3.1.5 Error Detection & Correction Flashcards
Electromagnetic interface EMI
Strong sources of electrical energy and radio frequencies can cause disruption to other signals.
One signal that can get disrupted is the path taken by data.
As a result binary 1s and 0s could get ‘flipped’
So an Original signal that was supposed to represent some information could get garbled, that data received might represent something else entirely.
e.g 1000110 in ASCII = F, but if the second last bit flips it becomes 1000100= D.
For successful data transmission, protocols in the OSI model need to contain methods for detecting these possible errors.
Parity Bit
Some methods of representing data in binary do not use all 8 bits in a byte.
This ‘spare’ bit can then be used to make the total number of 1s up to either an even or odd number.
even parity
1110011 becomes 11110011
weaknesses in this method of error detection
cannot work out which bit is different from the transmitted signal
Parity Bit points for exam answers
“11011010 includes a parity bit”
5 1s so odd parity
original binary pattern only had 4 1s
if another 1 added at start then that must have been there to equal an ODD number of bits because the 8 bits transmitted, includes the parity and is the correct data.
in question might tell original and received data- you have to describe how error will be picked up.
e.g. if 10011010 is received- even no. of 1s, odd parity was being used= error.
however, if 10011010 received error will not be detected as odd no. of 1s.
Checksum
Data is transmitted in packets.
packet contain data being transmitted and
source/ destination of IP address
no. of packets that this packet is part of.
Before data sent, binary total calculated.
value is sent as a byte at the end of the packet- checksum.
when packet received- as part of protocol being used- checksum recalculated.
if values do not match, data have to be retransmitted- correction.
better for checking larger amounts of data.
Negatives of Checksum
adding new data- all adding new data.
not suited to small amounts of data.
suited to large.
Cyclic redundancy check
uses same principal of checksum.
value added to end of a packet.
value is much more complex than just ‘a total of all the 1s in the rest of the packet’
CRC produced by modulus arithmetic.
CRC method used for both transmission and storage of data.
Negative of CRC
Sum more complicated to work out.
Echo checking
receiving device sends data back to transmitting device.
transmitting device compares this data with original data.
transmitting device retransmits data if there was an error.
AKA loop back check.
Negative of Echo checking
Data traffic
Double
Positive od Echo checking
spot any error
