Gigabit Ethernet. Gigabit Ethernet Shielded Twisted Pair

No sooner had the milk on the lips of the newly born Fast Ethernet standard, as they say, had yet been dried before the 802 committee began work on new version(1995). It was dubbed Gigabit Ethernet almost immediately, and in 1998 new standard has already been ratified by the IEEE under the official name 802.3z. Thus, the developers emphasized that this is the latest development in the 802.3 line (unless someone urgently comes up with calling the standards, say, 802.3y. At one time, Bernard Shaw proposed to expand the English alphabet and include in it, in particular, the letter "S", but was not convincing.).

The main prerequisites for 802.3z were the same as when 802.3u was created - to increase the speed by 10 times while maintaining backward compatibility with legacy Ethernet networks. Specifically, Gigabit Ethernet had to provide unacknowledged datagram service in both one-way and multicast. At the same time, it was necessary to keep the 48-bit addressing scheme and frame format unchanged, including the upper and lower limits of its size. The new standard fulfills all of these requirements.

Gigabit Ethernet networks are built on a point-to-point basis, they do not use a mono channel, as in the original 10-megabit Ethernet, which, by the way, is now called classic Ethernet. The simplest gigabit network shown in diagram "a" consists of two computers directly connected to each other. In a more general case, however, there is a switch or hub to which many computers are connected; it is also possible to install additional switches or hubs (scheme "b"). But in any case, two devices are always connected to one Gigabit Ethernet cable, no more, no less.

Gigabit Ethernet can operate in two modes: full duplex and half duplex. "Normal" is considered full duplex, and traffic can go simultaneously in both directions. This mode is used when there is a central switch connected to peripheral computers or switches. In this configuration, all lines are buffered so that subscribers can send data whenever they want. The sender is not listening to the channel because he has no one to compete with. On the line between the computer and the switch, the computer is the only potential sender; the transmission will be successful even if the switch is transmitting at the same time (full-duplex line). Since there is no competition in this case, the CSMA / CD protocol is not applied, therefore the maximum cable length is determined solely by the signal strength, and the issues of the propagation time of the noise burst do not arise here. Switches can operate at mixed speeds; moreover, they automatically select the optimum speed. Self-tuning is supported in the same way as Fast Ethernet.

Half-duplex operation is used when computers are not connected to a switch, but to a hub. The hub does not buffer incoming frames. Instead, it electrically connects all lines, simulating a conventional Ethernet mono link. Collisions are possible in this mode, so CSMA / CD is used. Since a frame of the smallest size (i.e. 64-byte) can be transmitted 100 times faster than in a classic Ethernet network, the maximum segment length must be correspondingly reduced by a factor of 100. It is 25 m - it is at this distance between the stations that the noise burst is guaranteed to reach the sender before the end of its transmission. If the cable had a length of 2500 m, then the sender of a 64-byte frame at 1 Gbit / s would have time to do a lot even while its frame only traveled a tenth of the way in one direction, not to mention that the signal should also come back.

The 802.3z development committee quite rightly noted that 25 m is an unacceptably short length, and introduced two new features that allowed to expand the radius of the segments. The first is called media expansion. This extension consists only in the fact that the equipment inserts its own padding field, which stretches a normal frame to 512 bytes. Since this field is added by the sender and removed by the recipient, the software has nothing to do with it. Of course, wasting 512 bytes to transfer 46 bytes is somewhat wasteful in terms of bandwidth efficiency. The transmission efficiency is only 9%.

The second property that allows you to increase the allowable segment length is bursting frames. This means that the sender can send not a single frame, but a packet that combines many frames at once. If the total packet length is less than 512 bytes, then, as in the previous case, hardware filling with dummy data is performed. If there are enough frames waiting for transmission to fill such a large packet, then the system operation turns out to be very efficient. This is, of course, preferable to media expansion. These methods made it possible to increase the maximum segment length to 200 m, which is probably already quite acceptable for organizations.

It is difficult to imagine an organization that would spend a lot of effort and money installing cards for a high-performance gigabit Ethernet network, and then connecting computers with hubs that simulate the work of classic Ethernet with all its collisions and other problems. Hubs are, of course, cheaper than switches, but Gigabit Ethernet interface cards are still relatively expensive, so the savings on buying a hub instead of a switch are not worth it. In addition, it drastically reduces performance, and it becomes generally unclear why the money was spent on gigabit boards. However, backward compatibility is something sacred in the computing industry, so 802.3z does it anyway.

Gigabit Ethernet supports both copper and fiber optic cables. Running at 1 Gbps means the light source has to be turned on and off approximately once every nanosecond. LEDs simply cannot work that fast, so lasers have to be used here. The standard provides for two operating wavelengths: 0.85 µm (short wavelength) and 1.3 µm (long). Lasers rated at 0.85 microns are cheaper, but do not work with singlemode cables.

Three fiber diameters are officially allowed: 10, 50 and 62.5 microns. The first is for single mode transmission, the other two for multimode transmission. Not all of the six combinations are allowed, and the maximum segment length depends on the selected combination. The numbers in the table are the best case. In particular, the 5 km cable can only be used with a 1.3 µm laser working with 10 µm single mode fiber. This option, apparently, is the best for highways of various kinds of campuses and industrial areas. It is expected to be the most popular despite being the most expensive.

1000Base-CX uses a short shielded copper cable. The problem is that it is compressed by competitors both from the top (1000Base-LX) and from the bottom (1000Base-T). As a result, it is doubtful that he will gain widespread public acceptance.

Finally, another cable option is a bundle of four unshielded twisted pairs. Since such wiring is almost ubiquitous, it looks like it will be the most popular Gigabit Ethernet.

The new standard uses new rules for coding signals transmitted over optical fibers. Manchester code at a data rate of 1 Gbps would require a signal change rate of 2 Gbaud. It is too complex and takes up too much bandwidth. Instead of Manchester coding, a scheme called 8V / 10V is used. As you might guess from the name, each byte consisting of 8 bits is encoded for transmission over the fiber in ten bits. Since 1024 resultant codewords are possible for each input byte, this method allows some freedom in the choice of codewords. The following rules are taken into account:

No codeword should have more than four identical bits in a row;

No codeword should contain more than six zeros or six ones.

Why exactly these rules?

First, they provide enough state changes in the data stream to keep the receiver in sync with the transmitter.

Secondly, the number of zeros and ones is trying to roughly equalize. In addition, many of the incoming bytes have two possible codewords associated with them. When the encoder has a choice of codewords, it is likely to choose the one that will equalize the number of zeros and ones.

A balanced number of zeros and ones is therefore given such importance that it is necessary to keep the constant component of the signal as low as possible. Then she will be able to pass through the converters without changes. Computer science people are not happy with the fact that converting devices dictate certain rules for coding signals, but life is life.

