Factors affecting the data transfer rate when using ADSL technology.

Everyone has repeatedly heard about networks of the second, third and fourth generation of mobile communications. Some may have already read about the networks of the future - the fifth generation. But questions - what does G, E, 3G, H, 3G+, 4G or LTE mean on a smartphone screen and what is faster among this are still of concern to many people. We will answer them.

These icons indicate the type of connection your smartphone, tablet or modem has to the mobile network.

1. G(GPRS - General Packet Radio Services): The slowest and most obsolete packet data connection option. First standard mobile internet, performed by an add-on over GSM (after a CSD connection up to 9.6 kbps). The maximum speed of the GPRS channel is 171.2 kbps. At the same time, the real one, as a rule, is an order of magnitude lower, and the Internet here is not always functional in principle.

2. E(EDGE or EGPRS - Enhanced Data rates for GSM Evolution): Faster add-on over 2G and 2.5G. Technology of digital data transmission. The speed of EDGE is about 3 times higher than GPRS: up to 474.6 kbps. However, it also belongs to the second generation of wireless communication and is already outdated. The real EDGE speed is usually kept in the region of 150-200 kbps and directly depends on the location of the subscriber - that is, the workload of the base station in a particular area.

3. 3 G(Third Generation - third generation). Here, not only data transfer is possible over the network, but also “voices”. The quality of voice transmission in 3G networks (if both interlocutors are within their range) can be an order of magnitude higher than in 2G (GSM). Internet speed in 3G is also much higher, and its quality, as a rule, is already quite sufficient for comfortable work on mobile devices and even desktop computers via USB modems. At the same time, your current position may affect the data transfer rate, incl. whether you are in one place or moving in transport:

  • Stay still: typically up to 2 Mbps
  • Drive at speeds up to 3 km/h: up to 384 kbps
  • Travel at speeds up to 120 km/h: up to 144 kbps.

4. 3,5 G.3G+,h,H+(HSPDA - High-Speed ​​Downlink Packet Access): The next high-speed packet data add-on is already over 3G. In this case, the data transfer rate is very close to 4G and in H mode it is up to 42 Mbps. In real life, mobile Internet in this mode average works for mobile operators at speeds of 3-12 Mbps (sometimes higher). For those who do not understand: it is very fast and quite enough to watch online video in not too high quality (resolution) or download heavy files with a stable connection.

Also in 3G there was a video call function:

5. 4G, LTE(Long-Term Evolution - long-term development, the fourth generation of mobile Internet). This technology is used only for data transmission (not for "voice"). The maximum download speed here is up to 326 Mbps, upload - 172.8 Mbps. The real values ​​are again an order of magnitude lower than the declared ones, but they still amount to tens of megabits per second (in practice, often comparable to mode H; in Moscow, usually 10-50 Mbps). At the same time, faster PING and the technology itself make 4G the most preferred standard for mobile Internet in modems. Smartphones and tablets in 4G (LTE) networks hold a battery charge longer than in 3G.

6. LTE-A(LTE Advanced - LTE upgrade). The peak data transfer rate here is up to 1 Gbps. In reality, the Internet is capable of operating at speeds up to 300 Mbps (5 times faster than regular LTE).

7. VoLTE(Voice over LTE - voice over LTE, as an additional development of technology): a technology for transmitting voice calls over LTE networks based on IP Multimedia Subsystem (IMS). The connection speed is up to 5 times faster compared to 2G/3G, and the quality of the conversation itself and voice transmission is even higher and cleaner.

8. 5 G(fifth generation cellular communication based on IMT-2020). The standard of the future is still under development and testing. The data transfer rate in the commercial version of the networks is promised to be up to 30 times higher than LTE: the maximum data transfer can be up to 10 Gb / s.

Of course, you can use any of the above technologies if your equipment supports it. Also, its work depends on the capabilities of the mobile operator itself at a particular location of the subscriber and his tariff plan.

© Vladislav Timoshenko

One of the main characteristics of Internet connection from the point of view of the end user is the speed of Internet access. Almost all providers, when offering their services, draw the attention of potential subscribers to the maximum possible theoretical speed values ​​that are possible for a specific connection method.

For example, providers providing Internet access via local networks (Ethernet technology) draw the attention of potential subscribers to the possibility of accessing their resources at speeds up to 100 Mbps. Providers that use the xDSL family of technologies most often talk about the independence of the speed characteristics of the subscriber from the channel load by other subscribers, guaranteeing speed in the final section to the subscriber up to 1-8 Mbps (depending on the tariff plan). Cable TV providers using DOCSIS protocols draw the attention of subscribers to the fact that there are no restrictions on the Internet access speed from the provider, which currently makes it possible to theoretically access various Network resources at a speed up to 38 Mbps.

Such communication primarily focuses the attention of the end user on the benefits of certain technologies that are used by providers in providing access to the Internet. At the same time, providers, of course, do not provide subscribers with fairly simple, but at the same time important practical knowledge necessary for competent and adequate perception specifications provided services. As mentioned above, one of these main characteristics is access speed.

Let's try to figure it out. Since it is questions related to the speed of access to certain Internet resources that most often mislead not only novice users, but also quite experienced subscribers. Which, however, is quite natural, since here they are directly confronted with such special technical questions as metrics, characteristics and diagnostics of telecommunication communication channels. But without a minimum understanding of the basic principles of this area of ​​knowledge, it is impossible to adequately and competently use the Internet.

First, let's start with the theory.

About bits, bytes and speed

So what are bits and bytes. Without going into theoretical details, we note that a bit is the smallest unit of measurement for the amount of information. Along with a bit, a byte is actively used. A byte is equal to 8 bits.

Since bits and bytes are very small quantities, they are mainly used with prefixes kilo-, mega- and giga-. Common units and their abbreviations are shown in Table 1.

Table 1

Name

English abbreviation

Russian abbreviation

Meaning

kilobyte

KByte (KB)

1000 kilobits

megabyte

MB (MB)

1024 kilobytes

1000 megabits

gigabyte

GB (GB)

1024 megabytes

Connection speed- this is a certain amount of information received or sent in a certain unit of time.

In this case, it is customary to consider a second as a unit of time, and a bit as a unit of the amount of information. By analogy with the unit of measurement for the speed of physical bodies - a meter per second (m / s), the speed of an Internet connection is usually indicated in bits per second (and its derivatives - kbps, Mbps, gigabit / s).

Thus, if at a particular moment in time the speed of receiving or receiving information is, for example, 1 Mbps, then this means that the connection has a bandwidth of 1000 kilobits per second, or 128 kilobytes per second.

