Ethernet

Ethernet Cable Types

Cat5, Cat6, Cat6a specifications, fiber optic cables, and physical layer requirements.

Basics: the transmission medium

A computer network moves information between points through a transmission medium. This physical channel carries electrical, optical, or radio signals.

The choice of medium in a network design depends on several criteria:

  • Distance: How far does the signal need to travel without significant degradation?
  • Environment: In what physical conditions will the medium be installed? Is it a quiet office, a factory floor with heavy machinery, or outdoors?
  • Bandwidth and Speed: How much data needs to be transmitted per second?
  • Cost: What is the budget for the material and its installation?

Wired Ethernet mainly uses copper and fiber-optic cabling. Each standard defines requirements for a particular medium and reach. For twisted pair, a 100 m channel limit is common in 100BASE-TX and 1000BASE-T; the exact result depends on the channel category, PHY parameters, and installation conditions.

Copper Cabling in Ethernet Networks

Copper cabling has been used in local Ethernet networks for decades. It carries data as electrical signals encoded by the physical layer. Ethernet copper links use coaxial cable in historical and specialized systems, while twisted pair is the usual LAN medium.

1. Coaxial Cable

Coaxial cable, or "coax," consists of a central copper conductor, a dielectric, a metallic shield made of braid or foil, and an outer jacket. The shield limits electromagnetic coupling and protects the signal from some external interference.

  • Historical role:Coax was an Ethernet medium in the 10BASE-5 ("Thicknet") and 10BASE-2 ("Thinnet") standards.
  • Current use: Twisted pair has replaced coax in local Ethernet LANs, but coax remains common in cable television (CATV), satellite installations, and cable modems. The required impedance depends on the system.
  • Impedance: Coaxial cables are defined by their characteristic . Video installations often use 75 Ω, while many data and RF systems use 50 Ω.

2. Twisted-Pair Cable

Twisted pair is the most common cabling in Ethernet LANs. A typical installation cable contains four pairs of insulated copper wires, but 10BASE-T and 100BASE-TX use two pairs, while 1000BASE-T uses all four. In 1000BASE-T, every pair carries signals in both directions, so the PHY must cancel echo and crosstalk between pairs.

Twisting wires and limiting interference

Twisting improves line symmetry and limits two types of interference:

  • Noise from external sources (EMI): An electromagnetic field, for example from a motor, can induce a voltage in a pair. Twisting balances the exposure of both wires, while a differential receiver rejects much of the common-mode interference.
  • Crosstalk from adjacent pairs: A signal in one pair can couple into another pair. Different twist rates, cable geometry, and pair balance limit this coupling; cabling standards define the applicable limits.

Twisted-pair types: shielding and nomenclature

Twisted-pair cables are also described by their shielding. In the X/YTP notation, the first letter identifies the overall shield and the second identifies shielding around individual pairs; U, F, and S mean no shield, foil, and braid.

  • U/UTP (Unshielded Twisted Pair):Commonly called UTP. It has no overall shield and no individual pair shields, so interference rejection relies on the pair construction and twisting. It is widely used in office and home environments.
  • F/UTP (Foiled with Unshielded Twisted Pair):Often called FTP. It has an overall foil shield ('F') around four pairs without individual shields ('UTP'). The shield limits external EMI when it is bonded correctly.
  • S/FTP (Shielded with Foiled Twisted Pair):Each pair has its own foil shield ('FTP'), which limits crosstalk, and the cable has an additional overall braided shield ('S'). This construction can be selected for high-speed links, including 10 Gigabit Ethernet, and for environments with substantial interference.
  • SF/UTP (Shielded and Foiled with Unshielded Twisted Pair):This cable has two overall shields, a braid ('S') and a foil ('F'), while the internal pairs have no individual shields ('UTP').

A shielded cable must be bonded to the installation's equipotential system according to the design. Incorrect bonding can reduce shielding effectiveness and increase susceptibility to interference.

Twisted-pair performance: categories

Besides shielding, twisted-pair cables are classified into categories (TIA) or classes (ISO/IEC) based on electrical performance, including . A higher category usually provides more performance margin, but the supported data rate also depends on the Ethernet PHY, channel length, and installation conditions. The rating of the horizontal cable does not automatically describe the whole channel: connectors, patch cords, routing, and their compliance with the class also matter.

CategoryBandwidthCommon Application(s)
Category 316 MHzLegacy telephone systems, 10Base-T Ethernet (10 Mbps). Now obsolete for data networks.
Category 5e100 MHzCommon in installations. 100BASE-TX requires at least Category 5, while 1000BASE-T uses four pairs of Category 5 or better; Cat5e is a common choice for 1 Gbps channels.
Category 6250 MHzProvides more margin and less crosstalk than Cat 5e. Supports 1 Gbps and usually 10GBASE-T up to 55 meters, depending on channel parameters.
Category 6A500 MHzThe "A" stands for Augmented. Intended to support 10GBASE-T (10 Gbps) over a channel up to 100 meters long.
Category 7/7A600 / 1000 MHzCabling aligned with ISO Classes F/FA, usually shielded (for example S/FTP), with 600/1000 MHz parameters. It supports 10 Gbps; full class compliance may require an interface specified by ISO/IEC 11801 because 8P8C does not provide every configuration's full parameters.

Fiber-optic cabling

Fiber-optic cable transmits information as light pulses traveling through a glass core. The light remains in the core by a principle called , reflecting from the boundary with cladding that has a lower refractive index.

Fiber-optic advantages

  • High capacity: One fiber can carry multiple channels, and WDM systems reach aggregate rates in the terabits-per-second range.
  • Long Distances: Light signals suffer much lower than electrical signals, so with an appropriate power budget they can travel many kilometers without active regeneration.
  • EMI immunity: A glass fiber does not conduct electricity and does not pick up electromagnetic interference or crosstalk. A complete cable may still contain metallic elements, so lightning protection must be assessed for the whole construction.
  • Security: A fiber does not radiate a radio signal, so passive interception is harder than with copper. It does not remove the risk of tapping; detection requires monitoring link parameters.

Types of Fiber Optic Cable

Multimode Fiber (MMF)

MMF has a larger core in which multiple light modes propagate. Their different arrival times cause modal dispersion and limit reach. It is used over short links in LANs and data centers, usually hundreds of meters; selected systems reach approximately 2 km. The value depends on the fiber type, wavelength, transceiver, and required data rate; some 850 nm, 1 Gbps systems specify reaches of about 550 m.

Single-mode Fiber (SMF)

SMF has a small core in which one mode propagates. This removes modal dispersion, while chromatic and polarization-mode dispersion remain. With a suitable power budget and regeneration equipment, single-mode fiber supports long links and is widely used in long-haul, metropolitan, and carrier networks.

Fiber-optic transmission windows

Attenuation in fiber glass depends on wavelength. Ranges with favorable transmission parameters are known as "transmission windows". Selecting a window also requires the fiber type, light source, receiver sensitivity, and the link's complete power budget to be considered.

  • First window (850 nm): Used primarily with multimode fiber for short-range applications such as LANs.
  • Second window (1310 nm): Offers lower attenuation than the region around 850 nm. ITU-T G.652 single-mode fiber has its zero-dispersion wavelength in this region, which is used by high-speed systems.
  • Third window (1550 nm): In the usual telecommunications range it provides the lowest attenuation and is used for long-haul and undersea systems. WDM systems operate in its C band, approximately 1530-1565 nm.
  • L band and the region around 1625 nm (1565-1625 nm):The L band extends the range used by WDM and allows additional channels on one fiber. The term "fourth window" is used for the region around 1625 nm in some classifications.

Related articles