Study On Wavelet Packet Modulation System Based On Channel Identification

Since the three generation (3G) of mobile communication has come, mobile users are increasing quickly. Besides providing high data rate, the mobile communication systems are required to support real-time speech service and non real-time services such as pictures and data files etc. However, the complex wireless channels and limited frequency spectrum source put forward stricter requirement to the wireless transmission. As a key wireless transmission technique, the select and design of wireless modulation plays an important role in bettering wireless communication performances. Nowadays people prefer to using multicarrier modulation (MCM) technique in order to provide the wide-band, high-speed rate services. Wavelet packet modulation (WPM) is such kind of typical implementation of MCM, it has potential ability for multipath fading mitigation. In this thesis, the flexibility of WPM on subchannel distribution and the multirate diversity strategy are taken into account in the study of performance improvement of WPM.1.This thesis discussed the multiresolution analysis and reviewed fast discrete wavelet transform and fast discrete wavelet packet transform. The theory and implement of the WPM have been given. In WPM the bit error rate (BER) of subchannels performance in the additive white gaussian noise (AWGN) channel, Rayleigh flat fading channel and frequency selective fading channel has been shown.2. Based on the complex wireless channels and analysis of channel character, the BER of the subchannels of WPM in different channel types are analyzed, and an optimum wavelet packet tree which has good performance is achieved by implementing tree pruning for the full wavelet packet tree in the frequency selective fading channel.3.The multirate diversity of WPM is utilized, and both channel identification and joint estimation strategies at the receiver are accomplished to improve BER of WPM in different channel types. As optimization of demodulation, joint estimation strategies and channel identification are banded together in this thesis. Simulation results show that both BER and noise immunity of WPM in different channel types are observably improved, and the channel identification plays an important role in choosing suitable demodulation strategies to improve performance of WPM system before demodulation.4. The flexibility of WPM on subchannel distribution and the tree structure features are taken into account in the study of channel identification. Channel identification is accomplished by identifying the channel characteristics which are exhibited with one-way analysis of variance (ANOVA) scheme. Simulation results show that the channel identification is exact steady and effective.