What does bandwidth 20 40 mean on a router. Three options for high-speed Wi-Fi: hopes and fears. Using optional equipment

This article will be of great interest to owners Ubiquiti M2.

So we bought a couple UBNT M2(it doesn't matter if NanoStation or NanoBrigde). Installed. One was posted as AP, the second - as station, guided them on a signal. Link got up. Now I would like to make the link as stable as possible

The first thing we do is launch Tools->Site Survey from two sides.

Fig.1.

If we see more than two stations in the list, then we perform the following actions: Channel Width tab Wireless set to 20MHz.

The fact is that only 60 MHz is allocated for the 2.4 GHz band. With a channel width of 40 MHz, the station occupies 2/3 of the available range - and itself interferes with everyone, and everyone interferes with it, and no one can work.

The second thing to do is to come up with a load. All changes need to be checked when passing traffic. Without traffic, the station can connect at 130/130 and under load sag up to 26/26. As a load, the built-in speed test is only suitable for showing off to friends - it overestimates the speed too much.

You can use sites on the Internet to check, or run a torrent with a lot of movies to download.

Fig.2. Station operation with a channel width of 40 MHz.

Figure 2 shows an example of a failed setup. At 40 MHz TX/RX Rate should be about 300/300. And our station operates at a speed lower than it is possible to operate even at 20 MHz. The test was made in the morning when the activity of foreign stations is low. The more activity of foreign stations, the worse our speed.

Fig.3. Station operation with a channel width of 20 MHz.

By going to 20 MHz, we lost a little in speed, but we noticeably increased stability. Each step 40->20->10->5 increases the signal by 3dB and reduces the noise level by 3dB.

The next step is to select a frequency. For this purpose, you can look at it for a long time, or you can run it. Foreign stations are a source of noise for us. Noise interferes with reception, but noise does not interfere with transmission. Therefore, you need to choose the frequency at the station for which the download speed is more important. Figure 4 shows that the least used frequencies are about channel 5 and about 12-13 channels.

Fig.4.

There is also a very good option - Channel Shifting tab Wireless. It shifts the frequency grid by 3 MHz. In a noisy air, this allows you to squeeze out a couple of megabits.

In firmware 5.5, it became possible to work at 25 and 30 MHz. The transition to 25 MHz allows you to increase the bandwidth, while not losing much in stability.

Fig.5. Station operation with a channel width of 25 MHz.

Power selection. Work at a power of more than 20 dBm is undesirable. The greater the transmitter power, there is more out-of-band radiation, the more the station clogs the entire range.

It is necessary to ensure that at the input to the receiver it turns out from -60 to -70 dBm. If we have -50, then we need to reduce the power. Working with such a signal is harmful to the receiver.

If it turns out -80, then either you need to use antennas with a large gain, or to achieve line of sight.

AirMax is one of the features of UBNT stations. This is the polling protocol for the development of UBNT. Designed to partially compensate for the shortcomings of the 802.11 a / b / g / n standards when used outdoors. But Ubiquiti obviously over-praised him, because he does not always work well. Therefore, the inclusion of AirMax is an individual matter. In some cases, it allows you to increase the real speed, and in others - lowers it. According to my observations, the maximum speed decreases with clean ether (morning-night), and with clogged ether (evening), the speed increases.

Aggregation.

Can be found on the tab. Advanced Wireless Settings. Quantity frames I would leave 32, but with bytes you can experiment: decreasing increases stability, and increasing increases speed

In a highly noisy air, the decrease bytes increases both speed and stability.

Special for ASP24.

I did not touch on one important point - the use of 40 MHz wide networks in the 2.4 GHz band. Apparently in vain, because rooted in the minds of readers gg opinion (not without efforts on the part of the founding fathers of the resource) categorically does not accept the very idea of ​​​​the possibility of using "wide" networks in the 2.4 GHz range - which is easy to verify by reading the comments under the mentioned article. Today I will try to dot, if not all, then many points over the “i” regarding this issue. And at the same time I will destroy a couple more myths and legends that have developed around the work of Wi-Fi networks (hello to Adam Savage and Jamie Hyneman).

What are the arguments of the opponents of 40 MHz networks based on? On that:

1. there are catastrophically few non-overlapping channels in the 2.4 GHz Wi-Fi band, so the minimum channel width of 20 MHz is our (their) everything;
2. 40 MHz networks create strong interference with other nearby Wi-Fi networks. Horror!

Well, let's debunk the myths one by one.

On the dangers of public opinion

Established public opinion does not necessarily mean that it is automatically correct. After all, this opinion is formed under the influence of certain individuals who formed and defended it. And many of these individuals, to put it mildly, were far from being the smartest. It was thanks to the deep-rooted public opinion that Giordano Bruno burned down, Galileo suffered, Georg Ohm lost his job, and so on. etc. Frankly laughed at the "public" opinion and Albert Einstein. Now I will prove to you that the great physicist was right...

So, in every second, if not in every first article on Wi-Fi networks, we are persistently explained that in the 2.4 GHz band there are only 3 non-overlapping (i.e., not creating strong interference with each other) channels - 1, 6 and 11. What kind of 40 MHz channel width can we talk about in this case, if one "wide" network "eats" b about most of the available radio band?! The opinion about 3 non-overlapping channels is so firmly rooted in the mind of the people that I will not even argue with it. I'll just say that this is a blatant lie. Complete nonsense. Bullshit. Zvezdezh. Call it what you want. If you lean out a little and look out of the public tank, then the reality will turn out to be noticeably better: in the European region, where we also belong, 4 non-overlapping 20 MHz channels are available in the 2.4 GHz Wi-Fi band: 1, 5, 9 and 13 Only this way and nothing else. Only equipment bought directly in the USA and brought to Ukraine, or stitched with American firmware, does not allow working in these bands, but such devices are scanty. Therefore, even within one small cramped room, two independent "wide" 40 MHz Wi-Fi networks can work quite successfully, without interfering with each other at all.

But what about interference with neighboring networks? After all, we are all very worried about the quality of Wi-Fi connections among our neighbors and, in general, for the world of Wi-Fi all over the world!

Misunderstanding

In support of their "theory of the harmfulness of wide networks," 20 MHz apologists sing a tune in chorus about strong interference from a 40 MHz network to neighboring Wi-Fi networks. As a convincing argument, they even cite graphs of programs showing the presence of a mass of some kind of Wi-Fi networks around.

However, the problem is that even people who seem to be well versed in the subject of Wi-Fi have a poor idea of ​​\u200b\u200bwhat exactly these graphs show. What can I say for other users. So, these graphs do not show at all what we are used to seeing on the diagrams comparing the performance of processors or video cards there. But ordinary people interpret what they see in this way. Moreover, it is realistic to be afraid that the 40 MHz network will “drown out” with its “powerful” signal all these weak sprout networks nearby. The problem is not even that 40 MHz channel width has nothing to do with network power at all. The problem is that "Decibel" and "Decl" in the understanding of most of these people mean about the same thing. No, I don't blame them at all. This is fine. But let me try to explain the difference in plain language.