Gigabit Ethernet, built on 1000Base-T, uses a different encoding scheme, since it is difficult to change the signal state within 1 ns for a copper cable. Here, 4 twisted pairs of category 5 are used, which makes it possible to transmit 4 symbols in parallel. Each character is encoded with one of five voltage levels. Thus, one signal can mean 00, 01, 10 or 11. There is also a special, service voltage value. There are 2 data bits per twisted pair, respectively, in one time interval the system transmits 8 bits over 4 twisted pairs. The clock frequency is 125 MHz, which allows you to work at a speed of 1 Gbps. A fifth voltage level has been added for special framing and control purposes.

1 Gbps is quite a lot. For example, if the receiver is distracted by some business for 1 ms and at the same time forgets or does not have time to free the buffer, this means that it "sleeps" about 1953 frames. There may be another situation: one computer outputs data over a gigabit network, while the other receives them over classic Ethernet. Probably, the first will quickly fill up the data of the second. The clipboard will overflow first. Based on this, it was decided to implement flow control into the system (this was the case with fast Ethernet, although these systems are quite different).

To implement flow control, one of the parties sends a service frame informing that the other party needs to pause for a while. Service frames are, in fact, ordinary Ethernet frames, in the Type field of which 0x8808 is written. The first two bytes of the data field are command, and the next, if necessary, contain the command parameters. To control the flow, frames of the PAUSE type are used, with the duration of the pause in units of the transmission time of the minimum frame specified as a parameter. For Gigabit Ethernet, this unit is 512 ns, and pauses can last up to 33.6 ms.

Gigabit Ethernet has been standardized and the 802 committee has become bored. Then the IEEE asked him to start work on 10 Gigabit Ethernet. Long attempts began to find in the English alphabet any letter after z. When it became obvious that such a letter does not exist in nature, it was decided to abandon the old approach and go to two-letter indices. So in 2002 the 802.3e standard appeared. Apparently, the emergence of 100-gigabit Ethernet is also just around the corner.

Decide if you need to improve your network.

• If you and your family members regularly upload large files, stream media on the Internet, or perform other tasks that heavily load your network, such as a file-hosted server, or play a game of Online Games, you would love to invest in an upgrade to Gigabit Ethernet.
• Medium and large enterprises require many users to be connected over the network and at the same time be able to increase their productivity.
• Individuals who use the Internet alone for non-resource-intensive network tasks like Email, instant messaging, or web surfing may not see any benefit in improving network access to Gigabit Ethernet.

Introduction

Networks based on 10/100 Mbps Ethernet will be more than enough for any task in small networks. But what about the future? Have you thought about streaming video over your home's network? Will 10/100 Ethernet cope with them?

In our first article on Gigabit Ethernet, we'll take a closer look at it and determine if you need it. We'll also try to find out what you need to create a gigabit-ready network and take a quick tour of gigabit equipment for small networks.

What is Gigabit Ethernet?

Gigabit Ethernet is also known as gigabit over copper or 1000BaseT... It is a regular Ethernet version operating at speeds up to 1,000 megabits per second, which is ten times faster than 100BaseT.

Gigabit Ethernet is based on the IEEE standard 802.3z which was approved in 1998. However, in June 1999, an addendum came out to it - the standard of gigabit Ethernet over copper twisted pair. 1000BaseT... It was this standard that was able to bring Gigabit Ethernet out of server rooms and backbones, ensuring its use in the same conditions as 10/100 Ethernet.

Prior to 1000BaseT, Gigabit Ethernet required the use of fiber-optic or shielded copper cables, which are hardly convenient for conventional wiring. local area networks... These cables (1000BaseSX, 1000BaseLX and 1000BaseCX) are still used in special applications today, so we will not cover them.

The 802.3z Gigabit Ethernet group has done an excellent job of releasing a universal standard ten times faster than 100BaseT. 1000BaseT is also backward compatible with 10/100 hardware, it uses CAT-5 cable (or higher category). By the way, today a typical network is built on the basis of the fifth category cable.

Do we need it?

The first literature on Gigabit Ethernet pointed to the enterprise market as an area of ​​application for the new standard, and most often to data warehouse connectivity. Since Gigabit Ethernet provides ten times the bandwidth of the traditional 100BaseT, a natural application of the standard is to connect high-bandwidth sites. It is the communication between servers, switches, and backbones. This is where Gigabit Ethernet is needed, needed and useful.

As the price of gigabit hardware declined, the scope of 1000BaseT expanded to include "power users" and workgroup computers using "bandwidth-hungry applications."

Since most small networks have modest data needs, they are unlikely to ever need 1000BaseT network bandwidth. Let's take a look at some typical small network applications and assess their need for Gigabit Ethernet.

Do we need him, continuation

• Transferring large files over the network

  Such an application is typical, rather, for small offices, especially in companies dealing with graphic design, architecture or other business related to processing files of tens to hundreds of megabytes in size. You can easily calculate that a 100MB file will be transferred over a 100BaseT network in just eight seconds [(100MB x 8bit / byte) / 100Mbps]. In reality, many factors degrade the transfer speed, so your file will take a little longer to transfer. Some of these factors are related to the operating system, the applications running, the amount of memory on your computers, processor speed, and age. (The age of the system affects the speed of the buses on the motherboard.)

  Another important factor is the speed of network equipment, and the move to gigabit equipment can eliminate potential bottlenecks and speed up the transfer of large volumes of files. Many will argue that getting speeds above 50 Mbps on a 100BaseT network is far from trivial. Gigabit Ethernet, on the other hand, will be able to provide throughput above 100 Mbps.

• Network redundancy devices

  You can think of this case as a variant of "large files". If your network is configured to back up all computers to a single file server, then Gigabit Ethernet will speed up the process. However, there is also a pitfall - an increase in the "pipe" of transmission to the server may not lead to a positive effect if the server does not have time to process the incoming data stream (this also applies to the backup media).

  To benefit from a high-speed network, you should equip your server with more memory and back up to a fast hard drive rather than tape or CDROM. As you can see, you need to thoroughly prepare for the transition to Gigabit Ethernet.

• Client-server applications

  This area of ​​application is again more common in small business networks than in home networks. A large amount of data can be transferred between the client and the server in such applications. The approach is the same: you need to analyze the amount of transmitted network data to see if the application can keep up with the increase in network bandwidth and if this data is enough to load the Gigabit Ethernet.

In truth, we believe that most home network builders are unlikely to find sufficient reason to buy gigabit equipment. In small business networks, moving to gigabit can help, but we recommend that you analyze the amount of data transferred first. Everything is clear with the current state. But what if you want to take into account the possibility of future upgrades. What do you need to do today to be ready for it? In the next part of our article, we will look at the changes that need to be made to the most expensive, most often and most time consuming, part of the network - cable.

Gigabit Ethernet Cable

As we mentioned in the introduction, one of the key requirements of the 1000BaseT standard is the use of Category 5 (CAT 5) or higher cable. That is, Gigabit Ethernet can work on the existing cable structure of the 5th category... Agree, this opportunity is very convenient. As a rule, all modern networks use Category 5 cable unless your network was installed in 1996 or earlier (the standard was approved in 1995). However, here exists several pitfalls.