Let us immediately point out one fairly popular misconception here, which can lead to a misunderstanding by subscribers of the conditions for the provision of services on the part of providers. This is especially true for those subscribers whose tariff plans take into account the amount of received / transmitted information. We are talking about the number of bytes in one kilobyte, which, according to international system SI units is equal to 10 3 bytes (or 1000 bytes) in 1 kilobyte. At the same time, 1024 bytes contains 1 kibibyte. Those who wish to get acquainted with the details in this matter can visit this resource - http://en.wikipedia.org/wiki/%D0%9A%D0%B8%D0%BB%D0%BE%D0%B1%D0%B0%D0%B9%D1%82

It is appropriate to remember this remark when reading a contract for the provision of services with a provider, where this ratio is usually explicitly stated. And it is it that is taken as a unit of measurement when taking into account the consumed traffic.

However, no matter how many bytes in one kilobyte (1000 or 1024) your provider uses in their accounting systems, it absolutely no effect on the total amount of the most recorded information, since it is only a relative accounting unit, with the help of which all calculations are carried out within the system.

On the influence of "urban conditions" on the speed of movement

Let's take a closer look at the issue of Internet access speed.

It must always be remembered that any provider can guarantee the speed closest to the maximum declared only on the "last mile", that is, on the section of the telecommunications network from the subscriber's equipment (usually a DSL or cable modem, network card, etc.) to the point enabling this subscriber on its technical platform (provider's server, router, etc.).

If the provider has its own network resources (file, multimedia, game servers), then usually the speed of access to them at certain points in time can also be close to the maximum declared. For example, when connected to local network according to Ethernet technology, it would seem that the speed of access to the provider's resources should be at least 100 Mbps. However, this is not the case. There are many reasons. One of least famous - full throughput"100-megabit" network cards actually do not exceed 80-90 Mbit / s (and for cheap models, and at all - 35-40 Mbps). Basically, even in this case, there will always be deviations from these values ​​even lower, associated with overloads in the segments (sections) of the provider's network, as well as a specific resource.

Let's assume that there is a server that is connected to the network using a 100-megabit network card. And at a certain point in time, 10 users connect to it at the same time, requesting the same large file (for example, a movie). In general, the theoretical bandwidth of the channel will be divided among these ten users in such a way that none of them will benefit. If this happens at the same time, then each of these 10 users will connect to the resource at a speed of 10 Mbps (100 Mbps / 10 users). Thus, the speed of access to this file will be no more than 10 Mbps for each user.

Of course, this example is somewhat arbitrary. In reality, the speed characteristics in this example will be somewhat higher, since the simultaneous access of many users to the same file is unlikely and, moreover, it is of an extremely limited temporal nature.

However, this example illustrates all the ambiguity and inconsistency of the speed characteristics of access to Internet resources. Indeed, by its nature, the Internet is a collection of disparate networks, computers, servers, united only general rules(protocols) of interaction with each other.

Various Internet resources are not only geographically located in different parts of the Earth, but are connected by various communication channels, each of which, in turn, has certain speed characteristics and is subject to various time periods different loads. It is also necessary to take into account the temporary, dynamic, constantly changing nature of Internet access routes from one point to another. For example, if at the moment one route can be used when accessing a certain site in Europe from Ukraine, then after some time (albeit a short one), this route may change, which will lead to indefinite changes in the speed and time of access to it.

Quite often in various parts of the Internet, failures occur on telecommunications backbones and main routers. In such cases, all traffic is redirected through backup, slower channels.

The Internet is a branched structure, where the speed of access depends on both the receiving party and the remote source. Often there is a situation when the speed of access to certain Internet resources is artificially limited to prevent large overloads and to avoid subsequent unavailability of the resource. A specific provider most often cannot control all such processes.

Naturally, such a general heterogeneity cannot but affect the speed of access to Internet resources. And no provider in the general case can guarantee not only a certain speed of access to a specific resource, but also the very possibility of such access. Only an extremely limited amount of resources, which are controlled by your provider and belong to its area of ​​responsibility, can be guaranteed (with a high degree of probability) available to its subscribers.

Everything The above can be illustrated by the following example well understood by residents of large cities.

Let's imagine that you are moving along city streets in a car whose manufacturer has indicated a maximum speed of 250 km/h on its speedometer. Naturally, any driver (and not only) is well aware of the fact that such a speed is possible only if a number of favorable factors are observed - the congestion of the route, signs traffic, the relief of the road and the quality of its surface, weather conditions, seasons, the technical condition of the car, the emotional and physical state of the driver, the presence of obstacles, etc. At the slightest difficulty of any nature, driving at the maximum possible speed specified by the car manufacturer is impossible. And even more so, such a movement is absolutely unrealistic in the evening, in conditions of heavy traffic, in city traffic jams. No one can guarantee not only the maximum speed of movement, but also the average speed from one point of the city to another. The traffic situation changes quickly and often unpredictably.

Of course, in such a situation, it would never occur to anyone to blame the car manufacturer for the fact that the maximum possible promised speed of 250 km / h is impossible in the described situation. The manufacturer, however, did everything in his power so that under certain conditions the car could show such high-speed characteristics, but no one can guarantee them at any time on any road. At the same time, the driver of the car is obliged and may keep the car in such technical condition, as well as to have the appropriate driving skills, which will allow, under other favorable conditions, to show high speed characteristics that are adequate to the situation.

I hope that this example served as a good illustration of the topic of Internet access speed.

Factors Affecting Internet Access Speed

Having read up to this point, you may get the impression that Internet providers cannot even theoretically provide any guaranteed speed of Internet access for their subscribers. That nothing depends on them and nothing can be done in case of any difficulties. And since nothing depends on the providers themselves, then what may depend on users in this issue? Naturally, this impression is completely wrong.

A lot depends on both the providers themselves and the users, despite all the actual ambiguity of the situation described above. Naturally, within the limits of authority and the limits of responsibility of each.

Subscribers should be aware that the quality of the Internet access service is largely due to the successful actions of their provider. It is he who is obliged to ensure the quality of the "last mile" to each subscriber. The provider must constantly work on expanding its external channels to the Internet, avoiding a "bottleneck" at the exit from its own network. The provider can constantly monitor (track) the situation in order to identify and prevent emergency and emergency situations in their area of ​​responsibility. And finally, it is the provider that provides subscribers not only with the technical ability to connect to the Internet. He sells " service access to the Internet”, while providing its subscribers with support, offering new opportunities and attractive services.

