31 May 2011

African Undersea Cables

Steve Song's map of African undersea cables strikingly illustrates the changing landscape of African network connectivity.  To get the most recent version of his map it is best to obtain it from his site.  Here is the version of May 2011 reproduced with his kind permission.
It is almost unimaginable that, until recently, South Africa's only real undersea connectivity to the outside world was via the SAT3/SAFE cable system - a system which is hard to spot on this map.

RAU on Netscape Navigator 3.01

I have just booted up one of my old laptops.  One of the prominent applications on the (rather short) list of Windows 95 applications was Netscape Navigator 3.01.  Upon opening it I found the following page still cached in it (dated 10 April 1997):
It should be noted that this is not a monochrome image.  The display of the laptop was (and still is) monochrome.    The smudges on the picture are imperfections developing on the display rather than dirt.

This was an early home page of the Rand Afrikaans University (RAU).  (RAU was merged with other academic institutions to form the University of Johannesburg in 2005.)  The homepage was typical of early university homepages.  Usually they were developed by some 'techie' without any corporate involvement.

Scrolling down one finds the following useful information:

The 'general information' section consists of six links.  Just finding something else to link to was an achievement - hence the links to sites such as CNN.

Wired: 1919 - from Shorpy

Shorpy is a great source of images from the distant past.  Quite a number of them deal with topics of interest to networking enthusiasts.

Here is one such example that dates from circa 1919 and takes a "behind the scenes look at communications tech some 80 years after the telegraph tapped out its first message."
Wired: 1919
(Used with kind permission of Shorpy)
To really appreciate the detail one should look at the large version of the picture available from the original Shorpy post.  The comments left by Shorpy readers are usually also worth reading to gain a deeper insight into the details depicted in the photo.

STP (Shielded Twisted Pairs)

As the name indicates, STP (Shielded Twisted Pairs) consists of pairs of wire that are shielded against electromagnetic interference.
Note the silver strands that 'wrap' the entire cable, as well as the silver foil that 'wraps' (or shields) each pair.

The connector (which is physically compatible with an RJ-45 socket) is also shielded from three sides.

This is an example of Cat 6 STP (Category 6 Shielded Twisted Pairs).

29 May 2011

50 micron optical fibre

Here is a 50μ optical fibre through which an ordinary laser pointer's light is conducted.  The fibre is a patch cable with SC connectors at both ends.  Here is the light emerging from the one end of the ST connector.  A matchstick has been included in the picture for a sense of scale.  The orange cladding of the patch cable is visible in the background.
When the laser at the other end is switched off, one gets an even better feeling for the diameter of the fibre.


Satellites use microwaves to transmit signals.  Such signals should not be obstructed by any solid objects between the sender and receiver.  The phase line of sight is often used to refer to this requirement for microwave antennas.

When using satellites line of sight is not a major concern - except at the point where the signal from the satellite has to enter the building.  Satellite reception dishes are therefore placed outside and the signal is carried into the building by (typically coaxial) cable.  The dish has a parabolic shape and focuses any signals that it reflects to the end of the arm attached to the bottom of the dish.
The device that is attached to this other end is the LNBF and must do many things.  Firstly, it guides the signal towards a tiny antenna. located inside it.  The part that does this is known as a feedhorn, and used to be a separate device.  However, in dishes that are intended for the consumer market, the feedhorn has been included in the LNBF.  In fact, the F in LNBF indicates that it is an LNB with a feedhorn, or an LNB-feed.  The LNBF then converts the microwave signal to a lower frequency that can be carried into the building by cable.  (Cable can carry microwave, but only over small distances because of attenuation.)  This process is known as block-downconversion.  This explains the B in LNBF.  The letters L and N represent the notion of low noise: the LNBF amplifies the signal it received before sending it out along the copper cable.  This has to be done in a special manner that does not also amplify noise that may have been present.  Exactly how this works is not important for our purposes - just note that the LN represents low noise amplification.

So, here we have a typical LNBF - mostly used to receive satellite television.  The matchstick has been included in the picture to give a sense of scale.
Note that this LNBF specifies that it is intended for Ku-band satellite.  This will be encountered again later.

Patch panel

In a typical current Ethernet installation cables will run from each of the network points in an office, building or even (small) organisation to a central point where they all can be connected.  In a very small installation this central point may consist of a switch or, in years gone by, a hub.  However, the moment the installation becomes somewhat bigger all these cables will converge at a patch panel.  A patch panel simply consists of a series of (RJ-45) sockets to which cables may be attached from the rear.  A 24-port patch panel is depicted below.
At the rear one finds the typical slots in to which wires may be pushed down.  In this case the colour coding makes provision for both T568A and T568B as explained in an earlier post on UTP cable.
After installation the network points may be connected with a switch, hub, or other device using patch cables. In the picture below the patch panel is on the bottom and blue and green patch cables are used to connect points to the switch at the top.

