Signal Theory – P2
Signal Processing
Signal processing is relating to the conversion between digital signals (binary) and analogue signals (electrical frequencies). This is usually done using a modem (modulator demodulator). A computer or transmission device sends bits (binary digits) to a modem which converts these to electrical pulses. A modem on the receiving end then coverts these electrical pulses back to 0’s and 1’s for another transmission device to be able to interpret the data that has been sent.
A bit is basically a string of 8 numbers that contains 0’s and 1’s. Each of these strings represents a number between 0 and 255 which is what the computer converts into useable data. To convert denary (standard number form) to binary you have to think of it as being similar to ones, tenths, hundredths and thousands (denary), but with 1 to 128 – doubling each time. You can have a number up to 255 because if you add all of the 8 numbers up you will get 255 (in binary this would be shown as “11111111” because it contains every number). I will show how numbers 1 – 7 are converted into binary below:
As you can see, if you add up the numbers in each of the rows they equal the same number that they are representing on the left. This is how a computer converts between denary and binary. Each of these binary digits (bits) makes one byte because there are 8 bits in a byte (the 0 column does not count as one).
Before being transmitted, bit sequences are also given a source and destination IP address so they know where they are going and who has sent them. Once they have been assigned this information this creates something called a data packet. Data packets also contain error control which is sent to help identify problems with the data once it has been received at the other end.
Because computers use binary which contains only 0’s and 1’s, it makes it possible to send data via an electrical signal (analogue). The modem converts these 1’s and 0’s into an electrical signal by essentially representing 1 as “on” and 0 as “off”. The positive frequency is equal to 1 and the negative frequency is equal to 0. The diagram below shows this:
When the signal arrives at the receiving end of the transmission, a modem then converts this analogue signal back into a digital signal which the transmission device (usually a computer) can understand. This is how modems modulate and demodulate.
Asynchronous Transmission
With asynchronous transmission, the receiver must acknowledge receipt of the data before the sender will transmit more. Asynchronous transmission uses start and stop bits to signify the beginning and end of a transmission. This means a transmission would become ten bits, for example; "1 0100 0001 0". The extra 1 at the start indicates to the receiver that the transmission is coming and secondly the 0 at the end signals that the transmission has ended. The zero and one at the beginning and end vary depending on the parity bits. The parity bits are to change the numbers in the bit sequence from odd to even and vice versa. This type of data is used when data is sent intermittently as opposed to a constant stream.
Synchronous Transmission
With synchronous transmission, both devices will synchronise with each other before any data is sent. Synchronous transmission does not use start and stop bits. Instead, it synchronises transmission speeds at both the sending and receiving ends by just using clock signals built into each component. A continual stream of data is then sent between the two devices. Because there are no start and stop bits the data rate is quicker but there are more errors because of this. The clocks can eventually go out of sync and the receiving device would have the wrong time that has been agreed in the protocol for sending/receiving data. This could lead to some bytes being corrupted because of bits being lost as the receiver cannot process them all as quick as the sender is transmitting them.
Transmission Media – P4
Different transmission media that are used are coaxial, optical fibre, unshielded twisted pair (UTP), and shielded twisted pair (STP) cabling. Wireless types are Bluetooth, WIFI, and satellite.
Coaxial cabling – this is used as a transmission line for radio frequency signals; in applications such as radio transmitters and receivers with their antennas, computer network connections, and distributing cable television signals. One advantage of coaxial cabling over other transmission media is that the electromagnetic field carrying the signal only exists in between the inner and outer conductor. This means the cabling can run next to other metal objects without electromagnetic power loss or interference. This supports up to 1 GHz bandwidth.
Optical fibre cabling – this is a transmission media that uses laser to transmit data. The cable permits transmission over longer distance at higher bandwidths than any other data transmission media. A cable contains hundreds of fibre optic strands which lasers reflect down off glass or plastic within the tubes. The different strands are different colours which change in accordance to the data that is being transmitted. Fibre optical cabling is often used instead of metal wiring because signals travel along them with less loss and they are also immune to electromagnetic interference. This supports up to 75 THz bandwidth.
Unshielded twisted pair (UPT) – these cables are usually found in Ethernet networks and telephone systems. The cables inside a twisted pair are all twisted in pairs; hence its name. The insulated wires are made from copper. A disadvantage of this type of cabling is that it is susceptible to electromagnetic interference as it is not shielded from this. This can degrade the signal quality; this is why it is not used in densely populated areas. An advantage of this cable is that it is much lower in cost than optical fibre and coaxial cabling. This supports up to 1 MHz bandwidth.
Shielded twisted pair (STP) - this is the same as UTP cabling except it has shielding around each of the twisted pairs which prevents electromagnetic interference. Because the shielding is made out of metal it can also serve as a ground. This is obviously a better cable for this reason but it is more expensive than UTP, especially if having to cable large areas. This supports up to 1 MHz bandwidth.
Bluetooth - Bluetooth is built into many modern devices to allow data to be transferred between them via a radio frequency. It is a short distance data transfer protocol. Bluetooth operates on a radio frequency of 2.4 GHz and only sends out signals of about 1 Milliwatt which will only be broadcast in a close proximity of about 10 metres. The lower power output reduces the risk of interference from other devices that operate on the same frequency.