Also, a lot of own responsibility for ensuring high speed of work on the Internet lies with the subscriber himself. Since providers in the vast majority of cases relieve themselves of responsibility for the correct configuration computer equipment subscriber. Namely, the speed characteristics of the work on the Internet largely depend on this. After all, with other equal conditions competent setting the subscriber's computer is the key to quick and trouble-free access to the Network with the help of this computer.

We list those factors that have a decisive influence on the speed of Internet access forend user, focusing this time on those that can be controlled and adjusted by the subscribers themselves. This list is not exhaustive and is not meant to be a complete guide to setting the relevant parameters. It only focuses the attention of users on the most important and sensitive points that must be considered in order to guarantee on their part, the creation of the most favorable conditions for obtaining high speeds of Internet access.

Internet connection speed depends on:

      • used tariff plan

        the presence of all kinds of viruses and similar programs, as well as programs that filter traffic

        operating system settings and correct operation of the hardware

        network card driver settings

        intermediate network equipment(router, etc.)

        use of wireless internet(wi-fi)

        quality"last mile"

Let's look at each of these factors in detail.

1. Tariff plan.

For example, the VOLIA company for home users offers tariff plans without established restrictions on the speed of data transfer from the provider to the subscriber. In them, the channel bandwidth to the subscriber is not artificially limited and is equal to the maximum possible with the currently used technology according to the DOCSIS 2.0 standard - i.e. up to 38 Mbps. This means that there are no speed limits on the part of the provider when transmitting information to the subscriber, except for the limitations caused by the technology itself. I.e 38 Mbps is the speed of possible information reception exclusively and only by the subscriber's modem. At the same time, the speed of information receipt to the modem itself is not regulated.

As mentioned above, it should always be taken into account that the maximum theoretically possible peak speed in normal conditions is never reached. The main reasons for this have been discussed in detail above.

On tariffs without traffic restrictions (the so-called unlimited tariffs) the speed of the Internet channel is already artificially limited by the provider. VOLIA provides such conditions in its business tariffs, which, however, are also available to home users.

Some providers may limit the connection speed for a very large amount of incoming traffic (hundreds of gigabytes per month). Despite the fact that they declare in the tariff plan the absence of a volume limit, providers can forcibly reduce the connection speed of clients consuming a large number of traffic.

2. The presence of all kinds of viruses and similar programs, as well as programs that filter traffic, with the ability to limit the channel width (firewalls, antiviruses with built-in filters, programs for peer-to-peer networks, etc.).

Despite some seeming banality and obviousness of this paragraph, I would like to draw attention to its particular importance in ensuring not only high speed Internet access, but also the overall security of the computer and the data on it.

The presence of other programs mentioned - firewalls, antiviruses - is now mandatory on any computer connected to the Internet. But in this regard, it must be taken into account that the work of these programs is to check all the volumes of information that goes and comes to the computer, including via the Internet. In this regard, they need a certain amount of time for this, which in itself slows down the exchange of information. But in this case, a small overhead is quite justified.

Currently popular programs for working in peer-to-peer networks (DC ++, Torrent, etc.), with unqualified settings, can not only significantly reduce the speed of information exchange, but also affect the overall performance of the computer. Therefore, if the user works in such networks, he must carefully approach the settings of the relevant programs and take into account their network activity when assessing the overall speed of their work on the Internet.

Reduces the speed of information exchange with the Internet and various software- automatic updates of various programs and the operating system itself in the background, "beacons" of licensed software. This software, using the default settings, can connect to some servers without the user's knowledge and receive / transmit service data, which reduces the overall useful connection speed. In this case, we can only advise users to be attentive to all installed software and to manual setting each of them, paying particular attention to the procedure and regulations for their automatic updating.

3. Operating system settings and correct operation of the hardware.

The settings of any operating system for working with the Internet refer to difficult questions their administration. And therefore, they must either remain as they were set by the manufacturer, or be corrected only by very experienced users who understand all the consequences of their actions. This applies primarily to operating systems of the Windows family, as the most common among most Internet users.

Here we will only pay attention to the possibility of some acceleration of work on the Internet with qualified fine-tuning of any operating system. In this regard, we can recommend one of the best programs in its class - TuneUp Utilities ( http://www.tune-up.com) that helps optimize many implicit settings in the Windows operating system. This can lead to an increase in Internet exchange speed and overall performance. In particular, the speed of information exchange on the Internet is affected by such parameters of the TCP / IP protocol as the MTU packet size, Rwin window size and the number of parallel connections.

More experienced users can use the program SG TCP Optimizer(http://www.speedguide.net). This utility allows you to optimize the order in which data packets are sent and received. To do this, on the “General Settings” tab (Fig. 1), you just need to drag the scale slider to a value that approximately corresponds to the maximum connection speed and set the “Optimal Settings” option. Then click "Apply Changes". Next, you need to reboot the operating system.

Here it is necessary to mention one of the widespread myths that supposedly the Windows operating system is for your needs (or rather, for the service QoS) always reserves 20% of the available channel bandwidth. This is not entirely true. More details on this issue can be found (then accept correct solution) here - http://support.microsoft.com/kb/316666/en and here http://www.winline.ru/articles/1383.php?SHOWALL_1=1

In addition to the TCP / IP protocol parameters, it is also necessary to take into account other settings for the operating system itself (startup programs, services, visual effects, etc.).

Once again, we draw attention to the need for qualified and conscious actions with this kind of software. Only a professional, competent and comprehensive approach to optimization issues can bring obvious benefits, and not cause harm.

Fig 1. SG TCP Optimizer

4. Installed network card drivers.

It has already been noted above that the total throughput of 100-Mbps network adapters actually does not exceed 80-90 Mbps, and often less (up to 40 Mbps). Currently, all network cards are mostly configured and configured automatically, which in most cases does not require manual intervention in their configuration. But in some cases, a simple adjustment of the network card driver can slightly increase its performance. Note in this connection that the maximum value of such parameters as number of TX descriptors(number of receive buffers) and number of RX descriptors(number of transmission buffers) has a particularly noticeable effect on the performance of a network card when exchanging large amounts of information.

In this regard, mention must be made of next moment, which will be of interest to users using cable modems. The fact is that a cable modem can be connected to a computer in two different ways:

    using an ethernet cable to a computer network card;

    using a USB cable to a computer USB port.

To get a high speed of working on the Internet, we can recommend connecting a cable modem to a computer only in the first way, i.e. using an ethernet cable. The fact is that in this case the modem and the user's computer will be connected at speeds up to 100 Mbps. In addition, when connected using an ethernet cable, you will not need to install special modem drivers. If you want to use a wired or wireless router with your cable modem, you must also use an ethernet cable to connect it to the modem.

