What Are 802.11ac and 802.11ac Wave 2

Channel Bandwidth

802.11ac adds 80 MHz and 160 MHz bandwidths. 802.11n supports 20 MHz and 40 MHz bandwidths, where 20 MHz bandwidth is mandatory and 40 MHz bandwidth is optional. 802.11ac supports 20 MHz, 40 MHz, 80 MHz, 80+80 MHz (non-continuous, non-overlapping), and 160 MHz, where 20 MHz, 40 MHz, and 80 MHz bandwidths are mandatory, and 80+80 MHz and 160 MHz bandwidths are optional. The following figure uses North American spectrum as an example and illustrates the differences between 802.11ac, 802.11n, and 802.11a. For 160 MHz bandwidth, 802.11ac supports 2 continuous or non-continuous 80 MHz channels.

Figure 1-2 802.11ac channel bandwidth

The variable bandwidth design reserves compatibility with small channel bandwidth. In addition, increased bandwidth also improves the maximum throughput and brings better user experience.

Channel bandwidth scalability also causes the interference when multiple channels are used. 802.11ac needs to manage channel bandwidth management efficiently to reduce channel interference and make full use of spectrum bandwidth.

Frequency Band

Original Wi-Fi systems define 2.4 GHz or 5 GHz frequency band. 802.11n supports both 2.4 GHz and 5 GHz frequency bands. There are obvious problems at the 2.4 GHz frequency band as Wi-Fi applications are increasingly used.

• Congested frequencies: A large number of non-Wi-Fi devices such as baby monitors, microwave ovens, and cordless telephones also work at the 2.4 GHz frequency band. Interferences from these devices affect Wi-Fi performance, and Wi-Fi cannot effectively solve these problems.
• Fewer frequency resources: The 2.4 GHz frequency band has only 83.5 MHz frequency resources. Fewer frequency resources indicate more frequency multiplexing and interferences. In addition, high-channel-bandwidth networking is limited, and the Wi-Fi maximum throughput cannot be fully used.

802.11ac does not support the 2.4 GHz frequency band. It prevents interferences at the 2.4 GHz frequency band and promotes popularity of terminals at the 5 GHz frequency band. In the 802.11n era, over half of terminals on the live network support only the 2.4 GHz frequency band.

Although 802.11ac defines the frequency band less than 6 GHz frequency band (excluding 2.4 GHz frequency band), the mainstream frequency band is still 5 GHz. 802.11ac is also called 5G Wi-Fi.

MCS

802.11n defines eight MCS modes for each MIMO combination. There are four modulation modes: BPSK, QPSK, 16QAM, and 64QAM.

To improve the maximum throughput, 802.11ac uses higher-order modulation 256Q-AM with improved modulation efficiency. 802.11ac supports code rates 3/4 and 5/6 and 10 MCS modes. Original 802.11 standards provide MCS coding for each MIMO combination, which is abandoned by 802.11ac. Therefore, there are only 10 MCS coding modes in 802.11ac. A higher MCS value indicates higher maximum throughput. This is because different modulation coding modes use different numbers of bits in each sub-carrier. Each sub-carrier represents 2 bits in BPSK mode, 4 bits in 16QAM mode, 6 bits in 64QAM mode, and 8 bits in 256QAM mode. The following constellation figure shows BPSK, QPSK, 16QAM, 64QAM, and 256QAM. A higher order modulation mode achieves a higher modulation efficiency. The modulation efficiency is not improved linearly. The modulation efficiency in latter modulation modes is slightly improved.

Figure 1-3 Different modulation modes

256QAM improves efficiency, but has strict requirements for the wireless environment and demands higher SNR than 64QAM. Therefore, MCS8 and MCS9 are often applicable to scenarios where STAs are close to APs. In the scenarios, serviceable signals are strong and interference signals are weak, meeting SNR (SNR = Useful signals/Interference signals) requirements.

