Unit of measurement Internet speed, how is Internet speed measured?

85 comments
Hello, dear readers of the site IT-uroki.ru!

You were probably interested in the speed of data transfer over the network (including the Internet), the speed of writing to a flash drive (or hard drive). Today we will understand the speed of information transfer in computer technology and find out how many megabytes are in a megabit !

The information from the previous lesson will be useful to you; if you haven’t read it yet, be sure to start there.

Let me remind you that in the last IT lesson we dealt with bits, bytes and multiple prefixes K, M, G, T and found out how many bytes are in a kilobyte (here is a link to lesson 15).

Do you remember? Then let's begin!

Other concepts related to Internet use

In addition to downloading and uploading, there are a number of other definitions that you need to know in order to avoid problems when working on the global network:

Oscillations. This is system instability that occurs in throughput, which negatively affects download and upload speeds, and can also distort the final indicators.
Congestion. This is a limitation of Internet bandwidth, in most cases caused by the presence of a large number of users in one channel at a given time.
LAN cable. This is a wire from the provider that connects to the corresponding connector on the motherboard of a desktop computer or laptop and provides an Internet connection.
Wi-Fi. This is a technology for wireless Internet data transmission.
Router. A router that allows you to share the Internet with other devices via Wi-Fi.

Modern household Wi-Fi routers are capable of distributing the Internet to 5 devices simultaneously. However, the overall network speed will drop, as it will be divided equally between connected devices.

Appearance of a Wi-Fi router.

Fourth gear - code MLT-3

The three-level transmission code MLT-3 ((Multi Level Transmission - 3) has much in common with the NRZ code. The most important difference is the three signal levels.

One corresponds to a transition from one signal level to another. The signal level changes sequentially, taking into account the previous transition. The maximum frequency of the signal corresponds to the transmission of a sequence of ones. When transmitting zeros, the signal does not change. Information transitions are recorded at the bit boundary. One signal cycle contains four bits.

The disadvantage of the MLT-3 code, as well as the NRZ code, is the lack of synchronization. This problem is solved by using a data transformation that eliminates long sequences of zeros and the possibility of desynchronization.

High-speed Internet via Wi-Fi

However, if you use a Wi-Fi router, the speed drop on the way to your devices via a wireless channel will in any case, to a greater or lesser extent, depending on your equipment. Speed losses in a wireless connection are caused by the use of limited channel width, encryption, and congestion of the channel you (or the router) have chosen by your neighbors.

In the first case, a special mode of a dual-band router (not every model supports it) will help increase the speed of receiving data from the Internet, in which two bands are used at once for the user’s output: 2.4 and 5 GHz.

If most of your neighbors use routers, then most of the transmission is carried out at a frequency of 2.4 Hz on standard channels 1/6/11. The choice of intermediate channels with less noisy neighbors will help improve the situation.

The wi-fi analyzer mobile application will allow you to select a more or less unloaded channel. However, it is worth remembering that for most client devices, the transmitter power is reduced by more than half on the “extreme” channels (1 and 11/13 for 2.4 GHz).

A more radical solution would be to buy a router that supports dual bands and uses the 5 GHz frequency. There are more channels to choose from in this range, but it is worth considering that its penetration ability through solid walls is much worse.

Here again you need to test. Our speed meter will help with this.

Read also: Color “Fire Red”— afbe, rgb(, , ), cmyk(, , , )

Wireless networks - how it works

Barker Codes

There are quite a lot of chip sequences that meet the specified autocorrelation requirements, but the so-called Barker codes are of particular interest to us, since they are used in the 802.11 protocol.

Barker codes have the best noise-like properties among known pseudo-random sequences, which has led to their widespread use.

The 802.11 family of protocols uses a Barker code that is 11 chips long (11100010010).

In order to transmit a signal, a logical one is transmitted by a direct Barker sequence, and a logical zero by an inverse sequence.

Speed 1 Mbps

The 802.11 standard provides two speed modes: 1 and 2 Mbit/s. To encode data at the physical layer, the DSSS method with 11-chip Barker codes is used. With an information speed of 1 Mbit/s, the speed of individual Barker sequence chips is 11×106 chip/s, and the spectrum width of such a signal is 22 MHz. Considering that the width of the frequency range is 83.5 MHz, we find that a total of 3 non-overlapping frequency channels can fit in this frequency range. The entire frequency range, however, is usually divided into 11 overlapping frequency channels of 22 MHz, spaced 5 MHz from each other. For example, the first channel occupies the frequency range from 2400 to 2423 MHz and is centered relative to the frequency of 2412 MHz. The second channel is centered at 2417 MHz, and the last, channel 11, is centered at 2462 MHz. When viewed this way, the first, sixth and 11th channels do not overlap with each other and have a 3 megahertz gap relative to each other. It is these three channels that can be used independently of each other.