And when connecting the modem and the computer with a USB cable, the speed of such a connection will depend on the version of the USB bus that is supported by the modem and on the number of other devices that are connected to the computer via USB ports (mouse, keyboard, external drives, printer, scanner, etc. .). Modems using USB 1.1 (1.0) will not exceed 12 Mbps on the modem-computer section (which is the maximum data transfer rate for these versions of USB). This can cause some kind of "bottleneck" between the modem and the computer. If your cable modem uses USB 2.0, then there is no such limitation.

In addition, the USB cable without special amplification has a maximum length of about 3 meters, which limits the possibility of placing equipment relative to each other. An Ethernet cable without signal amplification in such cases can be much longer (100 meters or more).

5. Settings of intermediate network equipment

At present, network equipment (routers) is becoming widespread among home users, which helps to connect several devices to the Internet, thus making it possible for several home computers (laptops, PDAs, etc.) to work on the Internet.

Without delving into this issue, we will only point out here the need for qualified configuration of this hardware, since being an intermediate link between the subscriber's computer and the provider's network (modem), routers can significantly slow down the exchange of information.

As an example, here is a picture of one of the router settings pages ASUS WL-500W(see Fig.2). For example, on the Bandwidth Management page, you can set various priority settings for the traffic passing through the router. The maximum value of the speed of information exchange on the Internet on a computer connected through this router was reached only when this parameter was set to ftp server. For other values ​​of this parameter, the speed drop reached 20%. Disabling the hardware firewall built into the router, although it will somewhat weaken the protection of computers connected to it, will increase the overall speed of information exchange through the router.

Fig.2

This example illustrates the need to correctly configure any intermediate equipment between the provider's network and user computers.

In addition, we recommend that you regularly update the firmware of your router. For example, one of the latest firmware updates for the router ZyXEL P-300W declares a significant increase in performance due to the optimization of the firmware code.

6. Wi-Fi connection

Internet connection using Wi-Fi technology is also becoming increasingly popular, which allows you to connect computers and laptops to the Internet without using wires. You can read more about Wi-Fi, for example, here - http://ru.wikipedia.org/wiki/Wi-Fi

As part of our topic, we only note that in order to get the maximum speed when connecting via Wi-Fi, you must use its latest standards - for example, 802.11g, be located as close as possible to the access point, if possible, place it within the visibility zone. If there are other Wi-Fi zones nearby own point access points must be set to a different transmission channel than those used by nearby access points.

In general, the issues of setting up and operating Wi-Fi are beyond the scope of our discussion, however, we note that the competent configuration of Wi-Fi equipment can significantly increase the speed of information exchange.

7. Last Mile Quality

As mentioned above, the term "last mile" means a channel connecting the end (client) equipment with the provider's access node. Depending on the provider and the technology they use, last mile may mean different technological solutions. If access is provided via ethernet technology, then the last mile means the section from the provider's switch port on its communication node to the client's port (network card).

When providing access through cable television networks using DOCSIS protocols, under the "last mile" you can accept a section of a coaxial cable network, which is limited, on the one hand, by the subscriber's equipment (television receiver, computer or other equipment), and on the other hand, by the connection point to provider's telecommunications network. It is also called a subscriber branch.

It is on the quality of this subscriber branch, which is the property of the subscriber and completely controlled by him, that not only the speed of information exchange with the provider's equipment, but also its stability depends. Therefore, each subscriber should pay some attention to it. No special skills and knowledge are usually required from the subscriber for this. During the initial connection, some providers (however, unfortunately, not all) can upgrade or shift the subscriber branch at their own expense.

With some skill and desire, cable provider users can indirectly control the quality of an existing subscriber spur. In addition to visual inspection of the cable and connecting devices (splitters, etc.) in his apartment, the subscriber has the opportunity to monitor the signal level on his modem. This is especially true in case of obvious violations and failures in access to Internet resources.

Depending on the model of your cable modem, you can check the signal levels on the line in different ways.

So on modems from Motorola (and others), it’s enough to go to the page at http://192.168.100.1/ Signal levels on Terayon TJ715x modems can be checked on the page http://192.168.100.1/diagnostics_page.html (after entering the password icu4at!) - fig.3. Signal levels can be seen in the "Connection" tab.

Fig.3.

For practical purposes, it is enough to view the values ​​of three signals: Upstream TX signal power, Downstream RX signal power, Downstream SNR. In order to eliminate problems on the subscriber branch, the signal values ​​should be as follows:

    Upstream TX signal power: +39...+51 dBmV, the best value is about 45 dBm (in the figure - 52 dBm)

    Downstream RX signal power: -15...+15 dBmV, the best value is about 0 dBmV (in the figure - 2.9 dBmV)

    Downstream SNR: not less than 32 (35) dBmV, the best value - the more the better (in the figure - 35.2 dBmV)

If the indicated signals are lower, it can be concluded that there are problems on the subscriber branch and the cable should be checked. To do this, you usually need to call the specialists of your provider.

Local network subscribers accessing the Internet using ethernet technology do not need such checks. They do not have such an opportunity and should immediately contact the specialists of their provider about the quality of their subscriber cable.

If xDSL technology is used to access the Internet (via telephone lines), then in order to obtain information about the quality of your line to the provider, you must also go to the Web interface of your modem. In general, the address is used for this http://192.168.1.1 . Some models of Dynamix DSL modem may have an address: http://10.0.0.2

After that, enter one of the following logins and passwords, case-sensitive (the specific value depends on the modem model) - Table 2.

Table 2.

Modem model

Username

Other models

(possible options)

Next, it is necessary to control such technical values, which characterize, among other things, the quality of the line, such as SNR Margin (Local/Remote), Line Attn (Local/Remote) other. Unfortunately, it is impossible to provide brief and universal advice on assessing these parameters within the framework of this article. We only note that in most cases, the speed characteristics of a DSL connection depend on the distance to the provider's equipment at the telephone exchange and on the cross-sectional size and quality of the connecting wires.

If you suspect a poor line quality with a DSL connection, you can contact special sources of information and the staff of your provider. If you get access to the Internet from the Ukrtelecom company, then the following link can be recommended as a good source of information about this service - http://ogo.in.ua

Tools for checking the speed of access to Internet resources and evaluating the results

How can you still find out the speed of your Internet connection?

