Telefonaktiebolaget Lm Ericsson (publ)

Gerd M&ouml;rsberger

Gian Huaman-Bollo

Helmut Leuschner

A turbo coder block having a parallelization of degree n achieves increased processing speed. Each parallelized turbo coder block includes a first storage unit to store n samples of an input signal and a second storage unit to store n samples of at least one output signal of the parallelized turbo coding block. The parallelized turbo coder block further includes a bank of n delay units and is adapted to parallel process n samples of the input signal such that two delay units of the bank directly receive subsets of the n samples of the input signal, and an output signal of one delay unit is supplied to two delay units in the parallelized turbo coder block.

Parallel turbo coder implementation

The invention relates to a parallel CRC generation circuit comprising an input register means (I), an output register means (C), a number of XOR gates (XOR -XOR ) and a coupling means (CM) that feeds predetermined ones of the output lines (C -C ) of the output register means (C) and output lines (I -I ) of the input register means (I) as inputs to the respective XOR gates. According to the invention a matrix representation of the state change based on the selected CRC polynomial is set up and evaluated, such that the coupling means (CM) only uses the minimum number of feedbacks of the output lines and feed-forwards of the output lines of the input register means (I). Thus, the parallel CRC calculation circuit according to the invention has no redundancy and uses only a minimum hardware amount.

Parallel CRC generation circuit for generating a CRC code

In a telecommunication system where a plurality of user channels are processed in a time-slotted manner, a transmitter and receiver can perform bit error rate measurements for a plurality of user channels by only using one PN-generator (T-PN) and one state memory (ISM). Whenever the beginning of a new time-slot is detected, a last-stored phase state is read out from the state memory (ISM). When detecting the end of the respective time-slot, the phase state then present in the PN-generator (T-PN) is stored into the state memory (ISM) to be used for re-initialization of the PN-generator (T-PN) for the same time-slot in a succeeding frame. Thus, one PN-generator (T-PN) is enough for generating PN-sequences for a great number of user channels.

Growing community of inventors

Bubenreuth, Germany

Gerd Mörsberger

Growing community of inventors

Bubenreuth, Germany

Gerd M&ouml;rsberger

Gerd Mörsberger