3.4 still networks

.

Hacking is generally taken to mean unauthorised access whether actual harm is done or not.

After professors Liv and Zempel from Israel.

Auckland Cathedral from the top of Symonds Street - near the Sheraton Hotel.

As you might imagine, run-length encoding an image like this doesn't achieve a great deal of compression.

On this page: [ software | integrity | security | applications ]

Software

Over the hardware sit the various protocols that are used for transmission (these may be implemented in hardware and/or software). At the top sits the application layer where users interact with the systems.

The kind of software described in the subject guide , the " communications software " would probably belong in the middle layer of this simplified diagram. This kind of software deals with connections between LAN's and WAN's and deals with aspects such as protocols and security .

Back to top

Integrity

Communication lines are subject to (electrical) noise . These may be caused by anything from wind on the cables and connectors to magnetic storms (eg, caused by sunspots).

These disturbances can cause bits to be lost or changed in the data. It is important to detect these corrupted signals. Some errors can be detected and corrected but usually when an error is detected, a re-transmission of data is requested.

The subject guide mentions parity checks and checksum (block character checks):

Parity checks

• The number of bits that are 1 are counted.
• An extra bit is added such that:
• There is an even or odd number of bits in the whole number

Examples:

Using this method an error may be detected.

The bits are transmitted as one unit (ie, despite the example above, the parity bit is not sent separately). Think of an example where an error will not be detected.

top of section

Checksum (BCC)

• Values of successive bytes are added together.
• The sum is transmitted.
• The sum is re-calculated by the receiver.
• If it is the same it is assumed no errors occurred.

Here is an example of the message "hello!" beiong sent with a block character check (bcc) every 3 bytes:

Message "Hello!" - checksum sent every 3 bytes:

Note that 281 can't be sent in an 8-bit byte and is therefore actually sent with the high bit truncated (or modulo 256):

  H    72   01001000
 e   101   01100101
 l   108   01101100

  sum 281   100011001 - transmitted as 00011001 (25)

Checksums (BCC's) can be combined with parity, consider the following:

  H   72   11001000
 a   97   01100001
 l  108   11101100

   sum = 1000010101 - transmitted as 00010101 (21)

In this example, which bit is the parity bit and is odd or even parity being used?

top of section

Exercises
The following sequence of bytes is received. BCC's are every 3rd byte. A parity bit is being used in the MSB of each character byte - ASCII code is used.

       C3 53 B2 C8 30 B1 3F 20

What is the message? What type of parity is being used?

Which character in which block is corrupted in the following sequence?

    50 E0 72 A3 69 74 F9 D6

Calculate the checksum for these words (just one checksum for the whole word) - use the table below:

• Hello
• HELLO
• Heklo
• Hfklo

An Excel spreadsheet can be used to help with these exercises. Download this sheet and give it a try.

Examples 3 and 4 might be the result of corruption, comment on example 4.

top of section

Error correction
When errors in transmission are deteced they need to be corrected. The most common type of error correction is ARQ (automatic repeat request), re-sending continues until correct communication is achieved. Some codes allow a small number of corrupt bits to be corrected without retransmission ( forward error correction ).

This shows that parity checks can be applied to each character, as we have seen, but also the same bit from a number of characters can also be checked.

The errors in parity intersect at the incorrect bit and it can be "flipped" to achieve correction.

top of section                 Back to top

Security

The network brings with it some concerns about security as noted above.  Obviously it is easier to access information when it is on a computer that is physically linked to many other computers. Several precautions can be taken:

Privilege (or access permission)  
The network OS allows users and administrators to designate whether certain files are accessible to only themselves or to restricted groups of users on the system. Common forms of access are read , write , delete files and folders. These can be applied to single users or groups of users.  

Passwords
It is commonly said that no networked system is completely safe from hacking but they are safe enough for most users (just like the roads are safe enough for most people - that does not mean you will never be involved in an accident).