Once again, regardless of the Internet connection technology (Ethernet, xDSL, DOCSIS, etc.), the section of the network from the ISP to your computer is called the last mile. The speed of information transfer on the last mile (due to its controllability by the provider and the subscriber) is usually much higher than in other parts of the Network. Let's also assume that the provider itself does not artificially limit the speed on the last mile to its customers, that is, it provides Internet access without speed limits.

After the "last mile" there are trunk lines connecting your provider and the so-called Traffic exchange centers. These backbones, in turn, can pass through several intermediate providers on their networks. In this regard, it is important to know whether your provider has its own direct connection to the traffic exchange centers or uses intermediate networks for this purpose.

Traffic exchange centers (in Ukraine it is - U.A.- IX) between telecom operators and can, to some extent, be considered the place where “the Internet is located”. The speed of access to them must be measured.

What methods and methods are used for this?

Development metric parameters telecommunication channels and methods of their control is an actively discussed problem even among specialists. Some solutions are formulated in the form of the IPPM (Internet Protocol Performance Metrics) standard - http://www.ietf.org/html.charters/ippm-charter.html).

The significant quantitative parameters in this standard are:

    access time,

    loss rate,

    access time variation,

    bit bandwidth.

Experts note that the main disadvantage of this set of parameters for channel description and quality prediction is their high level of variability, that is, unpredictable variability, randomness of deviations. Nevertheless, it is the values ​​of these parameters obtained over a certain period of time under controlled conditions that can give an estimated picture of the quality of a particular communication channel.

In principle, it is not possible to measure "Internet speed" by the definition of this term itself. We can only talk about. The most correct in Ukraine for subscribers of most providers, as mentioned above, are speed measurements to the traffic exchange center UA-IX. The values ​​obtained in this case can show the state of the "last mile" and the provider's network, that is, what is in the area of ​​responsibility of both the subscriber and the provider.

Of course, measurements of access time and speed, packet loss rate, and the like are also possible. and to any other point on the Internet, but such data can only be of an estimated probabilistic nature. By themselves, without taking into account specific conditions at a certain point in time, they cannot serve as an assessment of the overall "Internet quality" and "access speed" as such. Unless, of course, the values ​​​​of these values ​​\u200b\u200bdid not reveal problems precisely on the last mile or directly in your provider's network.

All of the above must be taken into account when carrying out specific measurements in each case.

Let's take a closer look at the most simple methods and methods for checking the above parameters, which can be applied by ordinary users.

First of all, we note that during the speed test, it is necessary, if possible, to terminate (suspend) all software that uses the Internet connection, as well as in one way or another limiting the exchange rate: antiviruses (including anti-virus database updates), Windows updates, firewalls , download managers, peer-to-peer network clients, messaging programs (ICQ, QIP, etc.). All downloads must be stopped for the duration of the speed tests. Downloads include both direct downloading of files using download managers or a browser, and processes of waiting for the full loading of pages opened before passing tests.

If there are routers and other network equipment as intermediate links between the user's computer and the provider's network, it is desirable to temporarily abandon them if possible.

If you use any browser to get the test result, then in order to obtain the most accurate results, two conditions must be met:

1) it is necessary to enable the display of images in the browser settings;

2) during the test, no downloads, downloads, page navigations, or other actions related to network activity on the Internet should be carried out.

More correct, however, instead of a browser is to use special programs of download managers (FlashGet, Download Master, Free Download Manager, etc.). It has been noticed that browsers (eg Opera) incorrectly measure the data transfer speed (probably due to software restrictions on the priorities of viewing pages on sites and downloading large amounts of information). And special programs are deprived of this shortcoming. In addition, they can organize downloads in several streams. Which significantly speeds up the process and fully loads the existing channel.

To obtain more accurate reliable results, it is imperative to test from 3 to 10 times. Moreover, the faster your Internet connection, the large quantity times it is necessary to carry out tests. The final result can be considered the average value of the obtained value.

In any case, the difference between several test results taken one after the other should not exceed a few percent. If the difference is greater, then the source of this change in speed is somewhere nearby (the last mile, congestion in your provider's network, etc.).

The simplest and most easily accessible, but at the same time quite objective method of testing your Internet channel, can be considered the following: you need to download (or upload) something from / to an unloaded FTP server located somewhere nearby (for example, , on your ISP's network).

Usually every ISP has such an FTP server. Some of them even have files specially designed for such tests. For example, the company VOLIA (Kyiv) has a resource for these purposes ftp://ftp.volia.net with large files test10m (size 10 MB), test5m (size 5 MB).

To ensure full use of the bandwidth of your Internet connection is possible only if there are several simultaneous download sessions. One file download session usually does not allow you to fully load the communication channel due to some limitations of data transfer protocols. Therefore, it is necessary to upload several test files at the same time, thus creating the maximum possible load.

Such studies can reveal problems in the last mile (insufficient transmission bandwidth, packet loss, access time).

The vast majority of users are interested in the speed of access to those resources of the Network that are located outside the local network of their provider. In this case, special Internet resources can help - test sites that measure the speed of downloading or uploading between the user's computer and this resource.

The most popular resource for checking link speed in UA-IX can be considered http://itc.ua/modem

I would like to highlight this resource - http://www.dslreports.com/stest Uses some invisible MP3 file to test speed on it. Before testing, you should select the site from which you want to test the speed of access to your resource. The list of sites includes over 380 similar resources: from the USA and Argentina to the Philippines and New Zealand.

We emphasize once again that in carrying out all such measurements, one can only speak of measurement medium size speed at a given time to a specific node. This happens in the general case due to the uncertainty of routing between different segments (sections) and resources of the Network. The path from the user computer to the server can go through a different number of sections of the Network, each of which may have its own problems with the channel, as a result of which the speed of access to it measured by different tests may differ.

Currently, to check the speed of an Internet connection, the resource has become very popular. http://www.speedtest.net (Fig. 4). In a sense, it can even be considered a standard of this kind of sites.

Fig.4.

To check the access speed with it, you need to go to http://www.speedtest.net (Macromedia Flash must be installed to display it), after which your location will be automatically determined immediately. Further on the map, you can select the city to which you want to test the speed (for example, the closest city to the user available on the map is Kyiv). After some time, the result will appear - the value of the speed OT / To the resource at the moment. Figure 5 shows such a result obtained from the network of the provider VOLIA (Kyiv).

Fig.5

Let's now try to adequately evaluate the result obtained using this example. To do this, you will have to additionally perform some measurements.

