Dynamic wordlength calibration to reduce power dissipation in wireless OFDM systems

This paper describes low power architecture by using a dynamic wordlength technique in wireless orthogonal frequency division multiplexing (OFDM) system. The number of wordlength in digital signal processing (DSP) has to be carefully determined because wordlength affects system performance and hardware cost. Dynamic wordlength technique is applied to a fast Fourier transform (FFT) processor and a Viterbi decoder in OFDM receiver. The proposed method searches an optimum wordlength combination of FFT processor and Viterbi decoder by comparing output binary data while changing wordlengths. This operation is done by use of intervals in packet waiting. This approach leads to achieve the power reduction up to 23.9% with a desirable packet error rate (PER) in multipath channel environment.