• Four pairs required

  As seen from of this article 1000BaseT uses all four pairs of Category 5 (or higher) cable to create four 250 Mbps links. (Another coding scheme is also used - five-level pulse amplitude modulation - to stay within the 100 MHz CAT5 frequency range). As a result, we can use the existing CAT 5 cabling structure for Gigabit Ethernet.

  Since 10 / 100BaseT only uses two out of four CAT 5 pairs, some people did not plug in extra pairs when laying their networks. Pairs were used, for example, for a telephone or for Power over Ethernet (POE). Fortunately, gigabit NICs and switches are smart enough to fall back to 100BaseT if all four pairs are unavailable. Therefore, in any case, your network will work with gigabit switches and network cards, but you will not get high speed for the money you paid.

• Don't use cheap connectors

  Another problem for amateur networkers is poor crimping and cheap wall sockets. They lead to impedance mismatches, resulting in return loss and, as a result, reduced bandwidth. Of course, you can try head-on search for the cause, but you'd better get a network tester that can detect return loss and crosstalk. Or just put up with the low speed.

• Length and topology restrictions

  1000BaseT is limited to the same maximum segment length as 10 / 100BaseT. Thus, the maximum network diameter is 200 meters (from one computer to another through one switch). For the 1000BaseT topology, the same rules apply as for 100BaseT, except that only one repeater per network segment (or, more accurately, one “half-duplex collision domain”) is allowed. But since Gigabit Ethernet does not support half-duplex transmission, you can forget about the last requirement. In general, if your network was doing fine under 100BaseT, you should have no problem moving to gigabit.

Gigabit Ethernet Cable Continued

It is best to use a cable for laying new networks. CAT 5e... Although CAT 5 and CAT 5e both pass 100 MHz, CAT5e cable is manufactured with additional parameters important for better transmission of high frequency signals.

Review the following Belden documents to learn more about CAT 5e cable specifications (in English):

Although a modern CAT 5 cable will work just fine with 1000BaseT, you might be better off choosing CAT 5e if you want to guarantee high bandwidth. If you're hesitant, estimate the cost of a CAT 5 and CAT 5e cable and go your way.

The One Thing You Should Avoid Is Purchase Recommendations CAT 6 cable for gigabit Ethernet. CAT 6 was added to the TIA-568 standard in June 2002 and it skips frequencies up to 200 MHz... Sellers will most likely persuade you to buy the more expensive sixth category, but you will only need it if you plan to build a network. 10 Gbps Ethernet over copper wiring, which is hardly realistic at the moment. What about CAT 7 cable? Forget about it!

If you have a good amount, then it is better to spend it on network specialist which possesses sufficient experience in laying gigabit networks... A specialist will be able to correctly lay cables or check your existing network to work with gigabit Ethernet. When installing a CAT 6 cable, we highly recommend that you seek professional help, as this cable specifies the bend radius and special high-quality connectors.

Gigabit equipment

In a way, the question of "gigabit or not" could have been the subject of controversy a year or a couple of years ago. From the point of view of a SOHO buyer, the transition from 10 to 10/100 Mbps has already happened. New computers are equipped with 10/100 Ethernet ports, routers already use built-in 10/100 switches rather than 10BaseT hubs. However, such a change is not a consequence of the demands and wishes of home networkers. They are content with existing equipment.

For these changes, we should thank the corporate users who buy only 10/100 equipment in bulk today, which allows us to lower prices for it. Once consumer equipment makers discovered they could use 10BaseT chips versus 10/100 options expensive, they did not hesitate for a long time.

Thus, yesterday's 10BaseT hub architecture has quietly migrated to today's 10/100 switched networks. We will experience the same transition from 10/100 to 10/100/1000 Mbps. And although there is still a year or two left before the tipping point, the transition already started and prices continue to fall steadily.

All you need is to buy a gigabit network card and a gigabit switch. Let's take a closer look at them.

• Network cards

  Branded 32-bit PCI 10/100 / 1000BaseT network cards such as Intel PRO1000 MT, Netgear GA302T and SMC SMC9552TX cost from $ 40 to $ 70 on the Internet. Products from second-tier manufacturers are about $ 5 cheaper. And while gigabit NICs are about two and a half times more expensive than the average 10/100 cards, your wallet is unlikely to notice any difference at all, unless you buy them in bulk.

  You can find network cards that support not only the 32-bit PCI bus, but also the 64-bit one, but they also cost more. What you won't see are CardBus adapters for your laptops. For some reason, manufacturers believe that notebooks do not need gigabit networks at all.

• Switches

  But the price of 10/100/1000 switches makes you think ten times about the feasibility of switching to gigabit Ethernet. The good news is that transparent gigabit switches are now available, which are much cheaper than their managed counterparts for the enterprise market.

  A simple four-port 10/100/1000 Netgear GS104 switch can be purchased for less than $ 225. If you opt for lesser-known brands like TRENDnet's TEG-S40TXE, you can reduce the cost to $ 150. Few four ports - please. The eight-port version of the Netgear GS108 will set you back about $ 450, and the TRENDnet TEG-S80TXD about $ 280.

  Considering that a 5-port 10/100 switch costs only $ 20 today, the prices for gigabit will seem too high to some. But remember, until recently, you could only buy Managed Gigabit Switches at $ 100 + per port. Prices are heading in the right direction!

Do you have to change computers?

Here's a little secret to Gigabit Ethernet: Under Win98 or 98SE, you probably won't get any benefit from Gigabit speed. While you can try to improve throughput by editing the registry, you still don't get a significant performance boost over your current 10/100 hardware.

The problem lies in the Win98 TCP / IP stack, which was not designed with high-speed networking in mind. The stack has problems even using 100BaseT networks, what can we say about gigabit communication then! We'll come back to this issue in the second article, but for now, you should only consider Win2000 and WinXP to work with gigabit Ethernet.

With the last sentence we are by no means not assumes that only Windows 2000 and XP support gigabit network cards. We just haven't tested the performance on other operating systems, so please refrain from sarcastic comments!

If you're wondering if you'll have to throw out your good old computer and buy a new one to use Gigabit Ethernet, then our answer is “maybe”. Judging by our practical experience, one hertz of "modern" processors equals one bit per second of network bandwidth... One of the manufacturers of gigabit networking equipment agreed with us: any machine with clock frequency 700 MHz or lower will not be able to fully utilize the bandwidth of the Gigabit Ethernet. So even with the right operating system, old computers are gigabit Ethernet like a dead poultice. You will sooner see speeds 100-500 Mbps

The modern world is increasingly becoming dependent on the volume and flow of information going in various directions via wires and without them. It all started a long time ago and with more primitive means than today's achievements of the digital world. But we do not intend to describe all the types and methods by which one person brought the necessary information to the consciousness of another. In this article, I would like to offer the reader a story about the not so long ago created and now successfully developing standard for transmitting digital information, which is called Ethernet.

The birth of the very idea and the Ethernet technology took place within the walls of the Xerox PARC corporation, along with other first developments in the same direction. The official date for the invention of Ethernet was May 22, 1973, when Robert Metcalfe wrote a memo for the head of PARC on the potential of Ethernet technology. However, it was patented only a few years later.