How do decibels differ from other “parrots” that measure the performance of video cards and processors? Decibels help to display the difference between indicators, the value of which differs not by units and tens of values, but by an order of magnitude. For example, a difference in Wi-Fi signal strength of 10 dB means a difference of exactly 10 times, a difference of 20 dB is already 100 times, and 30 dB is a thousand times. It would be very difficult to visually depict the difference in such values ​​on a regular chart in "parrots". After all, the minimum value on the diagram corny runs the risk of being invisible to the "naked eye". So decibels come to the rescue. So, 5 dB is already a difference in signal power by 3.16 times, 1 dB - by 1.26 times. A difference of 1 or 5 dB is, of course, too little, although there are real networks that work quite normally even in such difficult conditions. But the 10-20 dB difference in signal strength that most users usually have (of course, signal strength measurements should be taken near the router or access point, and not on the balcony of a neighboring house) is already quite enough not to catch significant interference from other networks. And at the same time, do not interfere with the normal operation of these other networks, because the signal from our Wi-Fi device, spreading to the area of ​​\u200b\u200bthe other network, weakens proportionally. And it does not matter at all whether the width of the network used will be 20 or 40 MHz. Why do I think that a difference of 10-20 dB is enough?

Everyone interferes here!

I will tell you a terrible secret: non-overlapping Wi-Fi channels in the 2.4 GHz band do not physically exist. Generally. How so? It's just that application diagrams like inSSIDer, Acrylic Wi-Fi Home, Wifi Analyzer and the like don't show us the whole truth...

During operation, the Wi-Fi antenna emits not only a useful signal, but also interference - this is simply what it is supposed to do according to the laws of physics. The radiation power of the antenna is distributed approximately like this (according to Zyxel):

Here, for convenience, 0dB is taken as the zero level of maximum power, but the picture can be quite successfully extrapolated. As you can see, at a signal power of -28 dB from the maximum, even one channel successfully occupies a 40 MHz bandwidth. And at a signal level of more than -40 dB from the maximum, even the most remote channels 1 and 13 quite successfully “intersect”. Is this any significant problem for the operation of Wi-Fi networks? No. At the same time, some gagadget readers did not hesitate to post screenshots showing the difference in signal strength with neighboring networks by at least 30 dB, and at the same time they were absolutely sure that they were right about the impossibility of using “wide” 40 MHz Wi-Fi networks. True, in the end they could not explain the reason for their confidence ...

What for?

What is the whole garden for? What is the practical use of 40 MHz? And why is 20 MHz worse? I answer. On a specific example. With a channel width of 40 MHz, the performance of a wireless Wi-Fi network reaches 13-16 Mb / s, with a width of 20 MHz - only about 7-9 Mb / s. Is it worth sacrificing the speed of a Wi-Fi network for the sake of some ridiculous prejudice? I don't think it's worth it. However, you always have the right to your own opinion, indistinguishable from the public.

P.S. Even if your neighbor has built a powerful network, you can avoid significant interference from it simply by changing the polarization of the antennas of the router or access point, if the antennas allow this. Moreover, if there is strong interference from neighboring networks, many equipment manufacturers rightly recommend reducing the signal strength of your Wi-Fi network in order to improve communication. I will not go into details, but in this way it is simply easier for a router or access point to filter “strong” interference. However, this is a completely different story from the field of physics, which I am not going to write about here.

For many who are just starting their acquaintance with WiFi, the technical parameters of wireless equipment may seem confusing. Especially if the specification is in English, as is the case for MikroTik, Ubiquiti and other vendors.

Let's try to consider some of the most important parameters - what they mean, what they affect, in what cases and which ones you need to pay attention to.

Transmitter Power (Tx Power, Output Power)

Different units of measurement. Some manufacturers indicate power in mW, some are in dBm. Translate dBm to mW and vice versa, without bothering your head with recalculation formulas, you can with the help.

It is worth noting that the relationship between these two representations of power is non-linear. This is easy to see when comparing ready-made values ​​in the correspondence table, which is located on the same page as the above calculator:

• Power increase at 3 dBm gives an increase in mW 2 times.
• Power increase on 10 dBm gives an increase in mW 10 times.
• Power increase at 20 dBm gives an increase in mW 100 times.

That is, by reducing or increasing the power in the settings "only" by 3 dBm, we actually lower or increase it by 2 times.

The bigger, the better? Theoretically, there is a direct relationship - the more power, the better, The further the signal "beats", the greater the bandwidth (the amount of data transmitted). For point-to-point trunk channels with directional antennas raised in open spaces, this works. However, in many other cases, not everything is so straightforward.

• Interference in the city. Turned up to the maximum power can do more harm than help in urban conditions. Too strong a signal, re-reflecting from numerous obstacles, creates a lot of interference, and ultimately negates all the advantages of high power.
• Air pollution. An unreasonably strong signal "clogs" the transmission channel and interferes with other participants in the WiFi traffic.
• Synchronization with low-power devices. It may be necessary to reduce TX Power when connecting with low power devices. For good connection quality, especially for two-way high-capacity traffic such as interactive applications, online games, etc., you need to achieve speed symmetry for incoming and outgoing data. If the difference in signal strength between the transmitting and receiving devices is significant, this will not affect the connection in the best way.

Power should be exactly as much as necessary. Even when it is advised to first reset the power to a minimum and gradually increase, achieving the best signal quality. Wherein be aware of non-linear dependency between the power expressed in dBm and the actual energy power, which we talked about at the beginning of the article.

It is also important to take into account that the range and speed depend not only on power, but also on the gain (gain) of the antenna, receiver sensitivity, etc.

Receiver sensitivity (Sensitivity, Rx Power)

The sensitivity of a WiFi receiver is the minimum level of incoming signal that a device can receive. This value determines how weak signals the receiver can decode (demodulate).

Accordingly, you can select equipment for the conditions in which you want to raise a wireless connection.

"Weak" in this case is not necessarily - "not powerful enough". A weak signal can be either as a result of natural attenuation during transmission over a long distance (the farther from the source, the weaker the signal level), absorption by obstacles, or as a result of a poor (low) signal-to-noise ratio. The latter is important, since a high noise level muffles and distorts the main signal, to the extent that the receiving device cannot "separate" it from the general stream and decrypt it.

Sensitivity (RX Power) is the second important factor affecting the communication range and transmission speed. The greater the absolute value of the sensitivity, the better (eg -60dbm sensitivity is worse than -90dBm).

Why is sensitivity displayed with a minus sign?Sensitivity is defined like power in dBm, but with a minus sign. The reason for this is the definition of dBm as a unit of measurement. This is a relative value, and the starting point for it is 1 mW. 0 dBm = 1 mW. Moreover, the ratios and the scale of these values ​​are arranged in a peculiar way: with an increase in power in mW several times, power in dBm increases for several units(similar to power).

• The power of radio transmitters is greater than 1 mW, so it is expressed in positive terms.
• The sensitivity of radio transmitters, or more precisely, the level of the incoming signal, is always much less than 1 mW, so it is customary to express it in negative values.

Representing the sensitivity in mW is simply inconvenient, since figures such as 0.00000005 mW, for example, will appear there. And when expressing sensitivity in dBm, we see more understandable -73 dbm, -60dBm.

Sensitivity is an ambiguous parameter in the characteristics of access points, routers, etc. (however, like power, in fact). In reality, it depends on the signal transmission speed and in the equipment characteristics it is usually indicated not by one number, but by a whole table:

The screenshot from the specification lists the various WiFi signal transmission parameters (MCS0, MCS1, etc.) and how much signal strength and sensitivity the device shows with them.

Here we run into another question - what do all these abbreviations mean ( MCS0, MCS1, 64-QAM, etc.) in the specifications, and how can we still use them to determine the sensitivity of a point?

What is MCS (Modulation and Coding Scheme)?

MCS stands for "Modulations and Coding Schemes" in English. In everyday life, it is sometimes simply called "modulation", although in relation to MCS is not entirely true.

