
One of the key technologies for the development of wireless networks (for example, Wi-Fi) in recent years is the MIMO technology. MIMO is the multiple transmission of information from several transmitters and its reception, as well as processing at multiple receivers. The main objectives of MIMO - to increase the bandwidth of the wireless channel and the quality of communication.
The main method of increasing bandwidth in MIMO systems is multiplexing, that is, the parallel transmission of several streams of information from different antennas (about it below). Special cases of MIMO are transmission systems where a single antenna is used at the receiver or transmitter. Such systems are called Multiple-input single-output (MISO) and Single-input multiple-output (SIMO). They can not organize parallel transmission of multiple streams of information, but you can use additional antennas to improve the quality of receiving or transmitting a signal. In the description of access points of various vendors, we can find out how many transmitting and receiving antennas are on the device, how many MIMO spatial streams it supports. For example, it can be 3x4: 3, which means 3 transmitters, 4 receivers and 3 spatial streams. In addition to these parameters, you can find abbreviations or designations such as MRC, STBC, CSD, 802.11ac Tx BF, etc. All these technologies are also aimed at improving the signal quality. So, let's try to figure out what options of tweaks use modern access points to get your device a good signal. It should be noted that on Habré there are already articles with a rather detailed description of the operation of these technologies -
MIMO ,
OFDM ,
STBC and MRC . In this material, we would like to make a general overview on the technologies of improving the quality of communication, visually show how this or that function works and what kind of gain it gives. Considered to work from the point of view of 802.11 Wi-Fi, although, of course, these methods are used in other wireless standards (LTE, 802.16 WiMAX).
')
Spatial Multiplexing (MIMO SDM)
A key advantage of MIMO is the ability to transmit several independent information flows from different antennas on the same channel. This allows you to dramatically increase the bandwidth of the wireless channel. The technology is called spatial multiplexing, or SDM (Spatial Division Multiplexing). The main condition for MIMO SDM operation is multipath propagation. If we send data from two antennas, with direct visibility, the signal will come to the receiver at the same time, and we will receive their imposition (interference). So we will make only worse. But if during the passage the signal is reflected, refracted, etc., the recipient can recognize (correlate) the incoming signal for different streams. Then, the receiver calculates the current state of the transmission channels (streams) for each of the transmitting antennas based on pre-calibration (by service headers). And then with the help of mathematical transformations, restores the original threads. In the case of MIMO, the sender does not know about the status of the channel, that is, it does not optimize the signal during transmission. The access point and client transmit a certain number of streams supported by two parties. For example, if a client supports only one stream, the access point will also transmit a single stream.
It is worth noting that when transmitting multiple streams (and indeed when transmitting simultaneously from multiple antennas), the total radiated power is divided by the number of transmit antennas. For example, if we transmit a signal simultaneously from two antennas, then the signal power for each of them will be half the maximum. However, in this case, we transmit information on two or more channels at the same time. Also, by sharing SDM and multiple transmission (see below), we can increase the SNR value (signal-to-noise ratio) at the receiver.
MIMO systems continue to evolve and the 802.11ac (wave2) standard implements multiple simultaneous MIMO transmissions by several clients (Multiuser-MIMO). That is, if there are two clients supporting one and two streams, the MU-MIMO system will transmit them a signal simultaneously. As we remember, before the advent of MU-MIMO technology, only one system could transmit data at a time. The technology works only in the direction from the access point to the client (DownLink). Current access points allow you to work with three MU-MIMO clients and transmit up to three streams (in total). MU-MIMO technology requires support on both the access point and client device. It also requires additional calculations on the access point and imposes certain conditions on use. For example, its operation is impossible without prior calibration and adaptive transmission (Explicit Transmit Beamforming), which will be described below.

The development of mechanisms for multiple transmission / reception of course led to an increase in the number of antennas on 802.11n devices. Today, enterprise-level access points (802.11n / ac) have already become standard with 3-4 antennas. In this case, the number of spatial streams is often less than the number of antennas. In fact, are there many customers supporting, for example, 3 threads? Of course, not a lot. If this is a smartphone, then more often one spatial stream is supported. This allows the access point to use various techniques to optimize the reception and transmission of signals using free antennas.
Optimum weight addition (MRC)MRC allows you to improve the SNR value for the incoming signal (from the client to the access point). If there is an additional free receiver (s) on the access point, it adds the signal received on this receiver with the others. Since the receiver already has information on the current state of the transmission channel (for each of the transmitting antennas), it can calculate the signals (on each of the receiving antennas), align them and optimally add them, obtaining the best signal-to-noise ratio. Comparison of results for single and multiple streams with additional antennas and without shows that the MRC in some cases can significantly increase the value of SNR, and therefore increase the transmission speed, the range of the TD. MRC works only on the access point to improve the incoming signal from the client. The technology can be used in conjunction with others - CSD, SDM, STBC.
