Caleb koech -DCS/4492/2026
Emmanuel-DCS/03414/2025
Key Points
Parity Check: Detects odd‑number bit errors with minimal overhead.
Checksum: Common in TCP/IP for detecting varied error patterns.
Cyclic Redundancy Check (CRC): Powerful burst‑error detection used in
Ethernet and Wi‑Fi.
Hamming code: Corrects single‑bit errors in memory and satellite links.
Definition: Techniques used to identify when bits have been altered during
transmission over a noisy channel, ensuring data integrity.
Parity check
A single redundant parity bit is appended to each data unit.
Even Parity: Total number of 1s (including the parity bit) is even.
Odd Parity: Total number of 1s is odd.
Limitation: Detects only an odd number of bit errors; two flipped bits remain
undetected.
Two‑Dimensional Parity Check (Longitudinal
Redundancy Check)
Data is arranged in a rectangular table of rows and columns.
Parity bit calculated for each row.
Parity bit calculated for each column.
Strength: Improves detection of burst errors (multiple consecutive bit flips).
Checksum
Commonly used in upper‑layer protocols such as TCP/IP.
1. Generator (Sender):
Divide data into equal k‑bit segments.
Add segments using 1's complement arithmetic.
Complement the sum to obtain the checksum.
2. Receiver:
Add all received segments including the checksum.
If the final sum is all 1s (or 0 after a final complement), the data is
accepted.
Cyclic Redundancy Check (CRC)
A powerful detection method employed in Ethernet, Wi‑Fi, etc., based on binary
division.
Generator Polynomial (G): Agreed upon by sender and receiver.
Process:
1. Append n zeros to the data, where n = degree of G.
2. Divide the extended data by G using XOR division.
3. The remainder becomes the CRC bits, replacing the appended
zeros.
Verification (Receiver): Divide the received frame by the same G; a zero
remainder indicates no error.
Hamming Code (Error Correction)
Extends error detection to also correct single‑bit errors.
Multiple parity bits are placed at positions that are powers of 2 (e.g., 1, 2, 4, 8,
…).
Allows identification and correction of one erroneous bit within the data word.
Comparison of Error Detection/Correction Methods
Method Primary Use Error Overhead Typical
Detection (extra bits) Application
Capability
Simple Parity Low‑noise links Detects 1 bit per unit Basic serial
odd‑number links
errors
Two‑Dimensional Burst‑error Detects Row + column Storage
Parity prone links multi‑bit bursts bits devices, RAID
Checksum Software Detects many Depends on TCP/IP, UDP
protocols error patterns segment size
CRC High‑speed Detects burst Degree of Ethernet, Wi‑Fi
hardware and random polynomial
errors
Hamming Code Error Detects & Multiple bits Memory
correction corrects 1‑bit (log₂n) systems,
errors satellite links