Signal modulation has been used in radio engineering for quite a long time to coordinate spatial streams between various devices and improve transmission efficiency. Modulation is when a signal with information is superimposed on the carrier frequency, modified in a certain way (encryption, changing the amplitude, phase, etc.).

The result is a modulated signal. Over time, new, more efficient modulation methods are invented.

But the MCS index, which is set by the IEEE standards, means not just signal modulation, but a set of its transmission parameters:

• modulation type,
• information encoding speed,
• the number of spatial streams (antennas) used in the transmission,
• transmission channel width,
• duration of the guard interval.

The result is a certain channel rate obtained during signal transmission, taking into account each of these sets.

For example, if we choose from the above specification the best combination of power (26 dBm) and sensitivity (-96 dBm) is MCS0.

Let's take a look at the correspondence table and see what the transmission parameters of MCS0 are. To put it bluntly, sad parameters:

• 1 antenna (1 spatial stream)
• Transfer rates from 6.5 Mbps on a 20 MHz channel to 15 Mbps on a 40 MHz channel.

That is, the point gives the above power and signal sensitivity only at such low speeds.

When determining sensitivity (and power), we better focus on the MCS indices in the datasheet with more efficient, standard transmission parameters.

For example, in the same specification on Nanobeam, let's take MCS15: power 23 dBm, sensitivity -75 dBm. In the table, this index corresponds to 2 spatial streams (2 antennas) and a speed from 130 Mbps on a 20 MHz channel to 300 Mbps on a 40 MHz channel.

Actually, it is on these parameters (2 antennas, 20 MHz, 130/144.4 Mbps) that Nanobeam works in most cases (MCS15 in the Max Tx Rate field in AirOS is usually set by default).

Thus, the standard, that is, the most commonly used sensitivity: -75 dBm.

However, it should be taken into account that sometimes it is not high speed that is needed, but link stability, or range, in these cases, in the settings, you can change the modulation to MCS0 and other low channel rates.

The MCS index table (or speed table, as it is sometimes called) is also used for reverse search: they calculate what speed can be achieved at a certain power and sensitivity.

Bandwidth (Channel Sizes)

In WiFi, data transmission uses the division of the entire frequency into channels. This allows you to streamline the distribution of radio frequency air between different devices - each equipment can choose a less noisy channel for operation.

Simplified, such a division can be compared with a highway. Imagine what would happen if the entire road was one continuous lane (even one-way) with a stream of cars. But 3-4 lanes already bring a certain order to the movement.

We add and divide. The standard channel width in WiFi is 20 MHz. Starting with 802.11n, the possibility of link aggregation has been proposed and regulated. We take 2 channels at 20 MHz and get 1 at 40 MHz. For what? To increase speed and throughput. Wider bandwidth means more data can be transferred.

Disadvantage of wide channels: more interference and shorter transmission distance.

There is also a reverse modification of channels by manufacturers: a decrease in their width: 5, 10 MHz. Narrow channels give a longer transmission range, but a lower speed.

The modified channel width (reduced or increased) is The width of the line.

What does it affect: on the bandwidth and "range" of the signal, the presence of several bands - on the possibility of fine-tuning these characteristics.

Antenna Gain (Gain)

This is another important parameter that affects the signal range and bandwidth.

website

So we will not repeat ourselves once again, but rather we will note additional functionality that was not there before. Now, when you log in to the web interface for the first time, the Internet Access Setup Wizard starts. The user is prompted to either manually set all the parameters, or simply select the city and the name of the provider, and then enter the credentials, if any, are required to connect. The lists of cities and providers are not very large yet.

For NETGEAR Centria WNDR4700, an additional set of Genie utilities is still available for quick access to some settings, remote media playback, parental controls, wireless printing from iOS devices, and so on. Of the innovations, it is worth noting the function of generating a QR code for quick connection to a wireless network of mobile clients. Utilities are available for Windows, Mac OS X, Android and iOS. The ReadySHARE feature set is also supported for accessing data on drives and a printer / MFP connected to the router. They also include a built-in DLNA server and Time Machine support. There is another "cloud" service called ReadySHARE Cloud, which also provides remote access to files on drives. Moreover, mobile versions of the software are paid and, judging by the reviews, are far from ideal.

As for the possibility of the NAS component, then, in general, everything is standard. You can open network access to any folders or partitions on the HDD, add users, specify which users will have access to certain directories, and so on. Access to files from the local network is possible via SMB, HTTP and FTP, and from the external network - only via HTTPS and FTP. In advanced settings, you can format the internal hard drive and view its S.M.A.R.T. data. By default, an EXT4 file system is created, but the router will also cope with a drive that has FAT16 / 32, NTFS, EXT2 / 3/4 or HFS + partitions. The maximum supported volume in the latest firmware is 3 TB. To clear our conscience, we tried to shove a 4 TB drive into the router, but something went wrong with formatting, so we had to stop at a 2 TB drive.

And here is another feature, moreover, the most obvious, equipped only with NETGEAR Centria devices. It's about backup. Owners of Mac OS X, as already mentioned, can set up Time Machine to work with a router. In Windows 7, archiving over the network is only available for the Professional or Ultimate editions. To correct this omission, the ReadySHARE Vault utility is suitable. This program, which is installed on Windows machines, can make backups to the hard drive in the router. By default, she herself decides which files and how often to copy. The user, of course, himself can select files and folders for backup, as well as specify the frequency of backups or set a schedule. Optionally, backups can be protected with a password.

ReadySHARE Vault integrates well into the system. Items appear in the context menu for quickly deleting or adding objects to the list of reserved objects. From there, a dialog is also called showing the versions of objects, with the ability to quickly roll back to previous revisions of the file. Icons appear on the icons of files and folders, signaling the current state of the backup of these objects. And in the root, a pseudo-directory with a timeline with marks about backups is added. Here you can quickly select the desired version of saved files and folders and immediately restore them. In addition, there is a small utility to search for files by name among all the backups made for later opening or restoring. In general, a good replacement for Time Machine, albeit not so beautiful.

The hardware stuffing NETGEAR Centria is different from what we are used to seeing in top-end routers, which are often built on the basis of Broadcom products. In this case, the heart of the device is a RISC processor, or rather SoC AMCC APM82181 with a frequency of 1 GHz and a bunch of "body kit". We already met him in another NAS - WD My Book Live Duo . Atheros radio modules: AR9380 and AR9580. Separate internal antennas are connected to each module according to the 3T3R scheme. Gigabit switch from the same manufacturer - AR8327N. Built-in memory for firmware has 128 MB, and RAM - twice as much. Not bad? Oh yes, the technical characteristics are very good, but it's all the more disappointing that the firmware does not reveal the full potential of the hardware platform. What does it cost, for example, to add support for IP cameras, a download manager, or some kind of web server? Okay, let's not talk about sad things.

