance.

The throughput enhancements introduced in 802.11n are reused in 802.11ac. Some aspects of 802.11n, particularly A-MPDU and HT-immediate block ack, that were challenging to implement at the time 802.11n was developed are now widely adopted.

Reasons for change

Since the original 802.11 specification was completed, a number of amendments have introduced new PHY capabilities and with them enhanced performance. In addition, the 802.11e amendment, which primarily added QoS features, also enhanced MAC performance with the introduction of the TXOP concept and block acknowledgement. However, these MAC performance improvements were only slight, and with the potential for significantly higher PHY performance it was soon realized that the existing MAC protocol did not scale well with PHY data rate.

This section provides details on the MAC frame formats. The information provided here is sufficiently detailed to act as a reference for the topics discussed in this book, but it does not provide an exhaustive list of all field elements, particularly in the management frames. For a detailed treatment of the frame formats refer to the actual specification (IEEE, 2012a, 2012b).

General frame format

Each MAC frame consists of the following:

• a MAC header

• a variable length frame body that contains information specific to the frame type or subtype

• a frame check sequence or FCS that contains a 32-bit CRC.

This frame format consists of a set of fields that occur in a fixed order as illustrated in Figure 12.1. Not all fields are present in all frame types.

Frame Control field

The Frame Control field is shown in Figure 12.2 and is composed of a number of subfields described below

Protocol Version field

This field is 2 bits in length and is set to 0. The protocol version will only be changed when a fundamental incompatibility exists between a new revision and the prior edition of the standard, which to date has not happened.

Adaptive transmit beamforming was introduced with the 802.11n standard amendment. With the 802.11ac standard amendment, transmit beamforming is greatly simplified using the experience gained from industry attempts to implement the various 802.11n transmit beamforming options.

With transmit beamforming (TxBF), we apply weights to the transmitted signal to improve reception. The weights are adapted from knowledge of the propagation environment or channel state information (CSI). Since by definition transmit beamforming weights are derived from channel information, spatial expansion as defined in Section 6.2 is not considered transmit beamforming.

The key advantage with transmit beamforming is the ability to significantly improve link performance to a low cost, low complexity device. This advantage is illustrated in Figure 13.1, which depicts a beamforming device with four antennas. Such a device could be an AP or a home media gateway. The device at the other end of the link has only two antennas, typical of a small client device. Such a system would benefit from 4 × 2 transmit beamforming gain from device A to device B. However, when transmitting from device B to device A, the system gain would be matched with 2 × 4 SDM with MRC as described in Section 6.1. Therefore link performance would be balanced in both directions.

Wireless local area networking has experienced tremendous growth in the last ten years with the proliferation of IEEE 802.11 devices. Its beginnings date back to Hertz’s discovery of radio waves in 1888, followed by Marconi’s initial experimentation with transmission and reception of radio waves over long distances in 1894. In the following century, radio communication and radar proved to be invaluable to the military, which included the development of spread spectrum technology. The first packet-based wireless network, ALOHANET, was created by researchers at the University of Hawaii in 1971. Seven computers were deployed over four islands communicating with a central computer in a bi-directional star topology.

A milestone event for commercial wireless local area networks (WLANs) came about in 1985 when the United States Federal Communications Commission (FCC) allowed the use of the experimental industrial, scientific, and medical (ISM) radio bands for the commercial application of spread spectrum technology. Several generations of proprietary WLAN devices were developed to use these bands, including WaveLAN by Bell Labs. These initial systems were expensive and deployment was only feasible when running cable was difficult.

Having worked on the development of the 802.11n standard for some time, we presented a full day tutorial on the 802.11n physical layer (PHY) and medium access control (MAC) layer at the IEEE Globecom conference held in San Francisco in December 2006. Our objective was to provide a high level overview of the draft standard since, at the time, there was very little information on the details of the 802.11n standard available to those not intimately involved in its development. After the tutorial, we were approached by Phil Meyler of Cambridge University Press and asked to consider expanding the tutorial into a book.

Writing a book describing the standard was an intriguing prospect. We felt that a book would provide more opportunity to present the technical details in the standard than was possible with the tutorial. It would fill the gap we saw in the market for a detailed description of what is destined to be one of the most widely implemented wireless technologies. While the standard itself conveys details on what is needed for interoperability, it lacks the background on why particular options should be implemented, where particular aspects came from, the constraints under which they were designed, or the benefit they provide. All this we hoped to capture in the book. The benefits various features provide, particularly in the physical layer, are quantified by simulation results. We wanted to provide enough information to enable the reader to model the physical layer and benchmark their simulation against our results. Finally, with the amended standard now approaching 2500 pages, we hoped to provide an accessible window into the most important aspects, focusing on the throughput and robustness enhancements and the foundations on which these are built.