Differential Binary Phase Shift Key (DBPSK) is used to modulate a sinusoidal carrier signal (a process necessary to inform the carrier signal). In this case, information encoding occurs due to a phase shift of the sinusoidal signal relative to the previous signal state. Binary phase modulation provides two possible phase shift values - 0 and π. Then a logical zero can be transmitted by an in-phase signal (the phase shift is 0), and a logical one can be transmitted by a signal that is phase shifted by π.

Speed 2 Mbps

An information speed of 1 Mbit/s is mandatory in the IEEE 802.11 standard (Basic Access Rate), but a speed of 2 Mbit/s (Enhanced Access Rate) is optionally possible. To transmit data at this speed, the same DSSS technology with 11-chip Barker codes is used, but Differential Quadrature Phase Shiftey is used to modulate the carrier wave. With relative quadrature phase modulation, the phase shift can take four different values: 0, π/2, π and 3π/2. Using four different signal states, it is possible to encode a sequence of two information bits (dibits) in one discrete state and thereby double the information transmission rate. For example, dibit 00 may correspond to a phase shift of 0; dibit 01 - phase shift equal to π/2; dibit 11 - phase shift equal to π; dibit 10 - phase shift equal to 3π/2.

In conclusion, considering the physical layer of the 802.11 protocol, we note that at an information speed of 2 Mbit/s, the speed of individual Barker sequence chips remains the same, that is, 11×106 chip/s, and therefore the width of the spectrum of the transmitted signal does not change.

Physical layer of the 802.11b/b+ protocol

The IEEE 802.11b protocol, adopted in July 1999, is a kind of extension of the basic 802.11 protocol and, in addition to speeds of 1 and 2 Mbit/s, provides speeds of 5.5 and 11 Mbit/s. To operate at speeds of 1 and 2 Mbit/s, spectrum spreading technology using Barker codes is used, and for speeds of 5.5 and 11 Mbit/s so-called complementary codes (Complementary Code Keying, CCK) are used.

CCK sequences

Complementary codes or CCK sequences have the property that the sum of their autocorrelation functions for any cyclic shift other than zero is always zero.

The IEEE 802.11b standard deals with complex complementary 8-chip sequences defined on a set of complex elements.

Here it is worth making a small lyrical digression so as not to alienate the reader by the complexity of the mathematical apparatus used. The mathematics of complex numbers can evoke a lot of negative memories, being associated with something completely abstract. But in this case everything is quite simple. A complex representation of a signal is just a convenient mathematical apparatus for representing a phase-modulated signal.

Using the set of complex elements {1, –1, j, –j}, it is possible to form eight complex numbers that are identical in magnitude but differ in phase. That is, the elements of the 8-chip CCK sequence can take one of the following eight values: 1, –1, j, –j, 1+j, 1–j, –1+j, –1–j. The main difference between CCK sequences and the previously discussed Barker codes is that there is not a strictly defined sequence through which either a logical zero or a one could be encoded, but a whole set of sequences. Considering that each element of an 8-sip sequence can take one of eight values depending on the phase value, it is clear that 88 = 16777216 sequence options can be combined, however, not all of them will be complementary. But even taking into account the requirement of complementarity, a fairly large number of different CCK sequences can be formed. This circumstance makes it possible to encode several information bits in one transmitted symbol and thereby increase the information transmission speed.

Generally speaking, using CCK codes allows you to encode 8 bits per character at 11 Mbit/s and 4 bits per character at 5.5 Mbit/s. In both cases, the symbolic transmission rate is 1.385 × 106 symbols per second (11/8 = 5.5/4 = 1.385), and taking into account that each character is specified by an 8-chip sequence, we obtain that in both cases the transmission speed of individual chips is 11×106 chips per second. Accordingly, the signal spectrum width at both speeds of 11 Mbit/s and 5.5 Mbit/s is 22 MHz.