NETGEAR Centria WNDR4700
Network standards	IEEE 802.11a, IEEE 802.11b, IEEE 802.11g, IEEE 802.11n, IPv4, IPv6
WiFi speed	802.11a: 6.9, 12, 18, 24, 36, 48, 54Mbps 802.11b: 1, 2, 5.5, 11 Mbps 802.11g: 6.9, 12, 18, 24, 36, 48, 54Mbps 802.11n: up to 450 Mbps
Chipset/Controllers	AMCC APM82181 (1GHz) + Atheros AR9380 + Atheros AR9580 + Atheros AR8327N
Memory	128MB NAND/256MB DDR2 SDRAM
Frequencies	2.4-2.4835GHz /5.1-5.8GHz
Security	WPA2-PSK (AES), WPA-PSK (TKIP), WPS
Firewall	SPI, DoS Protection, URL/Network Services Filter
Network Services	UPnP, DLNA, DHCP, DDNS, Port Triger, Virtual Server, DMZ, Port Forwarding/Translation
WAN connection	Automatic IP, Static IP, PPPoE, PPTP, L2TP; IGMP Proxy
Guest network	1x2.4 GHz, 1x5 GHz
QoS	WMM, IP/MAC/Port Rules, Traffic Priority, LAN Port/Wi-Fi Priority
Print server	Yes, AirPrint
File Server	Samba, FTP, HTTP
Statistics, traffic monitor	Yes, notifications
Connectors and ports	4 x RJ45 10/100/1000 BaseT LAN+ 1 x RJ45 10/100/1000 BaseT WAN (802.3, MDI-X), USB 3.0x2
Drives	1 x 3.5″ SATA HDD, SD/MMC/MS/MS Pro card reader
Buttons	WPS, Wi-Fi, factory reset, power, SD card backup
Indicators	Power, WAN, Wi-Fi, USB, HDD
Power adapter	Input AC 110-240V 50-60Hz, Output DC 12V 5A
Dimensions, mm	256x206x85
Weight, g	870
Warranty	2 years
Price	9 000 rubles

The test bench configurations are the same as before. Desktop PC: Intel Core i7-2600K, 12 GB RAM, Killer NIC E2200 (LAN1). And K42JC with this filling: Intel Pentium 6100, 6 GB RAM, JMicron JMC250 (LAN2). Native adapters, NETGEAR's - WNDA4100 (WLAN1) and WNDA3100 (WLAN2). Synthetic tests were run on Ixia IxChariot 6.7 with the High Performance Throughput profile (see table below) and iperf 1.7.0. For Wi-Fi, WPA2-AES encryption was enabled, automatic channel selection was specified, and the speed was set to 450 Mbps for both bands. Other settings are left by default.

NETGEAR WNDR4700 Router
streams	1	2	4	8	16	32	64
Average speed Wi-Fi 802.11n 5 GHz, Mbps
WLAN1 → WLAN2	83	90	91	90	89	85	79
WLAN2 → WLAN1	87	92	92	92	90	85	73
WLAN1 ↔ WLAN2	91	93	93	91	89	83	73
LAN1 → WLAN2	214	282	276	289	262	247	229
WLAN2→LAN1	145	200	216	217	212	215	202
LAN1 ↔ WLAN2	225	238	233	227	224	220	206
Average speed Wi-Fi 802.11n 2.4 GHz, Mbps
WLAN1 → WLAN2	10	14	13	14	16	22	17
WLAN2 → WLAN1	15	17	19	14	17	11	16
WLAN1 ↔ WLAN2	17	18	18	18	16	10	-
LAN1 → WLAN2	61	59	74	68	66	69	58
WLAN2→LAN1	60	49	47	44	52	47	33
LAN1 ↔ WLAN2	60	62	50	40	44	38	31
Average LAN speed, Mbps
LAN1 → LAN2	890	923	921	915	905	901	821
LAN2 → LAN1	730	946	948	946	948	947	945
LAN1 ↔ LAN2 (½)	605	800	796	797	734	728	746

In the pristine 5 GHz band, the WNDR4700 showed excellent data transfer rates, but at 2.4 GHz, everything is not so rosy. And if the LAN ↔ WLAN route is still more or less decent, then working exclusively in the wireless segment does not cause joy - the connection speed was floating all the time, and one of the tests never completed successfully. But there is one nuance, even two. Firstly, the air is somewhat “polluted”, so the router automatically resets the channel width to 20 MHz, and the second adapter only supports 300 Mbps. Secondly, in the advanced Wi-Fi settings there is an interesting checkbox "Enable frequency sharing 20/40 MHz". Initially, it is enabled, as it should be in accordance with the rules of the Wi-Fi Alliance. If you remove it, the device will ignore the presence of neighboring networks and work to its fullest. At least try to work. It seems that NETGEAR was one of the last to give up and added this option to their routers.

It didn’t work out with WAN connections either, although everything is relative here. The direct connection speed was around 360 Mbps. This is where NAT hardware acceleration would come in handy. We managed to squeeze out about 111 Mbps via PPPoE, and the connection speed via L2TP and PPTP did not overcome the mark of 70 Mbps. The results by today's standards are not the best, although acceptable. Perhaps the firmware is to blame (we had version 1.0.0.50). In any case, somehow I can not believe that this is a hardware problem. After all, the router copes with the duties of a NAS, and not bad at that. For the test, a Hitachi Deskstar 7K3000 HDD was taken and formatted from the device's web interface (EXT4), and an external drive Apacer Share Steno AC202 (NTFS) was connected to the USB port. Then both drives were mounted as network drives in Windows 7 x64, and CrystalDiskMark 3.0.2 x64 was “baited” on them.

⇡ Conclusions

If we consider the functionality of the NETGEAR WNDR4700 from a "router" point of view, then everything is in order here. It is a pity, of course, that developers refuse to load powerful hardware to the fullest, deliberately not adding any goodies such as a download manager, a web server, and other little joys of a pirate or geek. I would also like to have higher WAN connection speeds and more stable Wi-Fi operation in the 2.4 GHz band. There are no complaints about the design of the case, just do not forget about its soiledness and the unceasing cooler.

As for the functions of NAS, then the issue is moot. If all you need is a simple NAS with the ability to backup and play files over the network, then yes, the WNDR4700 will suit you. Moreover, the data exchange speeds are quite decent. Again, not even the most expensive "dedicated" NAS usually has more functionality. In general, in the question "To take or not to take NETGEAR Centria?" everything is very individual. All the same, the price for it in Russia is not so small - about 8,500-9,000 rubles. There is also a NETGEAR WNDR4720 model, which already has a 2TB hard drive installed, but it has not yet reached our open spaces.

The implementation of Wi-Fi 802.11n in modern phones and tablets leaves a lot to be desired. The new standards 802.11ac and 802.11ad promise gigabit speeds in the future and have been discussed for years. Broadcom and other companies have been offering corresponding chipsets to manufacturers since mid-2012. When will they begin to be implemented and which devices will be the first to receive support for high-speed versions of Wi-Fi?

Tricks in implementing 802.11n

The history of the transition to new standards is repeated surprisingly accurately. One of the first smartphones in Russia to support the draft version of 802.11n was the HTC HD2, which appeared in 2009. Its speed was only slightly faster than that of smartphones with Wi-Fi version "g". It corresponded to the minimum implementation of version "n" and made you smile bitterly, remembering the promised 600 Mbps. Years have passed, the final version of the standard has long been approved, but everything remains the same.

Until now, most mobile devices support the 802.11n standard in its minimum version. One 20 MHz wide channel at 2.4 GHz - and that's it. This limits the theoretical speed limit to 72 Mbps. In real conditions, the actually demonstrated speeds turn out to be even less.

Real Wi-Fi connection speed (image: anandtech.com)

Please note: the version "g" and even "a" looks in practice quite competitive compared to the cut-down versions of Wi-Fi "n". Marketers, on the other hand, like to make references to the upper threshold of the standard - the notorious 600 Mbps. They could be achieved using four 40 MHz wide channels at 5 GHz, but this option is rarely found even in routers. Most mobile devices use one or two transceivers, each with its own antenna. Only in single laptops (for example, MacBook Pro) you can find three. Accordingly, the maximum speed is 3 x 150 = 450 Mbps. I think there is not a single smartphone or tablet in the world with three or four antennas.

