Error detection in serial data transmission

Errors in serial transmission are caused by noise, timing differences and interference. Common detection techniques are:

1. Parity check (VRC, Vertical Redundancy Check)

An extra bit is added to each character so that the total number of 1s is even (even parity) or odd (odd parity). The receiver recomputes it. Detects all single-bit errors (and any odd number of bit errors), but not an even number of errors. Example: data 1011001 (four 1s) -> even parity bit = 0; odd parity bit = 1.

2. Longitudinal redundancy check (LRC)

A parity bit is calculated for each bit column over a block of characters, and the resulting block check character is sent at the end. With VRC it detects most burst errors.

3. Checksum

Bytes of the block are added (modulo 256) and the two's complement of the sum is sent. The receiver adds all bytes plus the checksum; the result must be zero.

4. Cyclic redundancy check (CRC)

The data block is treated as a polynomial and divided (modulo 2) by a generator polynomial (e.g. CRC-16: x^16 + x^15 + x^2 + 1; CRC-CCITT: x^16 + x^12 + x^5 + 1). The remainder (CRC bits) is appended; the receiver divides again and a non-zero remainder means error. It detects all burst errors up to the CRC length and most others. Used in USB, Ethernet, disk.

5. UART-detected errors

• Parity error: parity does not match.
• Framing error: stop bit not found where expected (wrong baud rate or noise).
• Overrun error: a new character arrives before the previous one was read.
• Break detection: line low for more than one frame.

6. Others

Echo (loopback) checking, handshaking with ACK/NAK and retransmission (ARQ), and forward error correction (Hamming code) that can also correct errors.