Research On Beamforming Algorithm Of Broadband Non-Stationary Signals And OFDM System

Smart antenna technology has been concerned widely as a key technology in the 3G and B3G/4G mobile communications, which selects appropriate adaptive algorithms to form space directional beam dynamically by means of modern digital signal processing techniques, taking full advantage of mobile users signal to counteract or suppress the interference to the greatest extent. At the same time, by introducing space division multiple access and combing with other multiple accesses, smart antenna technology not only suppresses the co- channel interference effectively, and decreases bit error rate significantly, but also increases spectral efficiency and system capacity.This thesis mainly focuses on the core technology of smart antenna, the beamforming algorithm. The main work involves:1. The technologies of smart antenna and beamforming are introduced briefly; some common rules and algorithms are analyzed and compared, then the algorithm simulations and performance analysis are given.2. Aimed at the broadband and non-stationary character of modern communication signals, the performance of existing broadband and non-stationary beamforming algorithms are analyzed, using the methods based on time domain, frequency domain and the general transforming domain, and the problems of each method are summarized. On the basic of above, a typical signal mode—broadband Liner Frequency Modulation (LFM) signal model is established, and the beamforming algorithm of broadband LFM signal based on the fractional Fourier transform is proposed. This algorithm makes full use of the highly compact property of LFM on the fractional Fourier transform domain, and achieves beamforming on this domain. The simulation results verified the effectiveness of the algorithm.3. Considering the practical application of smart antenna, this thesis integrates smart antenna with the Orthogonal Frequency Division Multiple (OFDM) to achieve beamfomring. First of all, in the view of the current beamforming algorithm applied in OFDM system—blind algorithm and non-blind algorithm, some existing problems are analyzed. On the basic of those, a new bemforming algorithm for OFDM system is proposed. The algorithm projects the observations onto the over-complete dictionaries, which were built out of the preamble of OFDM frame, thereby obtains the estimation of DOA and steering vector of interested signal. Then through the robust beamforming algorithm, performance of the beamformer which has the directional error is improved. The algorithm does not need the complex channel estimation technology, neither affected by small number of pilot channels. Compared with the blind algorithm based on training symbols, the new algorithm has lower bit error rate.