Real Wi-Fi Data Rate - Continued (Image: anandtech.com)

More recently, some smartphone models began to support speeds of 150 Mbps. At MWC 2013, there was the Huawei Ascend P2, a mid-range smartphone with two Wi-Fi antennas, which was touted as an advantage. A little earlier, Ascend Mate was presented in a similar way. However, in addition to doubling the narrow channels, you can increase the width of the only one to 40 MHz, and the result will be the same - 150 Mbps.

It is noteworthy that the Wi-Fi speed does not depend on the price of the device. Not only the iPhone 5 and Huawei Ascend Mate, but also the budget Philips W626 can work on Wi-Fi “n” twice as fast as most others. The problem is that manufacturers usually do not indicate the features of a particular model. In the specifications everywhere they write "802.11 b / g / n" without any clarification.

"ad" version as a competitor to Bluetooth

With Wi-Fi of the following standards, the situation is even more interesting. Contrary to the designation, 802.11ad (WiGig) will not be the successor to 802.11ac. This parallel evolving standard is built from the ground up and is likely to replace Bluetooth soon. Its task is high-speed wireless communication over short distances. The table below shows some implementation details and theoretical speed limits for different Wi-Fi versions when using a single channel.

Tentatively, the 802.11ad standard will be limited to speeds up to 7 Gb / s, but the possibility of further increasing it is also being considered. Due to the characteristics of the propagation of a high-frequency signal, the devices must be in line of sight and within a few meters from each other. Unlike 802.11ac, WiGig is not backwards compatible with other versions of Wi-Fi because it operates at 60 GHz.

"ac" Version - Expectations and Concerns

Version "n" by the middle of the year will begin to displace 802.11ac. It has been developed since 2008 and the last draft version was announced only five years later. Now the readiness of the standard is estimated at 95%, whatever that means. Without waiting for the final official approval, manufacturers began to produce the corresponding microcircuits a year ago. Practice has shown that this approach was more than justified in the case of version "n". The hardware platform has not been modified, and software changes can easily be made by releasing a firmware update. One of the first modules to work according to the 802.11ac standard (backwards compatible with b/g/n) was released by TriQuint. The TQP6M0917 chip, which appeared in mid-2012, has dimensions of 4 x 4 x 0.5 mm, which allows it to be used in mobile technology.

Another major communications chipset company (Broadcom) believes that the first 802.11ac-enabled devices will be mass produced by the second half of 2013. Qualcomm agrees. Traditionally, routers and network adapters will come first. Smartphones and tablets with 802.11ac will become familiar a little later, but some of their representatives will go on sale in the very near future.

High-speed fifth-generation Wi-Fi is expected in the iPhone 5S (symbolically) and all smartphones based on the Qualcomm Snapdragon 800 platform. By analogy with the history of the implementation of the “n” version, most likely we are talking about the basic implementation and single-channel solutions. Depending on the channel width (from 80 to 160 MHz), the speed of new smartphones via Wi-Fi will be limited to a theoretical limit of 433 or 866 Mbps.

Smartphones with Broadcom BCM4335, Redpine Signals RS9117 and Qualcomm Atheros WCN3680 chips will connect at 433 Mbps. Higher speeds have so far been announced only in chips for laptops and routers.

Backward compatibility leaves another loophole for unscrupulous marketing. A device that supports draft 802.11ac can use the now-familiar channel widths of 20 and 40 MHz. With such a formal implementation, the speed bar will drop below the minimum 433 Mbps.

Among other important features of the standard, the technique for improving the quality of communication Beamforming is noted. It allows you to take into account the phase difference of the re-reflected signals and compensate for the resulting speed losses. Unfortunately, Beamforming involves the use of multiple antennas, which still limits its scope to laptops.

It is expected that in a number of use cases, the new standard will increase battery life. By transferring the same amount of data faster, the chip will be able to go into low power mode earlier.

As you can see from the presented examples, technically nothing prevents you from increasing the speed of data transfer over Wi-Fi right now. This does not require the introduction of new standards - the potential of the existing version of "n" in mobile devices is not even half revealed. If speed is critical for you, try testing your smartphone or tablet by connecting it to a decent router.

A WiFi connection may not always provide the same speed as a cable connection. Among the main reasons are incorrect router settings, conflicts with neighbors' access points, and the wrong choice of the location of the router. Speed ​​is also cut when using outdated hardware or older firmware versions.

How to determine that WiFi speed is being cut

Internet providers indicate in the contract the maximum possible access speed. The actual width of the throughput channel is usually lower than the declared one. At home, it is easy to check whether this is due to restrictions on the provider's side or using WiFi. To do this, connect an Ethernet cable directly to the device from which you access the Internet.

Open the Speedtest online service in any browser and click "Begin Test". The site will automatically determine the nearest server through which the speed test will be performed. The computer will exchange data with the selected server to find out the current Internet speed. Wait until the end of the operation, then remember or write down its result.

Then connect an internet cable to your router, turn it on and connect to WiFi from the same device you tested the speed on. Open the site again and repeat the measurement. If the results of the first and second tests differ significantly, the speed is cut precisely because of the use of wireless Internet.

Interference from neighbors' wireless equipment

Most often, this reason manifests itself in apartment buildings with a large number of installed WiFi access points. The wireless network can operate in one of two bands: 2.4 or 5 GHz. The first option is more common. In this case, the actual frequency can be from 2.412 to 2.484 GHz in 0.005 GHz steps, depending on the selected channel.

The 2.4 GHz band is divided into 14 segments, but not all of them may be available for legal use in a particular country. For example, in the USA only channels 1-11 are used, in Russia: 1-13, in Japan: 1-14. Selecting an incorrect value may violate the laws of the state in which the equipment is operated.

If your neighbors' access points use the same channel as your router, interference (radio wave overlay) occurs. As a result, the speed of the Internet over WiFi is cut. It is recommended to analyze the current frequency congestion. The most popular software tool used for this purpose is the inSSIDer utility developed by MetaGeek.

Install the program, run the executable file and click the "Start Scan" button in the upper left corner of the program window. The graph on the right will display the found WiFi networks and the channels on which they work. Find the range that contains the least number of networks with a high level of reception, and then select it in the control panel of the router.

Note! The width of each channel can be 20 or 40 MHz. Only channels 1, 6, and 11 do not overlap. Use one of these values ​​for the best network setup. You can also choose to automatically detect the least loaded frequencies in the router settings.

High band occupancy

In large cities, the number of available 2.4 GHz networks can be so high that changing the WiFi channel does not lead to the desired result. The data rate is cut even after selecting the freest segment of the frequency range. The optimal solution to this problem is the transition to the 5 GHz band, which has not yet received sufficient distribution.

Its use is possible on dual-band routers. Such routers create two networks at once, which have different names, encryption and authorization parameters. Client devices whose radio module supports 5 GHz operation will be able to connect to WiFi in this range. Legacy models will connect to the second network. With this scheme of work, a number of disadvantages should be taken into account, the main of which are:

1. Smaller coverage area in the presence of obstacles, due to the physical properties of radio waves of this length.
2. Lack of compatibility with older devices.
3. The high cost of dual-band equipment.

