Hawkynt

Turbo Code

Parallel concatenated convolutional codes with iterative decoding. First practical codes to closely approach Shannon limit. Used in 3G/4G mobile communications. Two recursive systematic convolutional encoders separated by interleaver. Iterative MAP/SOVA decoding with extrinsic information exchange.

Properties

Property	Value
Category	Error Correction
Sub-category	Concatenated Code
Security status	🎓 Educational Only
Complexity	Expert
Inventor	Claude Berrou, Alain Glavieux, Punya Thitimajshima
Year	1993
Origin	🇫🇷 France
Source	algorithms/ecc/turbo-code.js

Security

Status: 🎓 Educational Only

Known vulnerabilities

Issue	Description	Mitigation
Decoding Complexity	Iterative MAP/SOVA decoding requires significant computational resources. Complexity increases with frame length and iteration count.	—
Decoding Latency	Iterative decoding introduces latency proportional to iteration count. Critical for real-time applications.	—
Error Floor	Low error floors may occur at high SNR due to low-weight codewords. Mitigated by interleaver design.	—

Documentation

• Wikipedia - Turbo Code
• 3GPP TS 36.212 - LTE Multiplexing and Channel Coding
• Turbo Code Tutorial
• NASA Turbo Code Overview

References

• Original 1993 ICC Paper
• Berrou et al. - Near Shannon Limit Error-Correcting Coding
• 3GPP LTE Physical Layer Specification
• Iterative Decoding Survey

Test vectors

4 vectors ship with this algorithm and run in the test suite. Byte values are hexadecimal.

Vector 1 — Turbo code all zeros - K=4 rate 1/3

Field	Value
input	00000000
expected	000000000000000000000000

Vector 2 — Turbo code single bit - K=4 rate 1/3

Field	Value
input	01000000
expected	010101000101000000000101

Vector 3 — Turbo code pattern 1100 - K=4 rate 1/3

Field	Value
input	01010000
expected	010101010001000101000100

Vector 4 — Turbo code alternating - K=4 rate 1/3

Field	Value
input	01000100
expected	010101000100010101000001

---

← All algorithms

This site is open source. Improve this page.