However, you can help yourself by careful selection of adequate passwords . Most security breaches are the result of passwords that are easily guessed or even written down on a yellow sticky and attached to the user's monitor. It happens.

Passwords are designed to keep unauthorised users out of a system or parts of a system that are off limits to them.

Physical security
This can be the most effective means of security. A user can physically lock a door to prevent access to a workstation or group of workstations. Some workstations can be kept disconnected from the network at all times or only at times when synchronisation is needed. Subnetworks can be used, in school for example, to keep the student computer room separate from the assessment and administration database.

Sometimes access to a network might require a "key disc" or some small hardware device connected to a given computer port (a "dongle"). These can be removed to improve physical security.

Encryption
This is used to "scramble" or code data that is transmitted or held in computer systems such bthat it is unreadable . It is said that no code is actually unreadable but the real trick is to make it take so long to break a given code that by the time it is done, the data has no value.

Encryption can also be applied to data files that reside on disc.

Back to top

Applications

Many files are transmitted across all kinds of networks, including the internet (itself an extremely large WAN ). Any network has a certain (maximum) carrying capacity called its bandwidth . The more data that travels across a given network connection, the more bandwidth is used and, at some point near it's capacity, the slower the data is transmitted.

Since the bandwidth is fixed and increased use reduces speed, the more that volume can be reduced the more likely it is that higher speeds of transmission can be achieved. For this reason, files are often transferred in compressed format.

Both text and graphics documents can be reduced in size by compression techniques. The subject guide does not require you to know details but only principles so we'll look at those. The details are available elsewhere on the web (go look if you care).

Text compression
A couple of simple methods are tokenizing and Ziv-Lempel.

Tokenizing replaces certain common words or letter combinations with special symbols or tokens that take up only one byte. It is language dependent.

Ziv-Lempel uses pattern matching, it examines text and replaces any repeating pattern with a reference to the original pattern, for example, in the words of a famous cartoon character:

I am what I am.

The second I am is replace by a reference to the original: it's position and length:

I am what 1 4.

The 1 is the position of the repeating sequence and 4 is its length. This technique is not language dependent and works well with large pieces of text.

Graphics compression
Bitmaps are representations of pictures in which each picture element ( pixel ) is represented by a binary code corresponding to a given colour. In filing systems they are often distinguished with the file extension BMP .

The image below has 7 colours. Therefore, each pixel could be represented by a 3-bit colour code.

The top line of 16 blue pixels would therefore need 48 bits to store. However, starting at the top left, there are 19 blue pixels then 4 white pixels then 5 blue pixels again. If blue is 3 and white 7 then the early part of the picture could be coded as:                                  3 19 7 4 3 5 Here, most of the top two lines are stored in the same 48 bits.

This is RLE (run length encoding) as used in GIF images. It works well with "blocky" diagrams where there are large blocks of similar colour. RLE is "lossless" compression; all of the information in the original picture is preserved.

As an exercise you could compare the size of the RLE file with that of the full bitmap.

The image on the right shows an enlarged part of the sky. Many of the shades of blue are quite similar, we can compress the file by not storing information about every single shade of sky blue but just representing very similar ones with the same code. Some information is lost so this is known as "lossy" compression.

Back to top

related: [ Common core home | previous: networks (intro) | next: network applications ]

There are several "layers" to a network; the technical details are defined by the ISO (ISO).

However, we don't need the details but it should be apparent, as for computer systems in general, that the hardware (eg cables, other physical links) is at the base of everything.

Questions or problems related to this web site should be addressed to Richard Jones who asserts his right to be identified as the author and owner of these materials - unless otherwise indicated. Please feel free to use the material presented here and to create links to it for non-commercial purposes; an acknowledgement of the source is required by the Creative Commons licence. Use of materials from this site is conditional upon your having read the additional terms of use on the about page and the Creative Commons Licence. View privacy policy.