Considering the possible transmission speeds of 5.5 and 11 Mbit/s in the 802.11b protocol, we have so far left without addressing the question of why a speed of 5.5 Mbit/s is needed if the use of CCK sequences allows data to be transmitted at a speed of 11 Mbit/s . Theoretically, this is true, but only if you do not take into account the interference environment. In real conditions, the noise level of transmission channels and, accordingly, the ratio of noise and signal levels may be such that transmission at a high information speed, that is, when many information bits are encoded in one symbol, may be impossible due to their erroneous recognition. Without going into mathematical details, we only note that the higher the noise level of communication channels, the lower the information transmission speed. It is important that the receiver and transmitter correctly analyze the interference environment and select an acceptable transmission rate.

Binary Burst Convolutional Coding PBCC

To further consider the 802.11b/b+ protocol, we need to familiarize ourselves with another type of encoding - the so-called Packet Binary Convolutional Coding (PBCC).

The idea behind convolutional coding is as follows. The incoming sequence of information bits is converted in a special convolutional encoder so that each input bit corresponds to more than one output bit. That is, the convolutional encoder adds certain redundant information to the original sequence. If, for example, each input bit corresponds to two output bits, then we talk about convolutional coding with a rate of r = 1/2. If every two input bits correspond to three output bits, then the convolutional coding speed will be 2/3.

Any convolutional encoder is built on the basis of several sequentially connected storage cells and logical elements that connect these cells to each other. The number of storage cells determines the number of possible encoder states. If, for example, a convolutional encoder uses six memory cells, then the encoder stores information about six previous signal states, and taking into account the value of the input bit, we find that such an encoder uses seven bits of the input sequence. Such a convolutional encoder is called a seven-state encoder (K = 7).

The output bits generated in a convolutional encoder are determined by the values of the input bit and the bits stored in the storage cells, that is, the value of each output bit generated depends not only on the incoming information bit, but also on several previous bits.

PBCC technology uses seven-state convolutional encoders (K = 7) with a rate of r=1/2. The main advantage of convolutional encoders is the noise immunity of the sequence they generate. The fact is that with redundant coding, even in the event of reception errors, the original bit sequence can be accurately restored. A Viterbi decoder is used at the receiver side to restore the original bit sequence.

The dibit generated in the convolutional encoder is subsequently used as a transmitted symbol, but this dibit is first subjected to phase modulation. Moreover, depending on the transmission speed, binary, quadrature or even eight-position phase modulation is possible.

Packet convolutional coding is optionally provided as an alternative coding method in the 802.11b protocol at 5.5 and 11 Mbps bit rates. In addition, it is this encoding mode that forms the basis of the 802.11b+ protocol, an extension of the 802.11b protocol. Actually, the 802.11b+ protocol as such does not officially exist, but this extension is supported by many manufacturers of wireless devices. The 802.11b+ protocol provides another data transfer rate of 22 Mbit/s using PBCC technology.

At a transmission rate of 5.5 Mbit/s, binary phase modulation is used to modulate the bit generated by the convolutional encoder, and at a transmission rate of 11 Mbit/s, quadrature phase modulation is used. At the same time, for a speed of 11 Mbit/s, one input bit is encoded in each symbol and the bit rate corresponds to the symbol transmission rate, and at a speed of 5.5 Mbit/s, the bit rate is equal to half the symbol rate (since each input bit in this case corresponds to two output symbols). Therefore, for both 5.5 Mbit/s and 11 Mbit/s, the symbol rate is 11x106 symbols per second.

For a speed of 22 Mbit/s, compared to the PBCC scheme we have already considered, data transmission has two features. First of all, 8-position phase modulation (8-PSK) is used, that is, the signal phase can take on eight different values, which allows 3 bits to be encoded in one symbol. In addition, in addition to the convolutional encoder, a puncture encoder (Puncture) has been added to the circuit. The meaning of this solution is quite simple: the redundancy of the convolutional encoder, equal to 2 (for each input bit there are two output bits), is quite high and under certain noise conditions it is unnecessary, so the redundancy can be reduced so that, for example, every two input bits correspond to three output bits .

For this, you can, of course, develop an appropriate convolutional encoder, but it is better to add a special puncture encoder to the circuit, which will simply destroy extra bits.

Assume that the puncture encoder removes one bit from every four input bits. Then every four incoming bits will correspond to three outgoing ones. The speed of such an encoder is 4/3.