Single-user MIMO

MIMO falls into single-user MIMO and multi-user MIMO. MIMO uses spatial diversity and multiplexing. Although spatial diversity cannot directly improve the maximum throughput, the spatial diversity gain can increase the SNR so that a link can improve the capacity using a higher order modulation mode. Spatial multiplexing transmits multiple data streams of a single user or data streams of multiple users simultaneously without changing the channel bandwidth.

In Wi-Fi applications, Transmit Beamforming (TxBF) gains much attention. TxBF definition in 802.11n is complex, so TxBF is not well recognized in markets. 802.11ac simplifies the design.

1. 802.11n defines explicit and implicit Beamforming modes, but 802.11ac supports only explicit Beamforming.
   Figure 1-4 Explicit and implicit Beamforming
   
2. 802.11ac improves channel probe and feedback mechanisms. 802.11n uses the following modes to probe channels: Null Data Packets (NDPs) and staggered preamble. 802.11n defines three feedback formats: CSI, noncompressed, and compressed. It also defines immediate and delayed feedback modes. 802.11ac uses only NDPs to probe channels and supports only the compressed V matrix format and immediate mode.

   802.11n supports spatial multiplexing for multiple streams. 802.11nwasthe first to introduce MIMO technology to Wi-Fi. It supports a maximum of four streams and provides the maximum throughput of up to 600 Mbit/s, which is a qualitative leap compared with 802.11a/b/g.802.11ac supports a maximum of eight streams and provides the maximum throughput of 7 Gbit/s for a single user.

   Spatial diversity and multiplexing use the multi-antenna system. To support eight streams, APs and STAs require eight antennas, which is a great challenge to both APs and STAs. More antennas increase device complexity, dimensions, and costs. This is also the reason why the mainstream 802.11n APs use dual antennas and STAs use single antenna although 802.11n can support four streams.

   Figure 1-5 8x8 MIMO
   

Multi-user MIMO

The use of multiple streams increases the maximum throughput of a single user. However, many terminals, especially mobile smart terminals, use a single stream. A single-stream terminal takes more time on the air interface to transmit data of the same size than a multi-stream terminal. Therefore, single-stream terminals become the bottleneck for increasing access users. Multi-user MIMO is a good choice. An AP can send different data to multiple users (a maximum of four users) simultaneously without changing the user bandwidth and frequency band.

Figure 1-6 Comparisons between single-user MIMO and multi-user MIMO

When an AP in the same frequency band sends data to multiple users simultaneously, signals of streams sent to a user cause interference to signals of streams sent to another user.

Multi-user MIMO needs to work with TXBF to complete channel probe. The sender uses pre-coding technology to eliminate the interference according to the feedback matrix.

Figure 1-7 Interferences between multiple users

802.11ac supports only downlink multi-user MIMO and is able to transmit data to a maximum of four users. Uplink data is sent one by one, and cannot be sent simultaneously. When sizes of user packets to be transmitted simultaneously are different, frame padding is used. Scheduled BA mechanism is used to schedule ACK response messages of each user so that ACK messages are sent one by one.

When an AP supports Enhanced Distributed Channel Access (EDCA), priorities of different user services may be different. In this case, user service packets are sent to different AC queues. Multi-user MIMO uses the transmission opportunity (TXOP) to transmit packets with different priorities simultaneously.

Multi-user MIMO increases the number of concurrent users connected to a single AP. In scenarios using single-stream terminals, multi-user MIMO increases the number of concurrent users and an AP's downlink maximum throughput. When data streams are transmitted to multiple users, interference between streams affects higher order modulation mode. For example, 256QAM cannot be used in this scenario.

Dynamic Channel Management

802.11ac supports wide channel bandwidths from 20 MHz to 160 MHz, which also brings challenges to channel management. When different channel bandwidths are used, proper management methods must be used to reduce interference between channels and fully use channels.