Router Issues

The main mistake made by users when organizing a home WiFi network is the wrong choice of the location of the router. It leads to poor signal reception on client devices, due to which the Internet speed is cut. You can specify the signal level by the number of marks on the WiFi icon located in the tray (lower right corner) of the Windows operating system. On mobile devices, the Internet connection status and signal strength can be checked at the top of the screen, in the notification bar.

It is recommended to install the router in the central room of the room where it will be used. This arrangement ensures a high level of WiFi reception in all rooms of the apartment or office. When installed in the corner of a room, remote rooms will not be able to connect to the wireless network or will receive Internet at low speed.

Important! The quality of communication with the router is also affected by the power of the transmitter, the number of installed antennas and the distance from working sources of electromagnetic radiation. To prevent Internet speed from being cut, try to install the router away from microwave ovens, refrigerators and other household appliances.

Also check the correctness of the WiFi mode selection in the router settings. It is responsible for the maximum data transfer rate and backward compatibility with older devices. For example, if "11b only" is selected, the WiFi speed will be cut to 11Mbps, while "11g only" will limit the bandwidth to 54Mbps.

You can enter the web interface of the router at the address indicated on its bottom panel. For TP-Link models, the desired parameters are selected in the “Wireless Mode -> Wireless Settings” section. The recommended values ​​if there are older models on the network are "11bgn mixed" and "11bg mixed". If all home or office devices support the 802.11n standard, check the "11n only" box.

In the Wireless Security menu, set the security type to WPA/WPA2, as using the legacy WEP method cuts WiFi speed. Change the automatic encryption type selection to Advanced Encryption Standard (AES). It provides greater network security with less impact on data transfer rates.

Click the tab with advanced wireless network settings. On TP-Link it is "Wireless Mode -> Advanced Settings". Find and activate the "WiFi Multimedia" (WMM) setting. This protocol allows you to set a high priority for multimedia traffic, thereby speeding up its transmission.

In the settings of the connected devices, you must also activate this function. Open Device Manager in the Control Panel of your Windows operating system. Find your network adapter and go to its properties. On the "Advanced" tab, select the line "WMM" in the list on the left. On the right, specify the value "Enabled" or "Enabled". Save the configuration by clicking the "OK" button.

Another parameter that you should pay attention to when configuring the router is the transmitter power or "Tx Power". This value is indicated as a percentage of the maximum power of the equipment. If the hotspot is far away, set it to "100%" to improve WiFi reception.

Outdated device firmware

Manufacturers of routers and other wireless devices regularly optimize their software for maximum performance. You can download the new firmware version on the Internet, on the developer's website. The update is performed by downloading the file to the device through the admin panel. The path in the menu of routers of different brands is different:

• TP-Link: "System Tools -> Firmware Update";
• D-Link: "System -> Software Update";
• ASUS: "Administration -> Firmware Update";
• Zyxel: "System Information -> Updates";

Advice! When installing software, consider the hardware version of the router. It is indicated on the sticker or in the documentation for the device.

On client equipment (laptops, computers and other equipment connected to WiFi), you should check the versions of network drivers. Windows OS allows you to update the firmware through the control panel, in the "Device Manager" section. Open the "Network adapters" tab and select the radio module you are using. In the "Driver" section, click "Update" and select to automatically search for software on the Internet. After that, restart your computer and connect to the wireless Internet again.

Tutorial video: How and why the Internet speed is cut over WiFi

Using optional equipment

If, after fixing all problems, the Internet speed in remote rooms continues to cut, use additional equipment to amplify the signal. It includes: external antennas for routers, high-powered wireless adapters for computers, WiFi repeaters.

When choosing an antenna, consider the gain and type of connector with which it connects to the access point. Typically, manufacturers indicate a list of equipment recommended for use with certain device models. When connecting third-party antennas that have not been tested for compatibility, difficulties may arise with further warranty service.

The repeater allows you to increase coverage and get high speed Internet even at a considerable distance from the router. Thanks to the use of a built-in power supply, such devices have a compact size. To use them, just plug the device into a power outlet and press the WiFi Protected Setup (WPS) button on the case. After that, the same button must be pressed on the router itself or a quick connection must be activated via the web interface.

I think I won’t be mistaken if most of us have an Internet connection like this: there is some fairly high-speed wired channel to the apartment (now gigabit is not uncommon), and in the apartment it is met by a router that distributes this Internet to clients, giving them "black" ip and performing address translation.

Quite often, a strange situation is observed: with a high-speed wire, a very narrow wifi channel is heard from the router, which does not load even half of the wire. At the same time, although formally Wi-Fi, especially in its ac version, supports some huge speeds, when checking, it turns out that either Wi-Fi connects at a lower speed, or connects, but does not give out speed in practice, or loses packets, or all together.

At some point, I also encountered a similar problem, and decided to set up my Wi-Fi in a human way. Surprisingly, it took about 40 times longer than I expected. In addition, it somehow happened that all the Wi-Fi setup instructions that I found converged to one of two types: the first suggested putting the router higher and straightening the antenna, while reading the second, I lacked an honest understanding of spatial multiplexing algorithms .

Actually, this note is an attempt to fill a gap in the instructions. I will say right away that the task has not been fully resolved, despite decent progress, the connection stability could still be better, so I would be glad to hear the comments of colleagues on the topic described.

Chapter 1:

So the problem statement

The Wifi router offered by the provider has ceased to cope with its duties: there are long (30 seconds or more) periods when the ping to the access point does not pass, very long (about an hour) periods are observed when the ping to the access point reaches 3500 ms, there are long periods when the connection speed with the access point does not exceed 200 kbps.

Scanning the range using the inSSIDer windows utility produces the picture presented at the beginning of the article. There are 44 Wifi SSIDs in the 2.4 GHz band and one network in the 5.2 GHz band in the district.

Solution tools

Celeron 430 self-assembly computer, 2b Ram, SSD, fanless, two wireless network cards on a Ralink rt2800pci chip, Slackware Linux 14.2, Hostapd from Git as of September 2016.

Assembling the router is beyond the scope of this post, although I note that the Celeron 430 performs well in fanless mode. I note that the current configuration is the latest, but not final. Perhaps there are still improvements to be made.

Decision

In fact, the solution would, in a good way, be to run hostapd with minimal configuration changes. However, experience confirmed the truth of the saying “it was smooth on paper, but forgot about the ravines” so well that it took the writing of this article to systematize knowledge about all the non-obvious details. Also, I initially would like to avoid low-level details for the sake of harmony of presentation, but it turned out that this is impossible.

Chapter 2

A bit of theory

Frequencies

Wi-Fi is a standard for wireless networks. From the point of view of OSI L2, the access point implements a switch type hub, but most often it is also combined with an OSI L3 switch of the “router” type, which leads to a fair amount of confusion.

We will be most interested in the OSI L1 level, that is, in fact, the environment in which the packets go.

Wi-Fi is a radio system. As you know, a radio system consists of a receiver and a transmitter. In Wi-Fi, the access point and the client device perform both roles in turn.

The Wi-Fi transmitter operates on a certain frequency. These frequencies are numbered, and each number corresponds to a certain frequency. Important: despite the fact that for any integer there is a theoretical correspondence to this number of a certain frequency, Wi-Fi can only work in limited frequency bands (there are three of them, 2.4 GHz, 5.2 GHz, 5.7 GHz), and only on some of the numbers.

A complete list of correspondences can be found in Wikipedia, but it is important for us that when setting up an access point, you need to specify which channel the carrier frequency of our signal will be on.

An obscure detail: not all Wi-Fi standards support all frequencies.