If such an encoder is used in conjunction with a convolutional encoder with a speed of 1/2, then the total encoding speed will be 2/3, that is, for every two input bits there will be three output bits.

Having understood the principle of operation of the puncture encoder, let's return to the consideration of PBCC encoding at a speed of 22 Mbit/s in the 802.11b+ protocol.

The convolutional encoder (K = 7, R = 1/2) receives data at a speed of 22 Mbit/s. After adding redundancy in the convolutional encoder, the 44 Mbps bits are fed into a 4:3 puncture encoder, which reduces the redundancy so that for every four input bits there are three output bits. Consequently, after the puncture encoder, the flow rate will already be 33 Mbit/s (not the information speed, but the total speed, taking into account the added redundant bits). The resulting sequence is sent to an 8-PSK phase modulator, where every three bits are packed into one symbol. In this case, the transmission speed will be 11×106 characters per second, and the information speed will be 22 Mbit/s (Fig. 2).

How to test the speed of your Internet connection online?

In order to test your bandwidth, you can use one of the free Internet speed measurement resources: or.

Both sites measure the bandwidth from the server you choose to the computer on which the speed is measured. Since the length of the communication channel can be from several hundred meters to several thousand kilometers, it is recommended to choose the geographically closest server (although it may also be heavily loaded). It is better to conduct testing at a time when the activity of the provider's network clients is the least (for example, in the morning or late at night). The accuracy of Internet connection speed measurements is not ideal due to the large number of different factors that greatly affect the throughput, but it is quite capable of giving an idea of the real Internet connection speed.

The Internet provider allocates bandwidth to each subscriber for Internet access in accordance with the subscriber's tariff plan (the provider "cuts" the speed according to the tariff plan). However, many Internet browsers, as well as file download wizards and torrent clients, display the bandwidth of the communication channel not in megabits per second, but in megabytes per second, and this often causes confusion.

Let's test the speed of the Internet connection using the example of the resource. You need to click the “BEGIN TEST recommended server” button.

The resource will automatically select the server closest to you and begin testing Internet speed. The test result will be the channel throughput from the provider to the subscriber (“DOWNLOAD SPEED”) and the channel throughput from the subscriber to the provider (“UPLOAD SPEED”), which will be expressed in Megabits per second.

Encoding mechanism

Digital data transfer requires several mandatory operations:

synchronization of the clock frequency of the transmitter and receiver;
converting a sequence of bits into an electrical signal;
reducing the frequency of the electrical signal spectrum using filters;
transmission of reduced spectrum over a communication channel;
signal amplification and restoration of its shape by the receiver;
Converting an analog signal to a digital one.

Let's consider the relationship between clock frequency and bit sequence. The bit stream is transmitted at a speed determined by the number of bits per unit time. In other words, bits per second is the number of discrete signal changes per unit of time. Clock frequency, measured in hertz, is the number of sinusoidal changes in a signal per unit time.

This obvious correspondence has given rise to a misconception about the adequacy of the values of hertz and bits per second. In practice, everything is more complicated. The data transfer rate is usually higher than the clock frequency. To increase the transmission speed, the signal can go in parallel over several pairs. Data can be transmitted in bits or bytes. The encoded signal can have two, three, five or more levels. Some signal encoding methods require additional data encoding or synchronization, which reduces the transmission speed of information signals.

As can be seen from the table, there is no one-to-one correspondence between MHz and Mbit/s.

Channel category	3	4	5	5e	6
Frequency range, MHz	16	20	100	125	200
Maximum data transfer rate, Mbit/s	100	100	155/1000	1000	No protocols

Table 1. Relationship between channel category, frequency range and maximum data rate

Each protocol requires a certain spectrum width, or, if you prefer, the width of the information highway. Coding schemes are becoming more complex in order to make better use of information highways. As in the analogy with the engine, it is not at all necessary to spin it up to maximum speed; it is more advisable to engage the gear.

What determines the speed of an Internet connection?

The speed of the Internet connection depends on the tariff plan set by the provider.
The speed is also affected by the technology of the information transmission channel and the load on the Network by other users. If the overall channel capacity is limited, then the more users are on the Internet and the more information they download, the more the speed drops, since there is less “free space” left.
There is also a dependence on the loading speed of the sites you access. For example, if at the time of loading the server can provide data to the user at a speed of less than 10 Mbit/sec, then even if you have the maximum tariff plan, you will not achieve more.