In 1979, Metcalfe left Xerox and founded 3Com, whose main focus was to promote computers and local area networks (LANs). With the support of such renowned companies as DEC, Intel and Xerox, the Ethernet standard (DIX) was developed. After its official publication on September 30, 1980, it began a rivalry with two large patented technologies - token ring and ARCNET, which were later completely replaced, due to their lower efficiency and higher cost than Ethernet products.

Initially, according to the proposed standards (Ethernet v1.0 and Ethernet v2.0), they were going to use coaxial cable as a transmission medium, but later they had to abandon this technology and switch to using optical cables and twisted pair.

The main advantage at the beginning of the development of Ethernet technology was the method of access control. It implies multiple connections with carrier sense and collision detection (CSMA / CD, Carrier Sense Multiple Access with Collision Detection), the data transfer rate is 10 Mbps, the packet size is from 72 to 1526 bytes, it also describes the data encoding methods ... The limit value of workstations in one shared network segment is limited to 1024, but other smaller values ​​are possible when setting more stringent limits for the thin coaxial segment. But such a construction very soon became ineffective and was replaced in 1995 by the IEEE 802.3u Fast Ethernet standard with a speed of 100 Mbps, and later the IEEE 802.3z Gigabit Ethernet standard with a speed of 1000 Mbps was adopted. At the moment, 10 Gigabit Ethernet IEEE 802.3ae, which has a speed of 10,000 Mbit / s, is already in full use. In addition, we already have developments aimed at achieving a speed of 100,000 Mbit / s 100 Gigabit Ethernet, but first things first.

A very important position underlying the Ethernet standard is its frame format. However, there are quite a few options for it. Here is some of them:

Variant I is the firstborn and already out of use.

Ethernet Version 2 or Ethernet frame II, also called DIX (abbreviation of the first letters of the developers of DEC, Intel, Xerox) is the most common and is used to this day. Often used directly by the Internet Protocol.

Novell is an internal modification of IEEE 802.3 without LLC (Logical Link Control).

IEEE 802.2 LLC frame.

IEEE 802.2 LLC / SNAP frame.

Optionally, an Ethernet frame can contain an IEEE 802.1Q tag to identify the VLAN to which it is addressed and an IEEE 802.1p tag to indicate priority.

Some Hewlett-Packard Ethernet cards used an IEEE 802.12 frame that conforms to the 100VG-AnyLAN standard.

For different frame types, there are also different formats and MTU values.

Functional elements of technologyGigabit Ethernet

Note that manufacturers of Ethernet cards and other devices mainly include support for several previous baud rate standards in their products. By default, using autosensing of speed and duplex, the card drivers themselves determine the optimal mode of operation for the connection between the two devices, but, usually, there is also a manual choice. So, buying a device with an Ethernet 10/100/1000 port, we get the opportunity to work with 10BASE-T, 100BASE-TX, and 1000BASE-T technologies.

Here is the chronology of modifications Ethernet by dividing them by transmission rates.

First solutions:

Xerox Ethernet is the original technology, the speed of 3 Mbps, existed in two versions, Version 1 and Version 2, the frame format of the latest version is still widely used.

10BROAD36 - not widespread. One of the first standards to allow long distance work. Used broadband modulation technology similar to that used in cable modems. A coaxial cable was used as a data transmission medium.

1BASE5 - also known as StarLAN, was the first twisted pair Ethernet technology. It worked at a speed of 1 Mbit / s, but did not find commercial use.

More common and optimized for its time modifications of 10 Mbit / s Ethernet:

10BASE5, IEEE 802.3 (also called “Thick Ethernet”) was the original development of a 10 Mbps technology. The IEEE uses a 50 ohm coaxial cable (RG-8) with a maximum segment length of 500 meters.

10BASE2, IEEE 802.3a (called "Thin Ethernet") - uses RG-58 cable, with a maximum segment length of 200 meters. To connect computers to each other and connect the cable to the network card, you need a T-connector, and the cable must have a BNC connector. Terminators are required at each end. For many years this standard has been the main standard for Ethernet technology.

StarLAN 10 - The first design to use twisted pair cable for data transmission at 10 Mbps. Later, it evolved into the 10BASE-T standard.

10BASE-T, IEEE 802.3i - 4 twisted pair cables (two twisted pairs) of Category 3 or Category 5 are used for data transmission. The maximum segment length is 100 meters.

FOIRL - (acronym for Fiber-optic inter-repeater link). Basic standard for Ethernet technology using optical cable for data transmission. The maximum data transmission distance without repeater is 1 km.

10BASE-F, IEEE 802.3j - The main term for the 10 Mbit / s family of Eethernet standards using fiber optic cables up to 2 kilometers away: 10BASE-FL, 10BASE-FB, and 10BASE-FP. Of the above, only 10BASE-FL is widely used.

10BASE-FL (Fiber Link) - An improved version of the FOIRL standard. The improvement concerned an increase in the segment length up to 2 km.

10BASE-FB (Fiber Backbone) - Now an unused standard, it was intended for combining repeaters into a backbone.

• 10BASE-FP (Fiber Passive) - Passive star topology that does not require repeaters - developed but never implemented.

The most common and inexpensive choice at the time of writing Fast Ethernet (100 Mbps) ( Fast Ethernet):

100BASE-T - The main term for one of the three standards of 100 Mbit / s Ethernet, using twisted pair as the data transmission medium. Segment length up to 100 meters. Includes 100BASE-TX, 100BASE-T4 and 100BASE-T2.

100BASE-TX, IEEE 802.3u - Development of 10BASE-T technology, a star topology is used, a twisted pair cable of category 5 is used, which actually uses 2 pairs of conductors, the maximum data transfer rate is 100 Mbps.

100BASE-T4 - 100 Mbps Ethernet over Category 3 cable. All 4 pairs are used. Now it is practically not used. Data transmission is in half duplex mode.

100BASE-T2 - Not used. 100 Mbps Ethernet over Category 3 cable. Only 2 pairs are used. Full duplex transmission mode is supported, when signals propagate in opposite directions on each pair. The transfer rate in one direction is 50 Mbit / s.

100BASE-FX - 100 Mbps Ethernet over fiber optic cable. The maximum segment length is 400 meters in half duplex mode (for guaranteed collision detection) or 2 kilometers in full duplex mode over multimode fiber.

100BASE-LX - 100 Mbps Ethernet over fiber optic cable. The maximum segment length is 15 kilometers in full duplex mode over a pair of single-mode optical fibers at a wavelength of 1310 nm.

100BASE-LX WDM - 100 Mbps Ethernet over fiber optic cable. The maximum segment length is 15 kilometers in full duplex mode over one single-mode optical fiber at a wavelength of 1310 nm and 1550 nm. Interfaces are of two types, differ in the transmitter wavelength and are marked with either numbers (wavelength) or one Latin letter A (1310) or B (1550). In a pair, only paired interfaces can work, on the one hand a transmitter at 1310 nm, and on the other at 1550 nm.

Gigabit Ethernet

1000BASE-T, IEEE 802.3ab - 1 Gbps Ethernet standard. A twisted pair of category 5e or category 6 is used. All 4 pairs are involved in data transmission. The data transfer rate is 250 Mbps over one pair.