What does the information in Figure 5 directly indicate?

a) The access speed was measured between the user's computer located in the Volya network and a certain test server in Kyiv. In this case, this server is speedtest.dc.utel.ua. Information about this appears during the speed measurement process and is not reflected in the final result.

b) Access time to speedtest.dc.utel.ua at the time of testing is 15 milliseconds, which is quite good.

c) Speed ​​at the time of testing FROM server speedtest.dc.utel.uaTo the user (download) is 6402 Kbps. Speed ​​at the time of testing To server speedtest.dc.utel.ua FROM user (upload) is 224 Kbps.

That's probably all that can be said from the results of this test. And for many tests similar to it on similar resources. It is not possible to draw any truly meaningful conclusions from these results.

Therefore, additional research is needed.

First you need to test packet transmission delay time and packet loss rate on the route to speedtest.dc.utel.ua. To do this, in general, it is enough to use the program ping. With regard to our task (checking the speed of access to certain network objects), this program can be used as the main tool that provides measurements of the so-called round-trip time of packets along the route of interest back and forth. You can find many on the web detailed information about using this program.

Naturally ping can and should be applied to any Internet node, including when checking the quality of the "last mile". Packet losses recorded by this program when checking the provider's equipment closest to the subscriber, a large packet turnover time interval exceeding the allowable values ​​(usually over 100 milliseconds), with a high degree of probability indicate problems on the "last mile".

Here are the results of running the command ping speedtest.dc.utel.ua:

Packet exchange with speedtest.dc.utel.ua for 32 bytes:

Response from 213.186.113.11: number of bytes=32 time=20ms TTL=58

Response from 213.186.113.11: number of bytes=32 time=12ms TTL=58

Response from 213.186.113.11: number of bytes=32 time=11ms TTL=58

Ping stats for 213.186.113.11:

Packets: sent = 4, received = 4, lost = 0 (0% loss),

Approximate round-trip time in ms:

Minimum = 11ms, Maximum = 20ms, Average = 13ms

By default, 4 packets are sent and the packet return time from the endpoint is measured. Thus, we confirmed that out of 4 packets sent, responses also came from 4. That is, there is no packet loss. The average time for receiving and transmitting packets was 13 milliseconds.

The results of this simple test indicate that there are no problems on the line between the user's computer and the host. speedtest.dc.utel.ua.

Often in special sources recommendations are given to use ping with options like ping -f,pingl etc. Without being able to consider in detail all kinds of keys and parameters for launching the program within the framework of this publication, we note that in the vast majority of cases their use does not seem necessary to test exactly the speed of access to a particular network resource. After all, the result of executing the command ping can reliably speak only about the delay time and the very fact of packet transmission to the specified node. So ping allows you to quickly determine the presence or absence of connection problems at the physical (cable breakage, cable damage, network card failure, etc.) and software (for example, blocking all network traffic with firewalls) levels.

However, let's continue our measurements.

For this we use the program trace route which will not only show total time passing packets to speedtest.dc.utel.ua, but will also indicate to us the specific route along which packets will go to this node.

3 9 ms 12 ms 10 ms v64.TenGig3-1.opal.volia.net

4 10 ms 7 ms 8 ms v115.TenGig3-2.topaz.volia.net

5 8 ms 11 ms 11 ms utel-gw.ix.net.ua

6 11 ms 11 ms 14 ms dc-utel-m7i-2.utel.net.ua

7 13 ms 10 ms 11 ms speedtest.dc.utel.ua

The trace is complete.

As you can see, the results of running the program trace route also did not reveal any problems: between the user's computer and the node speedtest.dc.utel.ua there are 7 intermediate links (hops), the response time from each link is also quite acceptable - 8-13 milliseconds.

However, it should be noted here that, taking into account the high volatility (variability) of telecommunication channels, a single execution of commands ping and trace route can indicate the presence of problems only in the most extreme cases (complete loss of communication, large interference, etc.). Therefore, these programs are usually used only for an initial and quick assessment of the situation.

Windows has a program pathping- a utility for tracing a network route, which combines the functionality of utilities ping and tracert and has additional features. PathPing provides information about network latency (i.e., hidden, implicit processes) and data loss at intermediate nodes between the source and destination. Team pathping over a period of time (175 seconds) sends numerous ping requests to each router between the source and destination, and then calculates the results based on the packets received from each of them. Insofar as pathping shows the packet loss rate for each router or link, you can identify routers or subnets that are having network problems. Team pathping executes the equivalent of the command tracert action, identifying routers along the way.

Here is the output of the command pathping speedtest.dc.utel.ua

Route trace to speedtest.dc.utel.ua with max 30 hops:

3 v64.TenGig3-1.opal.volia.net

4 v115.TenGig3-2.topaz.volia.net

5 utel-gw.ix.net.ua

6 dc-utel-m7i-2.utel.net.ua

7 speedtest.dc.utel.ua

Calculation of statistics for: 175 sec. ...

Source node Route node

Hop MTU Lost/From Rev % Lost/From Rev % Address

3 12ms 0/ 100 = 0% 0/ 100 = 0% v64.TenGig3-1.opal.volia.net

0/ 100 = 0% |

4 16ms 0/ 100 = 0% 0/ 100 = 0% v115.TenGig3-2.topaz.volia.net

0/ 100 = 0% |

5 15 ms0/ 100 = 0% 0/ 100 = 0% utel-gw.ix.net.ua

1/ 100 = 1% |

6 10ms 1/ 100 = 1% 0/ 100 = 0% dc-utel-m7i-2.utel.net.ua

0/ 100 = 0% |

7 10 ms1/ 100 = 1% 0/ 100 = 0% speedtest.dc.utel.ua

The trace is complete.

AT in general terms repeating the results of the execution of previous commands, the program pathping showed very remarkable results. It turns out to be on the 5th and 7th sections of the packets (already in the area of ​​\u200b\u200bresponsibility of the company UTEL, which owns the server we are considering speedtest.dc.utel.ua) there is some packet loss within 1%. In itself, this value is small, but this fact requires additional study by the means available to us. Since the packet loss on the server to which we measure the speed of Internet access (using for example the site http://www.speedtest.net ) may indicate either a large load on this resource or other technical problems on it.

To continue measurements, we will use the MTR utility, another freely available program that performs tracert and ping for network diagnostics. Program website - http://www.bitwizard.nl/mtr. The Windows version (WinMTR) can be found here - Its peculiarity is that it traces the specified resource an unlimited number of times. Table 3 shows the meaning of the program columns.

