S 5.3 Selection of cable types suited in terms of communication technology

Initiation responsibility: Head of IT Section

Implementation responsibility: Network planner; Head of Site/Bldg Technical Service

From the communications technology standpoint, the choice of cable is determined by, among other things, the required data transfer rate (bandwidth) and the distances to be covered without repeaters. The requirements arising from the structural conditions at the installation location also have to be taken into account when making the selection. In the following, the respective advantages and disadvantages are described from an IT security point of view.

At present there can be considered to be two types of transmission media for cable (as opposed to wireless) communications: copper cable or an optical medium (optical fibre (FO)). Both of these types of media can be further divided into various subcategories. The most important of these for the medium of copper are the coaxial cable (one centre conductor with overall screening) and the multi-core copper cable with the wires twisted in pairs (twisted-pair cable, TP cable). These are explained in more detail below.

Twisted-pair cable

Twisted-pair cables are available in many different forms. These differ on the one hand according to the nature of their shielding and on the other according to their potential bandwidth. The current shielding classes are as follows:

• unshielded (UTP)

• unshielded but with overall screening ( screened-unshielded, S/UTP),

• shielded, where the individual pairs of wires are shielded ( shielded, STP)

• shielded TP cable with additional overall screening ( screened-shielded, S/STP).

In addition to their shielding designations, TP cables are divided into categories according to their bandwidth and other electrical properties, currently categories 1 to 5. A draft standard has been drawn up for categories 6 and 7. The rule here is: the higher the category, the higher the possible bandwidth. The bandwidth is determined by various physical properties of the cable. The commonly used UTP or STP cables in categories 3 to 5 can be used to transfer between 10 and 100 Mbit/s over a maximum length of 100 m with Ethernet or Fast Ethernet, while up to 155 Mbit/s can be transferred on category 5 cables with ATM. Cables in category 6 will have a bandwidth of 600 MHz and therefore allow data transfer rates of up to 1 Gbit/s.

TP cables are at present mainly used for star-type cabling and in some cases also for ring-type cable configurations.

Advantages:

• TP cables, and in particular the methods of preparing them for installation, are relatively cheap in comparison with optical fibres if bandwidth needs are low.

• TP cables up to category 5 are relatively easy to install and prepare.

• TP cables can be considered to be universal cables, because other services (such as telephony) can be used via these cables without major technical investment. If bandwidth needs are relatively low, existing telephone networks based on TP can also be used as data networks.

• It is easy to measure and test existing installations.

Disadvantages:

• Depending on the type of cable used (UTP to S/STP), the cable is surrounded by an electrodynamic field of greater or lesser strength. As a result of this, there is both a risk of interaction with other fields (for example from an adjacent cable or from power installations) and the possibility of data interception.

• The maximum cable length given the requirements commonly encountered today is limited to 100 m (including the necessary connecting cables and patch cables; cf. table below).

• In the case of unshielded installation cables (UTP) with a large number of pairs, there may be cross-talk between individual pairs.

Coaxial cable

Coaxial cables are mainly used for bus cabling or for connecting the network nodes in tree structures, for example. Thanks to the overall screening of the centre conductor, it can generally be assumed that the electromagnetic compatibility (EMC) of these cables is good.

Advantages:

• The bandwidth and unamplified transmission range are greater than in the case of TP cables. For both types of cable which can be used for the Ethernet protocol the upper limits, depending on cable type, are 185 and 500 m respectively (thin and thick Ethernet cable).

• The risk of cross-talk is lower with coaxial cables that with TP cables.

Disadvantages:

• Coaxial cables are generally more expensive than TP cables.

• Despite the coaxial design, the cables can be tapped and are sensitive to interference.

• Depending on the type of coaxial cable used, they have relatively wide bending radii and are therefore more difficult to install than TP cables, for example.

• Only a few network access protocols have been defined for coaxial cables. Fast Ethernet, for example, cannot be run on coaxial cables.

Optical fibres (FO)

Optical fibre cables use light in the visible to far infrared range for the transmission of signals. The structure of an optical fibre cable is similar to that of a coaxial cable. The actual optical fibre in the centre is surrounded by cladding, the optical properties of which are different from those of the core. Around the whole of this is another sheath to provide protection against mechanical and optical influences. Optical fibres are available in two versions: multimode and single-mode optical fibres. These two types differ primarily in terms of the possible bandwidth and the maximum length that can be attained without additional repeaters.

Optical fibre cables are generally used for bridging long distances (for example connections between buildings or floors) in a backbone, and in some cases in dual ring systems.

Advantages:

• The bandwidth and the unamplified range is greater than with copper cable.

• Tapping can be achieved only with substantial technical effort.

• Inadmissible rewiring/changes of line assignment can be easily identified by means of the available technology.

• Optical fibre cables are insensitive to all non-destructive ambient conditions, in particular to electromagnetic fields.

• Optical fibre cables require relatively little space.

• There is no cross-talk or interference between various optical fibres or an optical fibre and TP cables.

• With optical fibres the fire load is lower in comparison with copper cables, because they have less or different sheathing and also because as a rule less cable material is required over a given route.

Disadvantages:

• The installation costs for optical fibre cables are very much higher than for copper cables, mainly due to the splicing work required.

• The coupling components for the operation of optical fibres, especially for single-mode optical fibres, are more expensive than those for copper cables.

• Laying and preparing optical fibre cables requires special knowledge, special tools and special additional components (e.g. splice boxes).

• Optical fibres cannot be bent through just any radius, as required. This may make installation more difficult, as a result of the need to take account of possible and necessary cable routing options.

An overview of the length restrictions on cables for some of the common protocols (Ethernet, Fast Ethernet, FDDI and CDDI; cf. S 5.60 Selection of a suitable backbone technology) is given in the table below:

It should be noted that the lengths stated above are the maximum lengths in each case. This is often made up of the installation cable itself and the connecting cables (patch cables). For 10Base-T, for example, the length of the installation cable should therefore not exceed 90 m so as to leave sufficient length for patch cables. With some procedures (such as 100Base-FX) the length is reduced by the use of repeaters, depending also on the type of repeaters.

When installing new networks it makes sense to use optical fibres in the primary and secondary areas because they will be able to satisfy future requirements thanks to the high available bandwidth. For the tertiary area it needs to be examined whether the use of TP cables or optical fibres is possible and/or necessary according to technical and security criteria, and also whether it is justifiable from the economic standpoint (cf. also S 5.2 Selection of an appropriate network topography).