This chapter introduces some of the advanced channel access techniques in the 802.11 standard. In addition to, and built upon, the distributed channel access techniques described in Chapter 7, the 802.11 standard includes two centrally coordinated channel access techniques. The PCF was introduced in the original 802.11 specification and HCCA was introduced in the 802.11e amendment to support parameterized QoS and to fix some of the deficiencies in the PCF. The chapter then introduces new channel access techniques in the 802.11n amendment.

A very simple technique called the reverse direction protocol was introduced with one bit of additional signaling and some simple changes to the rules for operating a TXOP. This technique is particularly effective under EDCA for improving throughput for certain traffic patterns.

During the development of the 802.11n amendment, a strong interest emerged among many participants for improving the power efficiency of the MAC protocol. While outside the scope of the 802.11n PAR this resulted in the power-save multi-poll (PSMP).

PCF

Infrastructure network configurations may optionally include the point coordination function (PCF). With PCF, the point coordinator (PC), which resides in the AP, establishes a periodic contention free period (CFP) during which contention free access to the wireless medium is coordinated by the PC. During the CFP the NAVof all nearby stations is set to the maximum expected duration of the CFP. In addition, all frame transfers during the CFP use an inter-frame spacing that is less than that used to access the medium under DCF, preventing stations from gaining access to the medium using contention-based mechanisms. At the end of the CFP, the PC resets the NAV of all stations and regular contention-based access proceeds.

The 802.11n amendment introduced many optional features and the 802.11ac amendment introduces a few more. These optional features are geared toward specific market segments that will likely only be deployed in specific classes of devices. Many of the optional features are complex and, given time to market concerns and cost constraints, many implementations will only adopt a subset of the available features, perhaps phasing in features with time or in higher end products. Some features appear in multiple flavors. This is due to the many unknowns present at the time the features were being discussed during the standard development process. Very often there was no single clear direction to take that would clearly suit all situations and so variations were introduced.

The popularity of 802.11 also means that there are a large number of legacy 802.11 devices deployed making interoperability and coexistence with those devices essential.

This chapter discusses various features that help ensure interoperability and coexistence between 802.11ac, 802.11n, and legacy 802.11 devices.

Station capabilities and operation

With the large number of optional features in the 802.11n amendment and a few additional optional features in 802.11ac, a fair bit of signaling is required to establish device capabilities and ensure interoperability. Also, care must be taken to ensure that a feature used by one station does not adversely affect the operation of a neighboring station that is not directly involved in the data frame exchange.

While 802.11n was a revolutionary enhancement over 802.11a/g, and necessitated an entire book for proper presentation of the new technology, 802.11ac is more of an evolutionary improvement over 802.11n by providing wider bandwidth channels and multi-user MIMO. As such we felt that the treatment of new 802.11ac features could be addressed by a few extra chapters in an update of our original 802.11n book, now an 802.11n/ac book.

The new single user Very High Throughput physical layer packet structure is described in a new Chapter 7, including a description of 80 MHz and 160 MHz waveforms. The new downlink multi-user MIMO mechanism in 802.11ac is presented in a new Chapter 14. Enhancements to channel access for 802.11ac have been added to Chapter 11. Several new basic service set and clear channel assessment rules to manage 80 MHz and 160 MHz operation and coexistence are described in Chapter 11. Modifications to 802.11n channel model Doppler component are given in Chapter 3 and Chapter 13. Furthermore, we discuss the simplification of single user transmit beamforming in 802.11ac in Chapter 13.

Foreword

The first version of the 802.11 standard was ratified in 1997 after seven long years of development. However, initial adoption of this new technology was slow, partly because of the low penetration of devices that needed the “freedom of wireless.”

The real opportunity for 802.11 came with the increased popularity of laptop computers just a few years later. This popularity brought a rapidly growing user base wanting network connectivity not only while connected to an Ethernet cable at home or at work, but also in between: in hotels, airports, conference centers, restaurants, parks, etc. 802.11 provided a cheap and easy way to make laptop mobility a reality for anyone who wanted it.

However, technology by itself is rarely sufficient, particularly in the networking space, where interoperability of devices from multiple vendors is almost always the key to market success. Having been formed as WECA in 1999, the Wi-Fi Alliance was ready to provide certification of multi-vendor interoperability.

With the right technology from the IEEE 802.11 Working Group, certified interoperability from the Wi-Fi Alliance, and a real market need based on a growing installed base of laptops, the conditions were ripe for the Wi-Fi market to take off, and indeed it did. By 2007 virtually every new laptop contains Wi-Fi as standard equipment. More importantly, and unlike some other “successful” wireless technologies, many of these devices are used regularly. With this wide use came a growing understanding of the power of cheap, easy-to-deploy, and easy-to-manage interoperable Wi-Fi networks.