An old Wireless Access Point (WAP)

Below is my very first Wireless Access Point, which a decade ago cost as much as a computer did.
SMC2655W access point
Its age is evident from the fact that it 'boasts' a speed of 11Mbps - it is an IEEE 802.11b WAP.

Besides its low speed  its other 'feature' that distinguishes it from today's WAPs is its use of SNMP to configure it.  Modern consumer-oriented WAPs may typically be configured using a variety of application layer protocols, of which Web-based configuration is typically the most popular.  Other protocols supported may include Telnet and SSH.  SNMP is typically not supported on modern WAPs.

28 May 2011


Here is a picture of unshielded twisted pair (UTP) cable.  It is clear that each pair of wires is twisted around one another.  It is not entirely clear from this picture that the rate at which each pair is twisted is different.  It is also not obvious from this picture that the twisted pairs are then twisted around the other pairs.  Even though it may not be obvious, it happens to be true.

At the ends of such cables one expects plugs or sockets.  RJ-45 plugs are the ones to use with the cable depicted above.  Below is a picture of two unused RJ-45 plugs (from the top and bottom).  For the sake of comparison an RJ-11 telephone plug (attached to a telephone cable) is shown on the left.
To attach the cable one would separate the individual wires, push them into the plug in the correct order and then make the connection permanent by crimping it.  Crimping pushes the pieces of copper (that can be seen on the pug in the centre of the picture above) through the wire inserted below it.  This keeps the wire in place and provides an external contact via the copper strip that still protrudes slightly.  Crimping also pushes the sleeve (on the other end of the plug as the copper) in so that it helps to hold the cable in place.

The order in which wires are arranged depends on the wiring scheme used at a particular installation.  The following diagram depicts the two alternatives.  T568A starts with the green and white wire in position 1 and ends with the solid brown wire in position 8.  In contrast, T568B starts with the orange and white wire in position 1, but also ends with the solid brown wire in position 8.
     T568A                 T568B
1                         1                   
2                         2                   
3                         3                   
4                         4                   
5                         5                   
6                         6                   
7                         7                   
8                         8                   

The 'strange' use of blue in both standards in positions 4 and 5 is intended to ensure compatibility with telephone wiring.  Although the RJ-11 plug has four pins, ordinary telephone connections only use the middle two.  An RJ-11 plug physically fits into an RJ-45 socket and hence installations that use either of these wiring schemes can use the same wiring for telephone or network services.  (In some cases it may even be and/or.)

The next picture shows an RJ-45 socket from the rear.
The 'A' stamped on it indicates that it has been colour coded for the T568A wiring scheme.  In order to install it (using T568A) one merely places each wire into the groove next to its corresponding colour - and then pushes it down into the groove with something known as a pushdown tool.  As the wire is pushed down the copper contacts of the socket cut through the insulation of the wire.  In addition, the contacts are shaped in such a manner that they will retain the wire in the groove and maintain the contact.


Here is a piece of 'standard' coaxial cable.  (It was RG-58/U if I recall correctly.)

This is an example of so-called thinnet or, specifically in the Ethernet context, 10Base2 cable.  The name thinnet stems from the fact that it is much thinner than the original coaxial cable used for networking.  The name 10Base2 referred to the fact that the cable was used to carry baseband signals at 10Mbps over distances of up to 200m.

In order to connect computers to such  cables, BNC connectors were used.  In the picture below three such connectors are shown: a T-connector, a male connector crimped to a cable and a 50Ω terminator.  The matchstick is included to give a sense of size.

In normal use cable would be installed with male connectors at any point where it may be necessary to link a node to the network.  If no node was present, the two ends of the cable could be joined by using a female-to-female connector to join the two male ends.  However, if a node was present the connection would be made with a T connector as shown below.  A fly lead could then be used to connect this new point with the female connector on the network card of the node.
As is obvious from the picture, it is easy to break the connection again by twisting the bayonet part of the connector and then simply pulling the connectors apart.  The B in BNC indeed stands for bayonet.

Today these connectors are not often seen in real networks and a probably most useful the create BNC art.  Below is a piece by me, entitled 7 Tees terminated twice.

Lukasrand Tower

This is what Telkom's Lukasrand Tower normally looks like from my office window:

Its place in this network museum stems from the fact that it is used to relay communications.  Zoomed in a bit one can clearly see the microwave dishes (and other antennas) used for this.

(When these pictures were taken the tower still sported its 2010 Soccer World Cup livery.)


This picture demonstrates a TOSLINK optical fibre conducting light though multiple loops.  An ordinary laser pointer is aimed at the one end of the fibre (at the bottom right hand corner of the picture).  The light emerges at the other end (top left).

Here are the two TOSLINK connectors.  The matchstick between them provides a sense of scale.  The fibre has a diameter of about 1mm.