There are two Wi-Fi standards: a and b. "a" is older and operates in the 5GHz band, "b" is newer and operates in the 2.4GHz band. At the same time, b is slower (11 mbit instead of 54 mbit, that is, 1.2 megabytes per second instead of 7 megabytes per second), and the 2.4 GHz band already accommodates fewer stations. Why this is so is a mystery. It is doubly a mystery why there are practically no standard access points in nature.

(Image borrowed from Wikipedia.)

(Actually, I'm being a little disingenuous, because a also supports the 3.7 GHz frequency band. However, I haven't seen a single device that knows anything about this band.)

Wait, you ask, but there are also 802.11g, n, ac - standards, and they seem to just beat the unfortunate a and b in speed.

But no, I will answer you. The g standard is a belated attempt to bring speed b to speed a, in the 2.4 GHz band. But why, you answer me, did you even remember about b? The answer is because even though the ranges of both b and g are called 2.4, they are actually slightly different, and the range of b is one channel longer.

The standards n and ac have nothing to do with ranges at all - they regulate the speed, and nothing more. Standard point n can be either "in the base" a (and operate at 5 GHz), or "in the base" b and operate at 2.4 GHz. I don’t know about the ac standard point, because I haven’t seen it.

That is, when you buy an access point n, you need to look very carefully at what ranges this n works in.

It is important that at one time one Wi-Fi chip can only work in one range. If your access point claims that it can work in two at the same time, as, for example, free routers from popular providers Virgin or British Telecom do, then it actually has two chips.

Channel Width

Actually, I have to apologize because I said earlier that one range is longer than another without explaining what "longer" is. Generally speaking, not only the carrier frequency is important for signal transmission, but also the width of the coded stream. Width - this is what frequencies above and below the carrier the existing signal can climb. Usually (and fortunately in Wi-Fi), the channels are symmetrical, centered on the carrier.

So in Wi-Fi there can be channels with a width of 10, 20, 22, 40, 80 and 160 MHz. At the same time, I have never seen access points with a channel width of 10 MHz.

So, one of the most amazing properties of Wi-Fi is that despite the fact that the channels are numbered, they intersect. And not only with neighbors, but even with channels through 3 from yourself. In other words, in the 2.4 GHz band, only access points operating on channels 1, 6, and 11 do not intersect with 20 MHz streams. In other words, only three access points can work side by side so as not to interfere with each other.

What is an access point with a channel width of 40 MHz? The answer is - and this is an access point that occupies two channels (non-overlapping).

Question: and how many channels with a width of 80 and 160 MHz fit in the 2.4 GHz band?

Answer: No one.

The question is, what affects the width of the channel? I do not know the exact answer to this question, I could not check it.

I know that if the network intersects with other networks, the connection stability will be worse. Channel width of 40 MHz gives more crossovers and worse connection. According to the standard, if there are other working access points around, the 40 MHz mode should not be turned on.

Is it true that twice the channel width gives twice the bandwidth?
It seems to be, but it is impossible to verify.

Question: If my access point has three antennas, is it true that it can create three spatial streams and triple the connection speed?

Answer: unknown. It may turn out that out of the three antennas, two can only send, but not receive packets. And the signal speed will be asymmetrical.

Question: So how many megabits does one antenna give?

Answer: You can see here en.wikipedia.org/wiki/IEEE_802.11n-2009#Data_rates
The list is strange and non-linear.

Obviously, the most important parameter is the MCS index, which determines the speed.

Question: Where do these strange speeds come from?

Answer: There is such a thing as HT Capabilities. These are optional chips that can slightly correct the signal. Chips are very useful: SHORT-GI adds a little speed, about 20 Mbps, LDPC, RX STBC, TX STBC add stability (that is, they should reduce ping and packet loss). However, your hardware may simply not support them and still be quite “honest” 802.11n.

Signal strength

The easiest way to deal with poor communications is to fry more power into the transmitter. Wi-Fi has a transmission power of up to 30 dBm.

Chapter 3

The solution of the problem

From all of the above vinaigrette, it would seem that the following conclusion can be drawn: Wi-Fi can implement two “modes” of functioning. “Improving speed” and “Improving quality”.

The first, it would seem, should say: take the most unoccupied channel, channel width 40 MHz, more antennas (preferably 4), and add more Capabilities.

Second - remove everything except the basic n-mode, turn on more power, and turn on those Capabilities that add stability.

Recalling once again the proverb about ravines, we will describe what kind of uneven terrain awaits us when trying to implement plans 1 and 2.

Ravine zero

Although Ralink rt2x00 family chipsets are the most popular chipsets supporting the n standard and are found both in high-end (Cisco) and low-end (TRENDNET) cards, and moreover, they look exactly the same in lspci, they can have radically different functionality, in particular, support only the 2.4 band, only the 5GHz band, or support incomprehensibly limited parts of both bands. What is the difference is a mystery. It's also a mystery why a card with three antennas only supports Rx STBC in two streams. And why don't they both support LDPC.

First ravine

There are only three non-overlapping channels in the 2.4 band. We have already spoken on this topic and I will not repeat myself.

Second ravine

Not all channels allow you to increase the channel width to 40 MHz, moreover, what channel width the card agrees to depends on the card chipset, card manufacturer, processor load and weather on Mars.

The third and largest ravine

Regulatory domain

If the fact that the Wi-Fi standards themselves are a noble vinaigrette was not enough for you to be happy, then rejoice that every country in the world seeks to infringe and restrict Wi-Fi in all sorts of different ways. In the UK, things are still not so bad, unlike, say, the USA, where the Wi-Fi spectrum is regulated to the point of impossibility.

So, the regulatory domain may require restrictions on the power of the transmitter, on the ability to launch an access point on the channel, on acceptable modulation technologies on the channel, and also require some “spectrum pacification” technologies, such as DFS(dynamic frequency selection), radar detection (which each regdomain has its own, say, in the Americas almost everywhere offered by the FCC, in Europe it’s different, ETSI), or auto-bw (I don’t know what it is). At the same time, with many of them, the access point does not start.

Many regulatory domains simply ban certain frequencies altogether.

You can set the regulatory domain with the command:

Iw reg set NAME

The regulatory domain can be omitted, but then the system will be guided by the union of all restrictions, that is, the worst possible option.

Fortunately, firstly, data on regulatory domains is publicly available on the core website:

And you can search for them. In principle, it is probably possible to patch the kernel so that it ignores the regulatory domain, but this requires rebuilding the kernel, or at least the crda regulatory daemon.

Fortunately, the iw phy info command displays all the capabilities of our device, taking into account (!) the regulatory domain.

So, how do we fix the state of our Wi-Fi?

First, let's find a country in which Channel 13 is not banned. A path of at least half the frequency will be empty. Well, there are quite a few such countries, although some, while not prohibiting it in principle, however, prohibit either the high speed mode n on it, or the creation of an access point in general.

But one channel 13 is not enough for us - because we want a larger signal-to-noise ratio, which means we want to launch a point with a signal strength of 30. We are looking for-looking in CRDA, (2402 - 2482 @ 40), (30) 13 channel, width 40 MHz, signal strength 30. There is such a country, New Zealand.

But what is it, at 5 GHz, DFS is required. In general, this is theoretically a supported configuration, but for some reason it does not work.

An optional task that can be completed by people with advanced social skills:

Gather signatures / movement in support of accelerated relicensing of Wi-Fi bands in the ITU (well, or at least in your country) in general towards expansion. This is quite real, some deputies (and candidates for deputies), thirsting for political points, will be happy to help you.