802.11ac defines an enhanced Request to Send/Clear to Send (RTS/CTS) mechanism to determine when channels are available. The mechanism is as follows:

1. An 802.11ac device sends an RTS. Basic 802.11a transmission, which is 20 MHz wide, is replicated another three times to fill the 80 MHz or another seven times to fill 160 MHz. Each nearby device, regardless of whether the primary channel is the 20 MHz channel over the 80 MHz or 160 MHz channel, can receive the RTS. Each device that receives the RTS sets virtual sub-channels in busy state.
2. The device that receives the RTS checks whether the primary channel or sub-channels of the 80 MHz channel are busy. If some channel bandwidth is used, the receiver replies with a CTS with available bandwidth and reports repeated bandwidth.
3. A CTS is sent over each available 20 MHz sub-channel.

The sender can learn available and unavailable channels. Then data is sent only over available sub-channels.

Figure 1-8 Dynamic spectrum management

Figure 1-8 compares 802.11n and 802.11ac. In 802.11n, if a sub-channel is unavailable, the entire bandwidth is unavailable. In 802.11ac, if some sub-channels are unavailable, other sub-channels can still be used to send data.

Dynamic bandwidth management is designed for spectrum multiplexing. This function increases channel use efficiency and reduces interference between channels. Therefore, two APs can work in the same bandwidth channel.

Figure 1-9 Two APs over the same 80 MHz channel

Compatibility

802.11ac defines the following preamble formats: Greenfield and Mixed. Because Greenfield does not consider compatibility, 802.11ac does not use this format. 802.11ac improves the Mixed format to ensure compatibility with original 802.11 standards.

An 802.11ac device can detect the preamble and pilot in the frame format used by an access device to differentiate the 802.11 standard used by the access device and adapt to the access device. The following figure shows the formats of 802.11n and 802.11ac frames.

Figure 1-10 Formats of 802.11n and 802.11ac frames

The short training field (STF), long training field (LTF), and signal field (SIG) are used to ensure compatibility with 802.11a/b/g/n. The letter L indicates Legacy. The first symbol of VHT-SIG-A is BPSK modulated, and the second symbol is BPSK rotated by 90 degrees rotation (QBPSK) used to differentiate HT and VHT modes. VHT-STF in 802.11ac is used to enhance the automatic gain control in a MIMO transmission. VHT-LTF is used by the receiver to estimate the MIMO channel between the transmit and receive antennas. According to the total number of spatial streams, there can be 1, 2, 4, 6, or 8 VHT-LTFs. In 802.11ac, 1, 2, or 4 VHT-LTFs are used for mapping, and 6 or 8 VHT-LTFs are used for spatial streams. VHT-SIG-B indicates the length of data to be transmitted, modulation mode, and coding mode.

Frame Aggregation

On a Wi-Fi network, each frame is transmitted on an air interface in CSMA/CA mode. When many frames are transmitted, collisions reduce the air interface use efficiency. 802.11n starts to use frame aggregation at the MAC address layer. MSDUs or MPDUs are aggregated, and then encapsulated at the physical layer. This improves encapsulation efficiency and reduces usage and preemption on the air interface.

Figure 1-11 A-MSDU and A-MPDU

Figure 1-11 shows A-MSDU and A-MPDU encapsulation. The two aggregation modes can improve encapsulation efficiency, but A-MPDU has the following advantage that A-MSDU does not have: When an error occurs during transmission, A-MSDU needs to retransmit the entire aggregated frame, while A-MPDU only needs to retransmit the error data packets because each MPDU has its MAC address header. Therefore, A-MPDU is used more frequently than A-MSDU.

To further improve efficiency and reliability, 802.11ac increases the MPDU size and A-MPDU frame size. 802.11ac supports only A-MPDU.