Table 3

Column name

Meaning

Name/address of network equipment (server, router, etc.) through which packets pass

Sequence number of the equipment along the path to the end node

Packet Loss (%)

Number of sent packages, pcs

Number of packets received in response, pcs

The value of the best ping at the specified point of the route (msec)

The value of the worst ping at the specified point of the route (msec)

The value of the average ping at the specified point of the route (msec)

The result of the last (current) ping at the specified route point (msec)

Before starting the program, you need to select in its settings (Options) the time interval (Interval), for example, 1 sec and the packet size (Ping size) - 500 bytes. Thus, traffic to the studied node will be created, equal to only 4 kbit / s (= 500 * 8). The item "Resolve names" can be selected if it is necessary to display host names instead of their ip-address.

In the Host field, enter the name of the host - speedtest.dc.utel.ua. To reduce the statistical error, let's allow the program to send more than 100 packets (preferably 300-500). Figure 6 shows the result of WinMTR.

Fig.6.

In order to verify the reliability of the results obtained and their low dependence on a particular provider, simultaneously with the measurement from the network of the Volya company, a similar measurement was made from the network of the Beeline provider (Internet at home service). The result of this measurement is shown in Fig.7.

Fig.7.

From Fig.6 and 7 it is clearly seen that in the last sections (hops) of the path to the measured resource speedtest.dc.utel.ua(in the Beeline network, identified by the name rio.poltava.ua) packet loss is 17-18%. That is, such a number of sent packets are not returned back to the user. The reasons for this will be discussed briefly below.

For a more visual representation of network routes, packet transit time, and the level of their losses, you can also use the PingPlotter utility - http://www.pingplotter.com

Figure 8 shows the graphical results of measurements taken with its help on the route from the user's computer to the resource speedtest.dc.utel.ua from the network of the Volya provider, and in Fig. 9 - from the network of the Beeline provider. They also well illustrate the fact of losing many packets in the network section, which lies in the area of ​​​​responsibility of the Utel company and, probably, on the server itself speedtest.dc.utel.ua. Moreover, the results practically do not depend on the provider from whose network the measurements were made.

It is clearly seen that over 50% of packets are lost on the speedtest.dc.utel.ua resource at the time of measurement. Which is an extremely high value. And casts doubt on the correctness of the results of measuring the speed of access to speedtest.dc.utel.ua, shown in Figure 5.

Fig.8.

Fig.9.

It is very difficult for the end user to judge the reasons for such packet loss. In the general case, the reason that not all packets are returned to the recipient can be both the workload of the studied resource itself and critical interference on the communication lines. After all, transmission through the nodes of the Network is characterized not only by the duration of delays, but also by the probability of packet loss. The higher the traffic intensity, the higher the probability of packet loss. The results obtained may indicate a large load speedtest.dc.utel.ua.

This can also be judged by the magnitude of the so-called jitter (column jttr in Fig.8, 9), that is, the spread of the maximum and minimum packet transit time from its average value. Jitter is also most commonly caused by network delays due to crosstalk.

In the process of analyzing the reasons for the load of a particular resource (and this, as we found out above, is extremely important for an adequate understanding of the magnitude of the speed characteristics of access to this resource), it is necessary to take into account one more IMPLICIT factor that affects the load of the resource. This is the number and quality of sites located on the same server (on the same IP address). Let us explain what has been said on the example of the same speedtest.dc.utel.ua.

With the help of special service sites (for example: http://2ip.ru/server.php , http://ip-whois.net/site_one_ip.php , http://www.testip.ru/services/showsitesonip.html etc.) you can find out that on the same resource (server) on which the speedtest.dc.utel.ua work (at the time of writing) still 27 (!!!) various sites. And some of them have a fairly high attendance. Which undoubtedly contributes to the load on the server as a whole and is another reason for packet loss on the way, including to speedtest.dc.utel.ua.

Whatever the actual cause of this phenomenon in each particular case, we note that in the presence of such phenomena, it is NOT possible to consider reliable measurements of the speed to this resource. More precisely, let's say this: the measured speed is an objective indicator only taking into account all the real factors that influenced it at the time of measurement. By itself, the value of speed to a particular resource cannot speak separately neither about the quality of the line in general, nor about the quality of the last mile, nor about the presence of other significant factors. Only taking into account the entire complex of data can indicate the presence of problems in a certain area to the desired resource of the Network.

findings

1. The provider can guarantee the speed closest to the maximum declared only on the "last mile", that is, on the section of the telecommunications network from the subscriber's equipment (usually a DSL or cable modem, network card, etc.) to the point of inclusion of this subscriber on its technical site (provider's server, router, etc.)

2. Internet connection speed depends on a number of factors. The following are some of the bottoms that can be controlled and configured by the subscribers themselves:

      • the tariff plan used;

      • the presence of all kinds of viruses and similar programs, as well as programs that filter traffic.

        operating system settings and correct operation of hardware;

        network card driver settings;

        presence and configuration of intermediate network equipment (router, etc.)

        use of wireless internet (Wi-Fi)

        quality of the "last mile" (in the user's area of ​​responsibility - an apartment, a room, etc.)

3. The speed of access "to the Internet" is correctly measured to traffic exchange centers (in Ukraine it is UA-IX), taking into account a number of rules and regulations.

4. When carrying out all measurements, one can only talk about the value of the average velocity at a given moment of time up to a certain node. The measured speed is an objective indicator only taking into account all the real factors that influenced it at the time of measurement. By itself, the value of speed at a certain point in time to a specific resource cannot speak separately either about the quality of the line in general, or about the quality of the last mile, or about the presence of other significant factors. Only taking into account the whole complex of these and other data can indicate the presence of problems in a certain area to the desired Network resource. It also depends on the way of solving this or that problem.

Hello Irina!

Basically, you can watch movies online at any speed.

Another question is how much it will cost nerves and health at very low speeds.

I will give an example specifically for your reception speed in numbers. Do not delve into the numbers, the main thing is the meaning.

Judging by the designation 1.55 M b/s, you have a speed of 1.55 Megabyte per second. The capital letter "M" indicates this.

Movies online, depending on the format (in my "Additional Materials" there is a link to my lesson on this topic, on films, in particular) and quality, the size of an average-quality online movie can be from 300 megabytes to 5000 megabytes.

Let's take something in the middle, for example, the size of a movie is 1000 megabytes.

How is watching a movie online?

When you enter the page for watching a movie, a player appears on your screen for viewing with the buttons "Pause", "Playback", "Stop".

When you click "Play", the movie starts downloading to your computer.

As soon as a very small part of the movie, which the player is already capable of playing, is downloaded to your computer, the player (after a few seconds) starts showing you the movie.