1000BASE-TX, - 1 Gbps Ethernet standard using only Category 6 twisted pair. Transmitting and receiving pairs are physically separated by two pairs in each direction, which greatly simplifies the design of transceiver devices. The data transfer rate is 500 Mbps over one pair. Practically not used.

1000Base-X is a generic term for Gigabit Ethernet technology with pluggable GBIC or SFP transceivers.

1000BASE-SX, IEEE 802.3z - 1 Gbps Ethernet technology uses lasers with an allowable radiation length within the range of 770-860 nm, transmitter radiation power in the range from -10 to 0 dBm with an ON / OFF ratio (signal / no signal) not less than 9 dB. Receiver sensitivity 17 dBm, receiver saturation 0 dBm. Using multimode fiber, the signal transmission range without repeater is up to 550 meters.

1000BASE-LX, IEEE 802.3z - 1 Gbps Ethernet technology uses lasers with an allowable radiation length within the range of 1270-1355 nm, transmitter radiation power in the range from 13.5 to 3 dBm, with an ON / OFF ratio (there is a signal / no signal) not less than 9 dB. Receiver sensitivity 19 dBm, receiver saturation 3 dBm. When using multimode fiber, the signal transmission range without a repeater is up to 550 meters. Optimized for long distance using single mode fiber (up to 40 km).

1000BASE-CX - Gigabit Ethernet technology for short distances (up to 25 meters), uses a special copper cable (Shielded Twisted Pair (STP)) with a characteristic impedance of 150 ohms. Replaced by 1000BASE-T standard, and is not used now.

1000BASE-LH (Long Haul) - 1 Gbps Ethernet technology, uses a single-mode optical cable, the signal transmission range without a repeater is up to 100 kilometers.

Specifications of 1000Base-X standards

10 Gigabit Ethernet

Still quite expensive, but quite popular, the new 10 Gigabit Ethernet standard includes seven physical media standards for LAN, MAN and WAN. It is currently covered by the IEEE 802.3a amendment and should be included in the next revision of the IEEE 802.3 standard.

10GBASE-CX4 - 10 Gigabit Ethernet technology for short distances (up to 15 meters) using CX4 copper cable and InfiniBand connectors.

10GBASE-SR - 10 Gigabit Ethernet technology for short distances (up to 26 or 82 meters, depending on the cable type) using multimode fiber. It also supports distances up to 300 meters using new multimode fiber (2000 MHz / km).

10GBASE-LX4 - Uses wavelength division multiplexing to support distances from 240 to 300 meters over multimode fiber. Also supports distances up to 10 kilometers when using single mode fiber.

10GBASE-LR and 10GBASE-ER - these standards support distances up to 10 and 40 kilometers, respectively.

10GBASE-SW, 10GBASE-LW, and 10GBASE-EW - These standards use a physical interface that is speed and data format compatible with the OC-192 / STM-64 SONET / SDH interface. They are similar to the 10GBASE-SR, 10GBASE-LR and 10GBASE-ER standards respectively, as they use the same cable types and transmission distances.

10GBASE-T, IEEE 802.3an-2006 - adopted in June 2006 after 4 years of development. Uses shielded twisted pair cable. Distances - up to 100 meters.

And finally, what do we know about 100-Gigabit Ethernet(100-GE), which is still quite crude, but quite popular technology.

In April 2007, after the meeting of the IEEE 802.3 committee in Ottawa, the Higher Speed ​​Study Group (HSSG) took an opinion on the technical approaches to the formation of optical and copper 100-GE channels. At this time, it is finally formed working group 802.3ba on the development of the 100-GE specification.

As in previous developments, the 100-GE standard will take into account not only economic and technical capabilities its implementation, but also their backward compatibility with existing systems. At this time, the need for such speeds is indisputably proven by leading companies. Constantly growing volumes of personalized content, including when delivering videos from portals such as YouTube and other resources using IPTV and HDTV technologies. We should also mention video on demand. All this determines the need for 100 Gigabit Ethernet operators and service providers.

But against the background of a large selection of old and promising new technological approaches within the Ethernet group, we want to dwell in more detail on the technology, which today is only acquiring full-fledged mass use due to the decrease in the cost of its components. Gigabit Ethernet can fully support applications such as video streaming, video conferencing, and complex image transmission with increased bandwidth requirements. The benefits of higher transmission speeds in corporate and home networks are becoming more and more undeniable as prices for this class of equipment fall.

Now the IEEE standard has received the maximum popularity. Adopted in June 1998, it was approved as IEEE 802.3z. But at first, only an optical cable was used as a transmission medium. With the approval of the addition of the 802.3ab standard during the following year, Category 5 unshielded twisted pair became the transmission medium.

Gigabit Ethernet is a direct descendant of Ethernet and Fast Ethernet, which have proven themselves well over nearly twenty years of history, maintaining their reliability and future-proofing. Along with the foreseen backward compatibility with previous solutions (the cable structure remains unchanged), it provides a theoretical throughput of 1000 Mbps, which is approximately equal to 120 Mb per second. It should be noted that such capabilities are practically equal to the speed of a 32-bit PCI bus 33 MHz. That is why gigabit adapters are available both for 32-bit PCI (33 and 66 MHz) and for 64-bit bus. Along with this increase in speed, Gigabit Ethernet inherited all previous Ethernet features such as frame format, CSMA / CD (transmission sensitive collision detection multiple access) technology, full duplex, etc. Although high speeds have made their own innovations, it is precisely in the inheritance of old standards that the huge advantage and popularity of Gigabit Ethernet lies. Of course, other solutions are now proposed, such as ATM and Fiber Channel, but here the main advantage for the end user is immediately lost. The transition to another technology leads to a massive redesign and re-equipment of enterprise networks, while Gigabit Ethernet will allow for a smooth increase in speed and not change the cabling. This approach allowed Ethernet technology to take a dominant place in the field of network technologies and conquer more than 80 percent of the world information transmission market.

The structure of building an Ethernet network with smooth transitions to higher data rates.

Initially, all Ethernet standards were developed using only an optical cable as a transmission medium - so Gigabit Ethernet received a 1000BASE-X interface. It is based on the Fiber Channel physical layer standard (a technology for interworking workstations, storage devices, and edge nodes). Since this technology had already been approved earlier, this borrowing greatly reduced the development time for the Gigabit Ethernet standard. 1000BASE-X

We, as well as a common man in the street, were more interested in 1000Base-CX in view of its operation on shielded twisted pair (STP "twinax") for short distances and 1000BASE-T for unshielded twisted pair of category 5. The main difference between 1000BASE-T and Fast Ethernet 100BASE- TX became that all four pairs were used (in 100BASE-TX only two were used). At the same time, each pair can transmit data at a speed of 250 Mbps. The standard provides full duplex transmission, with the flow on each pair being provided in two directions simultaneously. Due to strong interference during such transmission, it was technically much more difficult to implement gigabit transmission over twisted pair than in 100BASE-TX, which required the development of a special scrambled noise-immune transmission, as well as an intelligent signal recognition and restoration unit at reception. As a coding method in the 1000BASE-T standard, 5-level pulse-amplitude coding PAM-5 was used.