WIFI – this is used for transmitting data over larger networks that Bluetooth. It is another networking protocol similar to Bluetooth with slightly different specifications. Bluetooth networks only support up to 8 simultaneous connections whilst Bluetooth supports up to 32. WIFI is similar to Bluetooth as it operates on the same frequency band; 2.4GHz. However, WIFI outputs a much stronger signal as it can broadcast from 50 to 100 metres depending on object interference. WIFI is convenient for small networks as it eliminates the task of wiring a network. But on the downside, it is prone to interference and if there are more than 3 different WIFI devices in a close area they are likely to interfere with each other and produce very poor signals.
Transmission Media Selection – M2
In certain situations, certain transmission media would be more appropriate to use than in other situations. Here I will discuss the appropriateness in different situations.
Under different circumstances, unshielded or shielded twisted pairs may be preferred. If a company has to install a lot of transmission cabling and they think it will be installed where there is little metal interference, then UTP may be preferred. This is because it would be much cheaper cabling a large area with this cable. If a company is installing a lot of cabling where there is already a lot of transmission cabling in the area or a lot of metal, then STP would be the better option.
If there is little metal or previous cabling in the area then there is a low risk of electromagnetic interference which would mean that UTP would be appropriate and the cheapest option. If there is a lot of metal or previous cabling in the area then STP would be the ideal option to eliminate the risk of electromagnetic interference. The only downside to this is that it would be much more expensive.
If there is a small network that is being created in a certain area of the business premises where it will not require a large amount of bandwidth, it may be appropriate to simply use a WIFI connection to supply this area with a connection. This is because it would be a lot cheaper than installing cables, and if the subnet does not require much bandwidth a WIFI connection would offer these necessary requirements. These could be used in certain areas of a business where this situation applies; but the WIFI networks cannot be too close to each other otherwise the signals may interfere with each other and it would slow down all of the WIFI networks. Therefore, most areas of a company would obviously need to be hard-wired. This may just be a quick, cheap an effective solution to supplying certain areas with a connection.
If a company requires a very high speed and large bandwidth connection, they may choose to install fibre optic cabling. An appropriate use for this would be having it installed to supply a large company with the internet. This way the cabling can handle the amount of bandwidth that the company require. If a standard twisted pair is used to supply this the bandwidth may not be good enough to supply a fast enough connection to all computer users within the company.
In certain situations, a company may prefer to have a satellite internet connection. This could be if the internet is slow in the area where the company is, or even if there is no internet in that area. A satellite connection can supply up to a 1 Gigabit per second connection. If this is the case, a coaxial cable would be used to link the satellite to the satellite receiver. This is because the coaxial cable can supply a fairly high bandwidth. If the company does have a satellite internet connection it may be appropriate for them to install optical fibre cabling for the backbone of the company’s network. This would ensure that all of the subnets on the main network would be supplied with enough bandwidth to make use of the internet connection. Twisted pair cabling may not have enough bandwidth to support the backbone of the network which in turn could “bottleneck” the fast connection speed. Another benefit of optical fibre is that there would be fewer transmission errors than using twisted pair.
If a computer uses wireless technologies they may prefer to use WIFI for a network as opposed to Bluetooth. This is because a Bluetooth network can only support up to 8 simultaneous connections and WIFI can support up to 32. Also Bluetooth only transmits in a 10 metres area whilst WIFI can transmit up to 100 metres. Therefore, a company may chose to use Bluetooth for transferring data between different devices in a small area but if they are looking to network a large group of computers then WIFI should be the preferred choice.
Techniques Used to Reduce Errors – M1
Transmission media are subject to channel noise or interference, and thus errors may occur during transmission from the source to a receiver. Various techniques are used in order to detect and reduce errors that occur when transferring data. These include parity bits, cyclic redundancy checks, checksums, and repetition codes.
Parity bits – these are bits that are added to the end of data to ensure the number of set bits is either even or odd. These are a simple technique put in place to detect errors. If an odd number of bits are transmitted incorrectly, the parity bit will come back incorrect which indicates there has been an error during transmission. Parity bits are only useful for detecting errors; they cannot correct any errors as there is no way to determine which particular bit is corrupted. If parity bits detect an error in transmission, the data must be discarded and resent again by the sender.
A disadvantage of parity bits is that when transferring data over a transmission media with interference or noise, it may take a long time for the data to be sent successfully because the data may need to keep being resent. In some cases the data may not even be transferred successfully at all. An advantage is that parity bits only use 1 bit of data.
Checksums - this is designed to detect accidental changes to computer data. A short fixed binary sequence is calculated for each block of data that is send and stores them both together. When a block is read or received the device repeats the calculation; if the new binary sequence does not match with the one calculated earlier, the block contains a data error and the device will need to request the data to be sent again.
Repetition codes – this is a coding technique detects whether a transmission route contains problems. A stream of data with repeated bits is sent over a channel to detect whether there is an error-free communication. When the bits arrive at the receiving end the device on the other end checks that the repeated bits are the same as what was transmitted. An example of a stream of bits that could be sent is “1011 1011 1011”. If the bits arrive at the receiving end as “1010 1011 1011” then the first block of bits do not match and there is obviously a problem with the transmission route.
Repetition codes are not very efficient because they only check the route the data is being sent and do not check whether the actual data has been transferred without errors. They can also be susceptible to problems if the error occurs in the same place for each group of bits sent. Using the previous example; if “1010 1010 1010” was continually being received, errors may not be detected so it is not completely accurate. An advantage of repetition codes is that they are an extremely simple technique.