Gigabit Ethernet

Gigabit Ethernet was originally defined in the IEEE 802.3z standard and is now merged into the 2002 edition of IEEE 802.3. It operates essentially like 100-Mbps Ethernet, except that it is 10 times faster. It uses CSMA/CD with support for one repeater per collision domain, and handles both half- and full-duplex operations. It uses a standard 802.3 frame format and frame size.

To avoid the need to reduce the size of a half-duplex Gigabit Ethernet network to 1/10th the size of a 100-Mbps Ethernet network, Gigabit Ethernet includes some minor changes to the MAC layer. It's helpful to compare the changes the standards developers made when they introduced 100-Mbps Ethernet to the changes they made when they introduced Gigabit Ethernet.

With half-duplex 10- and 100-Mbps Ethernet, the minimum frame size is equal to the maximum round-trip propagation delay of the network (less an allowance for jamming, synchronization, and so on). To ensure that a sender is still sending if a collision occurs on the opposite side of the largest possible network, a sender must send for the time it takes 512 bits (64 bytes) to travel to the opposite side of the network and back. This ensures that the sending Ethernet NIC will recognize the collision, back off, and retransmit.

To ensure that collision recognition works correctly on 100-Mbps Ethernet, where senders can send 10 times as fast as they can with 10-Mbps Ethernet, the standards developers reduced the maximum network size to about 1/10th the maximum size for 10-Mbps Ethernet. Using such an approach with half-duplex Gigabit Ethernet would have resulted in impracticably small networks, however. Instead, the minimum frame is maintained at 512 bits (64 bytes), but the sender is required to send for the time it takes 4096 bits (512 bytes) to travel to the opposite side of the network and back.

Frames that are shorter than 4096 bits are artificially extended by appending a carrier-extension field so that the frames are exactly 4096 bits in length. The contents of the carrier-extension field are nondata symbols. By increasing the minimum time a station sends, the standards can support a larger maximum network topology.

Carrier extension could cause performance degradation for some applications, although the developers included a workaround in the standards to avoid problems. Consider an application that is trying to output many small frames at a quick rate, for example a voice over IP (VoIP) application. If the interface adds bits to reach 4096 bits, will the application be able to send at the proper rate? When the application is finished and releases control of the medium, will some other application jump in, causing the first application to defer? The Gigabit Ethernet designers addressed these problems by allowing a sender to send multiple frames.

A Gigabit Ethernet station may choose to burst frames if there is a frame in its transmit queue when it has finished sending its first frame (plus extension, if one was necessary). The station may send again without contending for use of the channel. During the interframe gap, the station sends nondata symbols. Only the first frame requires a carrier extension. The station may start the transmission of frames for up to one burstLength. IEEE defines the burstLength parameter as 8192 bytes.

Gigabit Ethernet is most appropriate for building and campus backbone networks. It can act as a trunk network, theoretically aggregating traffic from up to ten 100-Mbps Ethernet segments. Despite the carrier-extension and bursting features of half-duplex Ethernet, Gigabit Ethernet usually uses full-duplex mode and connects a switch to another switch or a switch to a router. Gigabit Ethernet is also used on high-performance servers that require a large amount of bandwidth.

The 802.3 standard for Gigabit Ethernet specifies multimode and single-mode fiberoptic cabling, UTP cabling, and shielded twinax copper cabling. Table 10-4 shows the variations of Gigabit Ethernet.

Table 10-4. Gigabit Ethernet Specifications

1000BASE-SX

1000BASE-LX

1000BASE-CX

1000BASE-T

Type of

850-nanometer

1300 nanometer

Twinax

UTP

cabling

wavelength multimode fiber

wavelength multimode and single-mode fiber

Distance limitations (in meters)

220-550, depending on the cable

550 for multimode and 5000 for single mode

25

100 between a hub and station; a total network diameter of 200 meters

1000BASE-SX, also known as the short-wavelength specification (hence the S in the name), is appropriate for multimode horizontal cabling and backbone networks.

1000BASE-LX uses a longer wavelength (hence the L in the name), and supports both multimode and single-mode cabling. 1000BASE-LX is appropriate for building and campus backbone networks.

1000BASE-CX is appropriate for a telecommunications closet or computer room where the distance between devices is 25 meters or less. 1000BASE-CX runs over 150-ohm balanced, shielded, twinax cable.

1000BASE-T is intended for horizontal and work-area Category 5 or better UTP cabling. 1000BASE-T supports transmission over four pairs of UTP cable and covers a cabling distance of up to 100 meters, or a network diameter of 200 meters. Only one repeater is allowed.

Was this article helpful?

+1 0
Project Management Made Easy

Project Management Made Easy

What you need to know about… Project Management Made Easy! Project management consists of more than just a large building project and can encompass small projects as well. No matter what the size of your project, you need to have some sort of project management. How you manage your project has everything to do with its outcome.

Get My Free Ebook


Post a comment