The criteria for choosing a cable have also become more stringent. To reduce pickup, unidirectional transmission, return loss, delay and phase shift, Category 5e for unshielded twisted pair has been adopted.

Crimping cable for 1000BASE-T is performed according to one of the following schemes:

Straight-through cable.

Crossover cable.

Crimping diagrams of a cable for 1000BASE-T

The innovations also affected the level of the MAC-standard 1000BASE-T. In Ethernet networks, the maximum distance between stations (collision domain) is determined based on the minimum frame size (in the Ethernet IEEE 802.3 standard it was 64 bytes). The maximum segment length must be such that the transmitting station can detect a collision before the end of the frame transmission (the signal must have time to pass to the other end of the segment and return back). Accordingly, with an increase in the transmission rate, it is necessary either to increase the frame size, thereby increasing the minimum time for frame transmission, or to decrease the diameter of the collision domain.

When switching to Fast Ethernet, they used the second option and reduced the segment diameter. In Gigabit Ethernet, this was not acceptable. Indeed, in this case, the standard that inherited such components of Fast Ethernet as the minimum frame size, CSMA / CD and the time slot for collision detection will be able to work in collision domains with a diameter of no more than 20 meters. Therefore, it was proposed to increase the time for transmitting the minimum frame. Considering that for compatibility with previous Ethernet, the minimum frame size was left the same - 64 bytes, and an additional carrier extension field was added to the frame, which complements the frame to 512 bytes, but the field is not added in the case when the frame size is greater than 512 byte. Thus, the resulting minimum frame size turned out to be 512 bytes, the time for collision detection increased, and the segment diameter increased to the same 200 meters (in the case of 1000BASE-T). Symbols in the carrier extension field have no semantic meaning, the checksum is not calculated for them. When a frame is received, this field is discarded even at the MAC layer, so the higher layers continue to work with minimum frames of 64 bytes long.

But here too there were pitfalls. While the media expansion allowed for compatibility with previous standards, it wasted bandwidth. Loss can be as high as 448 bytes (512-64) per frame for short frames. Therefore, the 1000BASE-T standard was modernized - the concept of Packet Bursting was introduced. It allows you to use the expansion field much more effectively. And it works as follows: if the adapter or switch has several small frames that need to be sent, then the first of them is sent in the standard way, with the addition of an extension field up to 512 bytes. And all subsequent ones are sent in their original form (without the extension field), with a minimum interval of 96 bits between them. And, most importantly, this interframe gap is filled with media spread symbols. This happens until the total size of frames sent reaches the limit of 1518 bytes. Thus, the medium does not become silent throughout the transmission of small frames, so a collision can occur only at the first stage, when transmitting the first correct small frame with a carrier expansion field (512 bytes in size). This mechanism can significantly improve network performance, especially under heavy loads, by reducing the likelihood of collisions.

But this was not enough. Initially, Gigabit Ethernet only supported standard Ethernet frame sizes, from a minimum of 64 (padded to 512) to a maximum of 1518 bytes. Of these, 18 bytes are occupied by the standard service header, and for data there are from 46 to 1500 bytes, respectively. But even a 1500 byte data packet is too small in the case of a gigabit network. Especially for servers transferring large amounts of data. Let's count a little. To transfer a 1 gigabyte file over an unloaded Fast Ethernet network, the server processes 8200 packets / sec and takes at least 11 seconds to do this. In this case, the 200 MIPS computer will take about 10 percent of the time to handle interrupts alone. After all, the central processor must process (calculate the checksum, transfer data to memory) each packet that arrives.

Ethernet transmission characteristics.

In gigabit networks, the situation is even worse - the load on the processor increases by about an order of magnitude due to the reduction in the time interval between frames and, accordingly, interrupt requests to the processor. Table 1 shows that even under the best conditions (using frames of the maximum size), the frames are spaced from each other by a time interval not exceeding 12 μs. In the case of using smaller frames, this time interval only decreases. Therefore, in gigabit networks, oddly enough, it was the stage of processing frames by the processor that became the bottleneck. Therefore, at the dawn of the formation of Gigabit Ethernet, the actual transfer rates were far from the theoretical maximum - the processors simply could not cope with the load.

The obvious way out of this situation is the following:

increasing the time interval between frames;

shifting part of the load of processing frames from the central processor to the network adapter itself.

Both methods are currently implemented. In 1999, it was proposed to increase the packet size. Such packets were called Jumbo Frames, and their size could be from 1518 to 9018 bytes (currently, equipment from some manufacturers also supports large giga frame sizes). Jumbo Frames allowed to reduce the load on the central processor up to 6 times (proportional to its size) and, thus, significantly increase performance. For example, the maximum Jumbo Frame packet of 9018 bytes, in addition to the 18-byte header, contains 9000 bytes for data, which corresponds to six standard maximum Ethernet frames. The gain in performance is achieved not due to getting rid of several service headers (the traffic from their transmission does not exceed a few percent of the total bandwidth), but due to the reduction in the processing time of such a frame. More precisely, the time to process a frame remains the same, but instead of several small frames, each of which would require N processor cycles and one interrupt, we process only one, larger frame.

The fairly rapidly developing world of information processing speed provides ever faster and inexpensive solutions on the use of special hardware, to remove part of the traffic processing load from the central processor. Buffering technology is also used to interrupt the processor to process multiple frames at once. At this time, Gigabit Ethernet technology is becoming more and more available for use at home, which will directly interest the common user. More fast access to home resources will provide high-quality video viewing high resolution, will take less time to redistribute information and, finally, will allow live encoding of video streams to network drives.

In the preparation of the article, resource materials were used http://www.ixbt.com/ andhttp://www.wikipedia.org/.

Article read 15510 times

Now there is a lot of talk about the time to massively switch to gigabit speeds when connecting end users of local networks, and again the question is raised about the justification and progressiveness of solutions "fiber to the workplace", "fiber to home", etc. In this regard, this article, describing the standards not only for copper, but also mainly for fiber-optic GigE interfaces, will be quite appropriate and timely.

Gigabit Ethernet architecture

Figure 1 shows the structure of the Gigabit Ethernet layers. As in the Fast Ethernet standard, there is no universal signal coding scheme in Gigabit Ethernet that would be ideal for all physical interfaces - on the one hand, 1000Base-LX / SX / CX standards use 8B / 10B coding, and on the other On the other hand, for the 1000Base-T standard, a special extended line code TX / T2 is used. The encoding function is performed by the PCS encoding sublayer located below the independent GMII interface.

Rice. 1. Layer structure of Gigabit Ethernet standard, GII interface and Gigabit Ethernet transceiver

GMII interface. The Gigabit Media Independent Interface (GMII) provides interoperability between the MAC layer and the physical layer. The GMII interface is an extension of the MII interface and can support speeds of 10, 100 and 1000 Mbps. It has a separate 8-bit receiver and transmitter, and can support both half-duplex and full-duplex modes. In addition, the GMII interface carries one clock signal, and two line state signals - the first (in the ON state) indicates the presence of a carrier, and the second (in the ON state) indicates the absence of collisions - and several other signal channels. and food. The transceiver module, covering the physical layer and providing one of the physical media-dependent interfaces, can connect, for example, to a Gigabit Ethernet switch via a GMII interface.

