Application Pid Channel Estimation Method And Codec Dsp Realization

It is well known that the accurate and timely channel state information (CSI) is required in achieving the satisfactory performance of channel codes used in broadband mobile systems. The correct CSI can provide the exact priority probability to the decoder, especially for the code with iterative decoding. Therefore, the CSI estimation is a necessity in broadband mobile systems. The main work includes:1. This paper presents a blind estimation method which combines the CSI estimation with decoding. The channel information is estimated by a linear regression model using the channel decoding results. The fluctuation of the channel information is monitored and forecasted through the Proportional-Integral-Differential (PID) controller. The performance of the proposed method applied to Turbo codes in Rayleigh channels is analyzed and compared with that of traditional methods.2. An innovative and successful design for channel estimation in the orthogonal frequency division multiplexing (OFDM) system is presented. The method joints CSI estimation and decoding with the PID controller and wavelet decomposition. PID controller is discussed. In addition, the performance of this method and the comparison between this method and the traditional one in different channels are given.3. We present an efficient implementation of the encoder and decoder for Turbo code and low-density parity-check (LDPC) code on a fixed point digital signal processor. The block-operation and a simplified decoding algorithm were used to reduce the running time for LDPC code. On a Texas Instruments TMS320C64xx DSP at 600MHz, we get a great lot of testing data for the comparing analyses. The performance on the running time, throughput and BER and the influence of the frame length and SNR are given.