Error Correction and Detection
Digital bit plays a vital role in the processing and communication in which the signals are converted into digital bits. These digital bits are then processed and transmitted from one system to another system. These are data bits are grouped and transmitted which are called data stream. In this entire steam if any single bit changes during processing or communication, it leads a high error in the output. With increase in digital techniques the errors are also high due to noise and these effects the entire data steam if it is not corrected because these errors will be propagated if they are not detected and corrected.
The errors can occur in the systems or in the communication network and to avoid the errors in the communication network, the network should be able to transfer the data from one device to another device with highest accuracy. In the systems level the errors must be detected and corrected at the input acceptance check for reliable processing.
Bit Error: if a bit 1 is changed to 0 or 0 changed to 1 due to noise and it is called bit error.
Types of errors
Single Bit error: If a bit is differed from sending and receiving information that is called single bit error. In the below figure the 0 bit is changed to 1 in the receiver side due to this noise, this results in single bit error
Single bit errors are very rare in the communication system because the noise must not a short duration that it can only disturb only one bit. So these errors in the communication system are less however this type of errors can happen in parallel communication where the noise will distribute in all the communication lines.
Multiple bit error: If two or more bits are differed between sending and receiving data stream, then the error is called multiple bit error. The position of the bits may differ i.e. the bits may not be consecutive.
Here two bits are changed from 0 to 1 and 1 to 0 which leads to two bit difference between the sending and receiving information which is called multiple Bit Error. This type are errors are also common in parallel and serial communication and it leads to some difficulty in detecting and correcting.
Burst errors: The burst means two or more bits changed in the transmitted data and the error burst is calculated from starting error bit to the ending error bit. The Burst may contain zero error items also in addition to the error bits.
These errors are most common in serial communication where the noise can interfere the some amount of bits in sequence or in multiple bits. The number of corrupted bits depends on the data rte and noise duration. The error correction of this type is difficult which leads to incorporation of complex detecting and correcting algorithms.
Causes of Errors
Transmission Errors: These errors are occurred in the communication media due to signal distortion or attenuation.
Storage errors: These errors occurs due to data storage errors where the memory tracks will corrupt during writing of by any power supply failures.
Error detection is the process of verifying the received information whether it is correct or not at the receiver end with out having any information of sent original message. In sender side some redundant bits are added to the original message based some property of message signal (i.e. parity bits) and in the receiver side by scanning this redundant bits, the error in the message will be predicted.
Types of Error detection Checks
- Parity Checking
- Longitudinal Redundancy check (LRC)
- Cyclic Redundancy Check (CRC)
- Check Sum
Error connection is the next step of error detection. Once the receiver detects the error in the message the error will be corrected by the error correcting codes. It some cases it may ask the sender to resend the data. Hamming distance code is the best error correction technique used in the receiver to detect an error.