Physical coding sublayer PCS. When connecting 1000Base-X interfaces, the PCS sublayer uses 8B10B block redundant coding, borrowed from the ANSI X3T11 Fiber Channel standard. Similar to the considered FDDI standard, only on the basis of a more complex code table, every 8 input bits intended for transmission to a remote node are converted into 10-bit symbols (code groups). In addition, there are special 10-bit control characters in the output serial stream. An example of control characters are characters used to expand media (padding a Gigabit Ethernet frame to its minimum size of 512 bytes). When connecting the 1000Base-T interface, the PCS sublayer implements a special noise-immune coding to ensure transmission over UTP Cat.5 twisted pair at a distance of up to 100 meters - the TX / T2 line code developed by Level One Communications.

Two line status signals - carrier presence signal and no collision signal - are generated by this sublevel.

Sublevels PMA and PMD. The physical layer of Gigabit Ethernet uses multiple interfaces, including traditional Category 5 twisted pair, multimode and singlemode fiber. The PMA sublayer converts the parallel character stream from the PCS to a serial stream, and also converts (parallelizes) the incoming serial stream from the PMD. The PMD sublayer defines the optical / electrical characteristics of physical signals for different environments. In total, 4 different types of physical media interfaces are defined, which are reflected in the specification of the 802.3z (1000Base-X) and 802.3ab (1000Base-T) standards, (Fig. 2).

Rice. 2. Physical interfaces of the Gigabit Ethernet standard

1000Base-X interface

The 1000Base-X interface is based on the Fiber Channel physical layer standard. Fiber Channel is a technology that connects workstations, supercomputers, storage devices, and edge nodes. Fiber Channel has a 4-tier architecture. The two lower layers FC-0 (interfaces and media) and FC-1 (encoding / decoding) have been moved to Gigabit Ethernet. Because Fiber Channel is an approved technology, this move has greatly reduced the development time for the original Gigabit Ethernet standard.

The 8B / 10B block code is similar to the 4B / 5B code used in the FDDI standard. However, the 4B / 5B code was rejected in Fiber Channel because the code does not balance the direct current... Imbalance can potentially lead to data-dependent heating laser diodes because the transmitter can transmit more "1" bits (there is radiation) than "0" (there is no radiation), which can cause additional errors at high transmission rates.

1000Base-X is subdivided into three physical interfaces, the main characteristics of which are as follows:

The 1000Base-SX interface defines lasers with an allowable radiation length within the range of 770-860 nm, the transmitter radiation power in the range from -10 to 0 dBm, with an ON / OFF ratio (signal / no signal) not less than 9 dB. Receiver sensitivity -17 dBm, receiver saturation 0 dBm;

The 1000Base-LX interface detects lasers with an allowable radiation length within the range of 1270-1355 nm, the transmitter radiation power in the range from -13.5 to -3 dBm, with an ON / OFF ratio (there is a signal / no signal) of at least 9 dB. Receiver sensitivity -19 dBm, receiver saturation -3 dBm;

1000Base-CX shielded twisted pair (STP "twinax") over short distances.

For reference, Table 1 shows the main characteristics of the optical transceiver modules manufactured by the company. Hewlett Packard for standard interfaces 1000Base-SX (model HFBR-5305, = 850 nm) and 1000Base-LX (model HFCT-5305, = 1300 nm).

Table 1. Specifications of optical Gigabit Ethernet transceivers

The supported distances for 1000Base-X standards are shown in Table 2.

Table 2. Technical characteristics of optical Gigabit Ethernet transceivers

With 8B / 10B encoding, the optical line bit rate is 1250 bps. This means that the bandwidth of the allowed cable length must be greater than 625 MHz. From table. 2 shows that this criterion is met for lines 2-6. Due to the high transmission speed of Gigabit Ethernet, care should be taken when constructing long segments. Single-mode fiber is definitely preferred. In this case, the characteristics of optical transceivers can be significantly higher. For example, NBase manufactures switches with Gigabit Ethernet ports that provide distances of up to 40 km over single-mode fiber without retransmission (narrow-spectrum DFB lasers operating at 1550 nm are used).

features of using multimode fiber

There are a huge number of corporate networks in the world based on multimode fiber-optic cable, with 62.5 / 125 and 50/125 fibers. Therefore, it is natural that even at the stage of the formation of the Gigabit Ethernet standard, the problem arose of adapting this technology for use in existing multimode cable systems. In the course of research on the development of the 1000Base-SX and 1000Base-LX specifications, one very interesting anomaly was revealed associated with the use of laser transmitters in conjunction with multimode fiber.

The multimode fiber was designed to be used in conjunction with light emitting diodes (emission spectrum 30-50 ns). Incoherent radiation from such LEDs enters the fiber over the entire area of ​​the light-carrying core. As a result, a huge number of mode groups are excited in the fiber. The propagating signal lends itself well to description in the language of intermode dispersion. The efficiency of using such LEDs as transmitters in the Gigabit Ethernet standard is low due to the very high modulation frequency - the bit rate in the optical line is 1250 Mbaud, and the duration of one pulse is 0.8 ns. The maximum speed, when LEDs are still used for signal transmission over multimode fiber, is 622.08 Mbps (STM-4, taking into account the redundancy of the 8B / 10B code, the bit rate in the optical line is 777.6 Mbaud). Therefore, Gigabit Ethernet became the first standard to regulate the use of optical laser transmitters in conjunction with multimode fiber. The area of ​​input of radiation into the fiber from the laser is much smaller than the size of the core of a multimode fiber. This fact in itself does not yet lead to a problem. At the same time, in the technological process of production of standard commercial multimode fibers, some defects (deviations within the allowable range) that are not critical for traditional use of the fiber are allowed, which are most concentrated near the axis of the fiber core. Although such a multimode fiber fully meets the requirements of the standard, coherent laser light introduced into the center of such a fiber, passing through regions of inhomogeneity of the refractive index, is able to split into a small number of modes, which then propagate along the fiber by different optical paths and at different speeds. This phenomenon is known as differential mode delay DMD. As a result, a phase shift appears between the modes, leading to unwanted interference on the receiving side and to a significant increase in the number of errors (Fig. 3a). Note that the effect manifests itself only under the simultaneous combination of a number of circumstances: a less successful fiber, a less successful laser transmitter (of course, meeting the standard) and less successful radiation input into the fiber. On the physical side, the DMD effect is associated with the fact that the energy from a coherent source is distributed within a small number of modes, while an incoherent source uniformly excites a huge number of modes. Studies show that the effect is more pronounced when using long wavelength lasers (transparency window 1300 nm).

Fig. 3. Propagation of coherent radiation in a multimode fiber: a) Manifestation of the effect of differential mode delay (DMD) at axial coupling of radiation; b) Off-axis coupling of coherent radiation into a multimode fiber.