This is ravine number 4

The access point may not start with DFS, without explanation. So, which regulatory domain should we choose?

There is one! The freest country in the world, Venezuela. Its regulatory domain is VE.

A full 13 channels of the 2.4 band, with a power of 30 dBm, and a relatively relaxed 5 GHz band.

Asterisk challenge. If you have a complete disaster in your apartment, even worse than mine, there is a separate, bonus level for you.

Regulatory domain "JP", Japan, allows you to do a unique thing: run an access point on the mythical channel 14. True, only in mode b. (Remember, I said that there are still small differences between b and g?) So if everything is really bad for you, then channel 14 can be a salvation. But then again, it is physically supported by a few client devices and access points. Yes, and the maximum speed of 11 Mbps is somewhat discouraging.

Copy /etc/hostapd/hostapd.conf into two files, hostapd.conf.trendnet24 and hostapd.conf.cisco57

We trivially edit /etc/rc.d/rc.hostapd to run two copies of hostapd.

In the first one, we indicate channel 13. True, we indicate the signal width of 20 MHz (capability 40-INTOLERANT), because, firstly, this way we will be theoretically more stable, and secondly, “law-abiding” access points simply will not start at 40 MHz from -because of the clogged range. Set capability TX-STBC, RX-STBC12. We cry that capabilities LDPC, RX-STBC123 are not supported, and SHORT-GI-40 and SHORT-GI-20, although they are supported and slightly improve speed, but also slightly reduce stability, which means we remove them.

True, for amateurs, you can patch hostapd so that the force_ht40 option appears, but in my case it makes no sense.

If you are in a strange situation when access points turn on and off, then for special gourmets you can rebuild hostapd with the ACS_SURVEY option, and then the point itself will first scan the range and select the least “noisy” channel. Moreover, in theory, it should even be able to move at will from one channel to another. However, this option did not help me, alas :-(.

So, our two points in one case are ready, we start the service:

/etc/rc.d/rc.hostapd start

The points are starting successfully, but ...

But the one that works on the 5.7 range is not visible from the tablet. What the hell is this?

Ravine number 5

The damned regulatory domain works not only on the access point, but also on the receiving device.

In particular, my Microsoft Surface Pro 3, although made for the European market, basically does not support the 5.7 band. I had to switch to 5.2, but then at least the 40 MHz mode started up.

Ravine number 6

Everything started up. The points started, 2.4 shows a speed of 130 Mbps (would be SHORT-GI, it would be 144.4). Why a card with three antennas only supports 2 spatial streams is a mystery.

Ravine number 7

It started up, and sometimes the ping jumps up to 200, and that's it.

And the secret is not at all hidden in the access point. The fact is that according to Microsoft rules, Wi-Fi card drivers themselves must contain software for finding networks and connecting to them. It's like in the good old days, when a 56k modem had to have a dialer with it (which we all changed to Shiva, because the dialer that came with Internet Explorer 3.0 was too terrible) or the ADSL modem had to have a client PPPoE.

But even those who do not have a standard utility (that is, everyone in the world!), Microsoft took care of it by making the so-called “Wi-Fi auto-configuration”. This auto-configuration cheerfully spits on the fact that we are already connected to the network, and scans the range every X seconds. Windows 10 doesn't even have a "refresh networks" button. Works fine as long as there are two or three networks around. And when there are 44 of them, the system freezes and gives out a few seconds of 400 ping.

"Autoconfiguration" can be disabled with the command:

Netsh wlan set autoconfig enabled=no interface="???????????? ????" pause

Personally, I even made myself two batch files on the desktop “enable autoscan” and “disable autoscan”.

Yes, please note that if you have Russian Windows, then most likely the network interface will have a name in Russian in the IBM CP866 encoding.

Summery

I've rolled out a rather long sheet of text, and I should have ended it with a brief summary of the most important things:

1. The access point can only work in one range: 2.4 or 5.2 or 5.7. Choose carefully.
2. The best regulatory domain is VE.
3. The commands iw phy info, iw reg get will show you what you can do.
4. Channel 13 is usually empty.
5. ACS_SURVEY, 20MHz channel width, TX-STBC, RX-STBC123 will improve signal quality.
6. 40 MHz, more antennas, SHORT-GI will increase the speed.
7. hostapd -dddtK allows you to run hostapd in debug mode.
8. For amateurs, you can rebuild the core and CRDA, increasing the signal strength and removing the restrictions of the regulatory domain.
9. Auto-discovery of Wi-Fi in Windows is disabled with the command netsh wlan set autoconfig enabled=no interface="???????????? ????"
10 . Microsoft Surface Pro 3 does not support the 5.7 GHz band.

Afterword

I found most of the materials used in writing this guide either in google or in mana for iw, hostapd, hostapd_cli.

How exactly the speed depends on the number of antennas and what speed can be achieved with three antennas - I don’t know. Comments would be welcome.

In fact, the problem IS NOT SOLVED. At times, the ping still jumps to 400 and stays at that level, even for the “empty” 5.2 GHz band. Therefore:

I am looking for a Wi-Fi spectrum analyzer in Moscow, equipped with an operator, with whom I could check what the problem is, and whether it is that there is a very important and secret military institution nearby that no one knows about.

P.S

Wi-Fi operates at frequencies from 2 GHz to 60 GHz (less common formats). This gives us a wavelength of 150mm to 5mm. (Why do we even measure radio in frequencies and not in wavelengths? It's also more convenient!) I, in general, have an idea, buy wallpaper from a quarter-wavelength metal mesh (1 mm is enough) and make a Faraday cage to guarantee isolate yourself from neighboring Wi-Fi, and at the same time from all other radio equipment, such as DECT phones, microwaves and traffic radars (24 GHz). One problem is that it will block GSM / UMTS / LTE phones, but you can allocate a stationary charging point for them by the window.

I will be glad to answer your questions in the comments.

1. Measure your speed using the correct measure

The first mistake many people make is determining how fast their wireless connection is based on the "Speed" item in the Windows Wireless Properties window.

Figure 1. Ignore this number

In fact, this number is only remotely related to the actual throughput of the wireless connection. This outputs a value that tells the wireless adapter driver to output - link rate.

Link rate is also called PHY (physical layer - physical layer) - the maximum speed at which data can move through a wireless connection between a wireless client and a wireless router. For a 10/100 Ethernet network card, you'll typically see 100 Mbps, and for a gigabit network card, you'll see 1000 Mbps (if you're connected to the switch's gigabit port).

The receiver speed at the application layer will be much slower than the speed of the physical layer. Indeed, a link rate of "300 Mbps" usually corresponds to 50 to 90 Mbps at the TCP/UDP level (depending on the 802.11n router and adapter used).

The reason for such a significant difference is the large "overhead" associated with wireless connections (many bits are used to transmit information to other than the intended recipients; plus retransmission data due to the unreliability of wireless communication)

To get a more accurate measurement of wireless connection speed, you need to use methods that actually measure delivery speed. Namely:

• calculating the speed by dividing the file size by the transfer time. LAN Speed ​​Test does the same automatically under Windows
• copy files and use Networking monitor (Start > Run perfmon.msc) in XP
• using NetMeter while watching streaming video or transferring files
• using Iperf on the command line and a graphical shell for it Jperf . Convenient graphical interface + cisco router on the remote side allows you to check the speed of the communication channel

Of course, whichever method you use, you must first try the same method for a wired connection as for a wireless one. This will let you know what you are losing when using a wireless connection.