And the rest of the movie continues to swing in the background.

It turns out that the whole movie has not yet been downloaded, and you are already watching it.

So the problem with low reception speeds is that while you are viewing the first downloaded part, the next one does not have time to download to your computer. Then the Brakes begin. What hits the nerves.

But in this case, there is a way out. The truth takes some time. You put the player on "Pause" and do other things. After 5-10 minutes, turn on playback and watch the movie normally.

Now for your speeds.

Roughly speaking, a 1000 megabyte movie should be downloaded for

1000Mb / 1.55Mb/s = 645 seconds = 10 minutes

Judging by my observations, this is a fairly comfortable viewing of an average quality movie.

Now a little fly in the ointment.

The reception speed of 1.55 Mb/s does not mean at all that the movie will be downloaded at that speed.

1. In addition useful information(of the film itself) there is a lot of service information in the communication channel.

2. A lot depends on the workload of the site from which you are watching the movie.

3. Much depends on the workload of the channels through which you receive information.

4. A lot depends on how loaded your computer is directly - how many programs and processes are loaded and how much they “eat away” the memory and resources of the processor itself.

5. What is the quality of the movie you are watching - low, medium or high.

In general, we make discounts for factors on which we do not depend and it turns out (according to experience) somewhere around 20 - 40 minutes.

But, that's also acceptable. After all, the player has a "Pause" button.

And finally. The whole movie, in fact, is not downloaded to the computer. Parts only. Viewed parts are automatically deleted.

Regards, Oleg

300 Mbps two spatial stream and 40 MHz channel for receiving and transmitting. The actual data transfer rate in a wireless network depends on the features and settings of the client equipment, the number of clients in the network, obstacles in the signal path, as well as the presence of other wireless networks and radio interference in the same range.

150 Mbps - maximum physical layer speed according to the IEEE 802.11n standard when connected with adapters using one spatial stream and 40 MHz channel for receiving and transmitting.

Let's start with the fact that many users are wrongly guided by the connection speed in megabits per second (Mbps), which is displayed in the line Speed (Speed) tab General (General) in the window State (status) wireless connection in operating system Windows.

Users mistakenly think that this value represents the actual throughput of a particular network connection. This figure is displayed by the wireless adapter driver and shows what physical layer connection speed is currently used within the selected standard, that is, the operating system reports only the current (instantaneous) physical connection speed of 300 Mbps, but the actual connection throughput during transmission data can be in the range from 50 to 140 Mbps, depending on the settings of the access point with support for 802.11n, the number of client wireless adapters connected to it at the same time, and other factors.
The difference between the connection speed shown in Windows and the actual connection speed is primarily due to high overhead, network packet loss in the wireless environment, and retransmission costs.

To get a more or less reliable value of the actual data transfer rate in a wireless network, you can use one of the following methods:

    Start copying a large file in Windows and then calculate the speed at which this file was transferred using the file size and transfer time (Windows 7 long copy to additional information windows calculates a fairly reliable speed).

  • Network administrators can recommend the program (cross-platform console client-server program) or (graphical shell of the Iperf console program).

We draw your attention to the following:
In the technical specifications of devices, the connection speed is indicated in Megabits per second (Mbps), and in user programs (Internet browsers, download managers, p2p clients), the data transfer rate when downloading files (upload speed) is displayed in Kilobytes or Megabytes per second (KB/s, KB/s or MB/s, MB/s). These values ​​are often confused.
To convert Megabytes to Megabits, you need to multiply the value in Megabytes by 8. For example, if the Internet browser shows a download speed of 4 MB / s, then to convert to Megabits, you need to multiply this value by 8: 4 MB/s * 8 = 32 MB/s.
To convert from Megabits to Megabytes, you need to divide the value in Megabits by 8.

But back to Wi-Fi connection speed.

In real conditions, the throughput and coverage of a wireless network will vary depending on interference from other devices, obstacles, and other factors. We recommend that you familiarize yourself with: "What affects the operation of wireless WiFi networks? What can be a source of interference and what are their possible reasons?"

As we wrote above, in the Windows operating system, as well as in the utilities supplied with the wireless adapter, when connected, not the real data transfer rate is displayed, but the theoretical speed. The actual data transfer rate turns out to be about 3 times lower than the one indicated in the specifications for the device.
The fact is that at any given time, the access point (Internet center with an active access point) works with only one client Wi-Fi adapter from the entire Wi-Fi network. Data transmission takes place in half-duplex mode, i.e. in turn - from the access point to the client adapter, then vice versa, and so on. Simultaneous, parallel data transfer process (duplex) in Wi-Fi technology is not possible.
If there are two clients in the Wi-Fi network, then the access point will need to switch twice as often as if there was one client, because. Wi-Fi technology uses half-duplex data transmission. Accordingly, the actual data transfer rate between two adapters will be two times lower than the maximum real rate for one client (we are talking about data transfer from one computer to another through an access point over a Wi-Fi connection).

Depending on the distance of the Wi-Fi network client from the access point or on the presence of various interferences and obstacles, the theoretical and, as a result, the actual data transfer rate will change. Together with wireless adapters, the access point changes the signal parameters depending on the conditions on the air (distance, the presence of obstacles and interference, radio noise, and other factors).

Let's take an example. The transfer speed between two laptops connected directly via Wi-Fi is ~10 MB/s (one of the adapters works in access point mode, and the other in client mode), and the data transfer speed between the same laptops, but connected via Keenetic Keenetic is ~4 MB/s. That's the way it should be. The speed between two devices connected via a Wi-Fi access point will always be at least 2 times less than the speed between the same devices connected to each other directly, because. the frequency band is one and the adapters will be able to communicate with the access point only one by one.

Let's consider another example, when a wireless Wi-Fi network is created in the Keenetic Lite router supporting the IEEE 802.11n standard with a possible theoretical maximum speed of up to 150 Mbps. A laptop with a Wi-Fi adapter of the IEEE 802.11n standard (300 Mbps) is connected to the router and desktop computer with a Wi-Fi adapter of the IEEE 802.11g standard (54 Mbps).
In this example, the entire network has a maximum theoretical speed of 150 Mbps, because it is built on a router with an IEEE 802.11n 150 Mbps access point. The maximum real Wi-Fi speed will not exceed 50 Mbps. Since all Wi-Fi standards operating on the same frequency range are backward compatible with each other, you can connect to such a network using a Wi-Fi adapter of the IEEE 802.11g standard, 54 Mbps. At the same time, the maximum real speed will not exceed 20 Mbps.