Factors that also affect Internet speed:

When checking, the speed of the server you are accessing.
Setting up and speed of a Wi-Fi router if you are connected to the local network through it.
At the time of scanning, all programs and applications running on the computer.
Firewalls and antiviruses that run in the background.
Settings for your operating system and the computer itself.

Trunk Width - Required Frequency Band

The speed of movement depends not only on the capabilities of the car, but also on the quality of the highway. The same is true for data transmission. Let's consider the possibilities of information highways.

Signal encoding is a method of converting clock frequency into data rate. What is the purpose of the transformation? In order to increase the speed without changing the frequency range of the communication channel. Coding requires the use of more complex transmitting and receiving equipment. This is a minus. But when moving to higher speed protocols, you can use the same cables. And this is already a big plus.

For example, Fast Ethernet 100 Base T4 provides network speeds of 100 Mbps over Category 3 (16 MHz) cables. Gigabit Ethernet 1000 Base T is implemented in such a way that, based on category 5 channels (100 MHz), which has some reserve, it can transmit 1000 Mbit/s.

Megabits to Megabytes

This is where the real confusion begins. This is a very popular question that people contact our support team with. And it sounds something like this: I have a tariff of 80 MB/s, but when downloading files, the download speed does not rise above 10 MB/s. Why is that?

Let's figure it out. As we already understood, providers provide data exchange speeds in megabits per second. Megabits per second can be written as either Mbit/s or Mb/s .

And programs for downloading files (download managers, torrents, file sharing services) often display the speed not in megabits per second, but in megabytes per second! Megabytes per second can be written as MB/s or MB/s . Notice the difference?

When writing Mbit/s and MB/s, the difference is obvious. But when writing Mb/s and MB/s, it seems that this is the same designation. But the real difference is in the spelling of the letter B.

In megabits it is capitalized, and in megabytes it is capitalized. That's why:

Mb/s is megabits per second

MB/s is megabytes per second.

What is the difference between bytes and bits?

Binary prefixes for speed measurement

Binary prefixes are used to measure data transfer speed much less frequently. But if you do come across it, here are signs that will help you figure it out:

Binary prefixes are used to measure data transfer rate (bit/s)

Binary prefixes are used to measure data transfer speed (bytes/s)

The main problem is that the use of binary prefixes is not always indicated ! And it may turn out that behind the letter “M” lies not “mega”, but “mebi”.

If you have problems with binary prefixes and the difference between kilobits and kibibits, re-read the last lesson, everything will fall into place.

Is it possible to increase the speed

There are several tips to help increase the speed of your Internet connection:

Reduce browser load. Do not leave many tabs open, disable programs built into the browser. Try to use them only when needed.
Clean your browser. Clearing cache and search history also lightens the load. This can be done manually or using CCleaner. Restarting the browser also helps.
Install Adblock. Pop-ups and advertisements take away some of your download speed. Adblock will help you cope with this.
Reduce video quality. It is not necessary to choose the minimum 144-240p. You can sacrifice picture quality for the sake of uninterrupted playback if the video is turned on in the background.
Enable turbo mode. It is called differently in each browser. It is convenient when working with sites full of images. These settings work based on a compression algorithm.
Disable downloads. In the case of standard loading, the problem of low speed cannot be solved in any way. When using Torrent, you should disable already downloaded files.
Check cable integrity. Often the reason for packet loss lies in a damaged network cable or a broken connector (a working connector should fit into the socket with a characteristic click).
Disable background programs. You can find out which programs load the network through the command line (Win + R). In the line you need to type the command netstat -b 5 > activity.txt. After this, a list of applications will appear in the menu. Then you need to type Ctrl+C, after which the folder with the connected software will appear. Enter activity.txt into the command line again, press Enter and disable unnecessary programs.

What is uploading and loading

When using the Internet, a person often comes across terms such as downloading and uploading. The difference between these concepts is not always clear to the average user. To fully understand the topic of the article, it is necessary to understand this point.

Loading is a value that characterizes the speed of opening resources, sites, downloading various types of information, for example, films, TV series, pictures, etc.

Upload is an indicator that characterizes the speed of uploading personal information of a computer user to the global network. For example, when a person uploads his own photos to social networks, he posts personal data online. Downloading must be faster than uploading, otherwise the Internet will be practically unusable.

Uploading files by the user to the Internet.