An Optimized Resource Efficient Approximation of max Operator for Recalculation in Turbo Code Decoder

A family of powerful and resource efficient Turbo codes is developed for sensor networks. In the hardware realization of Turbo code decoder, recalculation of the forward or backward state metrics is an effective method to reduce the size of state metrics cache. To facilitate recalculation of the sta...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Proceedings of the ... Enterprise Systems Conference (Online) s. 168 - 172
Hlavní autoři: Zhan, Ming, Wu, Jun, Wen, Hong
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 01.09.2017
Témata:
Algorithm design and analysis
approximation
Approximation algorithms
Decoding
Hardware
MAP algorithm
max operator
Measurement
Quantization (signal)
recalculation
Turbo code
Turbo codes
ISSN:2572-6609
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:A family of powerful and resource efficient Turbo codes is developed for sensor networks. In the hardware realization of Turbo code decoder, recalculation of the forward or backward state metrics is an effective method to reduce the size of state metrics cache. To facilitate recalculation of the state metrics, an optimized max* operator is presented in this paper. Calculation of the proposed max* operator includes two correction terms, and each one is approximated by using simple comparing, shifting and addition operations. Moreover, approximation of the proposed max* operator is specially optimized for reverse calculation. Although there is small error in the reverse calculation, it has been shown that the maximum error is about three percent, and the proposed approximation is validity for reverse calculation.
ISSN:2572-6609
DOI:10.1109/ES.2017.34