SOM Training Approach for OFDM Receiver

This paper presents a novel approach for performance improvement in WLAN receiver employing Orthogonal frequency division multiplexing (OFDM) technology. As an effective technology for wireless communications, OFDM is an important multicarrier modulation method that offers high spectral efficiency, multipath delay spread tolerance, and immunity to frequency selective fading. However, some challenging issues still remain unresolved in OFDM based system design, such as its high peak to average power ratio (PAPR). Methods used to reduce PAPR factor carry several disadvantages. Neural approach presented in this paper is the new method which provides function approximation method. WLAN receiver under consideration is Hiperlan/2. Modulation method is 16QAM with code rate and AWGN channel model is used. Self organizing map (SOM) and parameterless self organizing map (PLSOM) structures are used to improve bit error rate (BER) performance, which was degraded because of nonlinear distortions due to high PAPR factor.