Diversity Transmission (CSD / SE)
Cyclic Shift Diversity (CSD) technology allows you to transfer copies of a single signal with additional free antennas. This is done alternately with a small interval (200 ns). If you transmit copies of one signal simultaneously from several antennas (the power is divided), you will not be able to receive the gain at the reception. If, however, the signal is transmitted independently (at maximum power) with a small interval from each of the antennas, it is possible to obtain the separation of the signal at the reception, and therefore improve the signal. The receiver in turn, according to a certain criterion, selects the best signal. The diversity transmission method is quite old and not very convenient for recognition on the receiver (requires computational power, does not scale well). However, it is supported on access points and works with previous generation clients — 802.11a / g. Modern standards (802.11n and later) use STBC or adaptive transmission (Beamforming).
Spatio-temporal block coding (STBC)
STBC allows to transmit different signals simultaneously from several antennas at several clock intervals. For transmission, the Alamouti scheme is used. For the simplest case of 2x1, this scheme allows for two time intervals to transmit two signals twice. At two intervals, one of the signals and the complex conjugation of the other signal are transmitted from different antennas. Thus, we get the separation of signals in time and space (two signals pass in different ways), increasing the resulting signal at the reception. From the point of view of reception, the STBC method is quite convenient, since does not require large computing power. As you might guess, STBC does not work simultaneously with CSD. In contrast to the MRC, which we reviewed earlier, STBC allows us to improve the signal quality from the access point to the client. Theoretically, work is supported in higher order modes or for multiple streams (for example, in 2x1 mode for two streams with four transmit antennas). STBC can be used simultaneously with MIMO SDM.
Performance impactSo, we have considered different methods of diversity (multiple) transmission / reception on access points. What is the advantage of using them, what real growth do they give? See charts *. In the first graph for MCS7 (one stream), we see that SE (CSD) does not provide significant improvements compared to the SISO (1x1) mode. STBC behaves much better: for an error rate of 1% (PER - Packet Error Rate) it is ~ 4 dB better than SE. MRC ** gives the greatest gain: almost 10 dB compared to the 1x1 mode! However, at lower speeds, the results are less exciting. For MCS0 (second graph), SNR indicators for STBC and SE (CSD) are generally comparable.
* taken from the book by Eldad Perahia, Robert Stacey. Next Generation Wireless LANs - 802.11n and 802.11ac
** Why MRC is betterIn fact, when comparing 2: 1 and 1: 2 systems for STBC and MRC, respectively, with the same power from both sides, STBC a priori will give a smaller result. This is due to the fact that when transmitting a signal from two antennas, the power is divided by two (-3 dB). When receiving the same access point receives a signal sent at full power. That is, the difference between STBC and MRC in mode 2 of the receiver or transmitter is at least 3 dB.
Adaptive Transmission (802.11ac Explicit Beamforming)
All the methods that we considered before were based on signal processing at the receiving side. That is, when transmitting information, it was the receiver that formed the communication channel matrix for the incoming signal from each of the transmitters. The transmitting side did not adjust the signal on the antennas between themselves, that is, sent the signal "blindly." In adaptive transmission, the main focus is on determining the channel status at the transmitter, in order to send a signal with optimal phase-amplitude characteristics. In other words, send a signal from several antennas so that on the receiving side to get the best quality. This can be done in different ways (without a response from the recipient, calibration with the recipient). In the 802.11ac standard, an approach has been implemented to obtain calibration information from the receiver. That is, the receiver reports how it hears a signal from each antenna of the access point. After that, based on the assumption that the channel is symmetrical in both directions, a transmission matrix with coefficients for a specific receiver is formed. In addition, the use of adaptive transmission allows you to distribute power between different streams (for example, to increase the power for streams by the best SNR). The graph shows that in comparison with the exploded transmission methods discussed earlier, the adaptive transmission mode allows you to get the greatest increase in speed when transmitting information to the client.

We considered various methods of multiple signal transmission in MIMO (Wi-Fi) systems - multiplexing, diversity of the signal at the reception and transmission, adaptive transmission, and also showed how they can increase. In reality, there will be a more complex picture. Additional factors affecting the operation of the wireless network are added (distance to the client, number of clients, load on the channel, client-supported transmission methods, etc.). An access point based on embedded algorithms decides which transmission methods to use at any given time.
References- Eldad Perahia, Robert Stacey. Next Generation Wireless LANs - 802.11n and 802.11ac, 2013
- M. Bakulin, V. Kreyndelin, L. Varukina MIMO Technology. Principles and algorithms, 2014