This anomaly in the worst case can lead to a decrease in the maximum segment length based on the multimode FOC. Since the standard is supposed to provide a 100% performance guarantee, the maximum segment length should be regulated taking into account the possible manifestation of the DMD effect.

1000Base-LX interface... In order to maintain a greater distance and avoid the unpredictability of the behavior of the Gigabit Ethernet link due to anomaly, it is proposed to inject radiation into the off-center part of the multimode fiber core. The radiation due to the aperture divergence manages to be evenly distributed over the entire fiber core, greatly weakening the manifestation of the effect, although the maximum segment length remains limited after that (Table 2). MCP (mode conditioning patch-cords) single-mode transitional optical cords are specially designed, in which one of the connectors (namely, the one that is planned to be mated with multimode fiber) has a slight offset from the axis of the fiber core. An optical cord with one connector being Duplex SC with an offset core and the other with a regular Duplex SC may be referred to as MCP Duplex SC - Duplex SC. Of course, such a cable is not suitable for use in traditional networks, for example, in Fast Ethernet, due to the large insertion loss at the interface with the MCP Duplex SC. The transient MCP can be a combined single-mode and multi-mode fiber and contain an inter-fiber bias element internally. Then, with a single-mode end, it is connected to a laser transmitter. As for the receiver, a standard multimode patch cord can be connected to it. The use of transitional MCP cords allows radiation to be introduced into a multimode fiber through a region offset by 10-15 microns from the axis (Fig. 3b). Thus, it remains possible to use 1000Base-LX interface ports with single-mode FOCs, since radiation will be injected there strictly in the center of the fiber core.

1000Base-SX interface... Since the 1000Base-SX interface is standardized only for use with multimode fiber, the offset of the radiation input area from the central axis of the fiber can be implemented inside the device itself, thereby eliminating the need to use an optical matching cord.

1000Base-T interface

1000Base-T is a standard Gigabit Ethernet interface for transmission over unshielded twisted pair Category 5 and higher over distances up to 100 meters. All four pairs of copper cable are used for transmission, the transmission rate for one pair is 250 Mbit / s. It is assumed that the standard will provide full-duplex transmission, and data on each pair will be transmitted simultaneously in two directions at once - dual duplex. 1000Base-T. Technically, it turned out to be quite difficult to implement 1 Gbps duplex transmission over UTP cat.5 twisted pair, much more difficult than in the 100Base-TX standard. The influence of near and far crosstalk from three adjacent twisted pairs on a given pair in a four-pair cable requires the development of a special scrambled noise-immune transmission, and an intelligent signal recognition and restoration unit at reception. Several coding methods were initially considered as candidates for approval in the 1000Base-T standard, including: 5-level pulse-amplitude coding PAM-5; quadrature amplitude modulation QAM-25, etc. Below are brief ideas of PAM-5, finally approved as a standard.

Why 5-level coding. Common 4-level coding processes incoming bits in pairs. In total, there are 4 different combinations - 00, 01, 10, 11. The transmitter can set its own voltage level of the transmitted signal for each pair of bits, which halves the modulation frequency of the four-level signal, 125 MHz instead of 250 MHz, (Fig. 4), and therefore radiation frequency. A fifth level has been added to create code redundancy. As a result, reception error correction becomes possible. This gives an additional 6 dB signal-to-noise ratio.

Fig. 4. PAM-4 4-level coding scheme

MAC level

The Gigabit Ethernet MAC layer uses the same CSMA / CD transfer protocol as its Ethernet and Fast Ethernet ancestors. The main restrictions on the maximum length of a segment (or collision domain) are determined by this protocol.

The Ethernet IEEE 802.3 standard has a minimum frame size of 64 bytes. It is the value of the minimum frame size that determines the maximum allowable distance between stations (diameter of the collision domain). The time that the station transmits such a frame - the channel time - is 512 BT or 51.2 μs. The maximum length of the Ethernet network is determined from the collision resolution condition, namely, the time it takes for the signal to reach the remote node and return RDT back should not exceed 512 BT (excluding the preamble).

When switching from Ethernet to Fast Ethernet, the transmission speed increases, and the translation time of a 64-byte frame is correspondingly reduced - it is equal to 512 BT or 5.12 μs (in Fast Ethernet 1 BT = 0.01 μs). In order to be able to detect all collisions before the end of the frame transmission, as before, one of the conditions must be met:

Fast Ethernet kept the same minimum frame size as Ethernet. This retained compatibility, but resulted in a significant reduction in the collision domain diameter.

Again, by virtue of its continuity, the Gigabit Ethernet standard must support the same minimum and maximum frame sizes that are accepted in Ethernet and Fast Ethernet. But as the transmission speed increases, the transmission time of a packet of the same length decreases accordingly. While maintaining the same minimum frame length, this would lead to a decrease in the network diameter, which would not exceed 20 meters, which could be of little use. Therefore, when developing the Gigabit Ethernet standard, it was decided to increase the channel time. In Gigabit Ethernet, it is 4096 BT and is 8 times faster than Ethernet and Fast Ethernet. However, to maintain compatibility with the Ethernet and Fast Ethernet standards, the minimum frame size was not increased, but an additional field was added to the frame, called "media extension".

carrier extension

Symbols in the additional field usually do not carry service information, but they fill the channel and increase the "collision window". As a result, the collision will be recorded by all stations with a larger collision domain diameter.

If the station wishes to transmit a short (less than 512 bytes) frame, this field is added to the transmission - a carrier extension that complements the frame to 512 bytes. The checksum field is calculated only for the original frame and does not apply to the extension field. When a frame is received, the extension field is discarded. Therefore, the LLC layer does not even know about the presence of the extension field. If the frame size is equal to or greater than 512 bytes, then there is no media extension field. Figure 5 shows the Gigabit Ethernet frame format when using a media extension.

Fig. 5. Gigabit Ethernet frame with media extension field.

packet bursting

Media expansion is the most natural solution to maintain Fast Ethernet compatibility and the same collision domain diameter. But it wasted bandwidth. Up to 448 bytes (512-64) can be wasted when transmitting a short frame. During the development stage of the Gigabit Ethernet standard, NBase Communications made a proposal to upgrade the standard. This upgrade, called batch congestion, allows for more efficient use of the extension field. If the station / switch has several small frames to send, then the first frame is padded with a carrier expansion field to 512 bytes and sent. The rest of the frames are sent after a minimum interframe interval of 96 bits, with one important exception - the interframe gap is filled with extension symbols (Fig. 6a). Thus, the medium does not become silent between the sending of short original frames, and no other device on the network can interfere with the transmission. Such frame alignment can occur until the total number of transmitted bytes exceeds 1518. Packet congestion reduces the likelihood of collisions, since an overloaded frame can collide only at the stage of transmission of its first original frame, including media expansion, which certainly increases network performance. especially at heavy loads (Fig. 6-b).

Fig. 6. Packet congestion: a) frame transmission; b) bandwidth behavior.

Based on the materials of the company "Telecom Transport"