Wider Channel Bonding

IEEE 802.11n supports only two bandwidth modes: 20 MHz and 40 MHz. The 20 MHz mode is mandatory while the 40 MHz mode is optional. IEEE 802.11ac supports 20 MHz, 40 MHz, 80 MHz, 80+80 MHz (non- adjacent), and 160 MHz channel bandwidth. The 20 MHz, 40 MHz, and 80 MHz modes are mandatory while the 80+80 MHz and 160 MHz modes are optional. 802.11ac Wave 1 defined by the WFA supports 20 MHz, 40 MHz, and 80 MHz channel bandwidth. 802.11ac Wave 2 defined by the WFA supports adjacent and non-adjacent 160 MHz channel bonding. Figure 1-12 uses the frequency spectrum in North America as an example to compare channel bonding in 802.11ac Wave 1, 802.11ac Wave 2, 802.11n, and 802.11a.

Figure 1-12 Bandwidth in 802.11ac Wave 2

As shown in Figure 1-12, if the channel bandwidth is 20 MHz, 40 MHz, or 80 MHz, there are 25, 12, or 6 channels respectively. If the channel bandwidth is 160 MHz, there are two adjacent channels. The 160 MHz channel can be a combination of two non-overlapping 80 MHz channels. Channel bonding allows a flexible combination of channels. For example, to avoid the use of DFS channels, users can bind two non-DFS 80 MHz channels into a 160 MHz channel. In 80+80 MHz channel bonding mode, up to 13 bonding methods are supported.

MU-MIMO

SU-MIMO can increase the throughput of a single user significantly. However, most STAs, especially mobile smart STAs, on live networks support one stream only. Compared with multi-stream STAs, single-stream STAs occupy air interfaces for a longer period when they transmit data of the same size. Therefore, single-stream STAs become a bottleneck for increasing the number of access users. MU-MIMO is a good solution to this problem. With the user bandwidth and frequency unchanged, an AP can concurrently transmit different data to four users at most. Figure 1-15 compares the SU-MIMO and MU-MIMO transmission modes of a 4x4 MIMO AP. In the SU-MIMO transmission mode, all antennas of the AP send the same data. Although this transmission mode provides diversity gains, the gains are limited. In the MU-MIMO transmission mode, antennas of the AP transmit different data to different users. A single AP can send four different data packets, increasing the efficiency by four times than that in single-MIMO transmission mode.

Figure 1-15 Comparison between SU-MIMO and MU-MIMO

MU-MIMO is also applicable to scenarios where both multi-stream and single-stream STAs exist. For example, Figure 1-16 shows two application scenarios: one dual-stream STA + two single-stream STAs and two dual-stream STAs.

Figure 1-16 Application scenarios where both multi-stream and single-stream STAs exist

MU-MIMO is an outstanding feature of 802.11ac Wave 2, which depends on explicit transmit beamforming (TxBF). This feature requires that STAs support explicit TxBF. The reason is that when an AP concurrently transmits data to multiple users over the same frequency, signals are interference to users who are not target receivers of the signals. MU-MIMO uses TxBF to detect channels and uses precoding technology based on the feedback to mitigate such interference.

Figure 1-17 Communication between one 3x3 MIMO AP and three 1x1 MIMO STAs

Figure 1-17 shows a MU-MIMO application scenario with one 3x3 MIMO AP and three 1x1 MIMO STAs. To obtain channel information about each STA, the AP sends a sounding frame to each STA. The STAs reply the AP with channel information. The AP uses precoding technology to implement beamforming to generate strong signals in the respective direction to each STA but weak signals in other directions (including directions to other STAs). In this way, the AP ensures good wireless coverage and mitigates interference to other users.

MU-MIMO applies to downlink transmission only and can concurrently transmit data to four users at most. In the uplink, data frames of a single user are transmitted one by one. If lengths of concurrently transmitted frames are different, frame padding is used to adjust the frame lengths. The scheduled BA mechanism is used to schedule ACK responses from each user so that ACK responses are sent one by one.

MU-MIMO increases the number of concurrent users on a single AP, enhancing the concurrent user access capability. In single-stream STA scenarios especially, MU-MIMO improves the downlink throughput of APs significantly. In multi-user transmission, interference between streams limits the usage of higher-order modulation modes, for example, 256QAM.