Research On Non-cooperated OFDM Reception

Orthogonal Frequency Division Multiplexing (OFDM) is widely used, due to it's advantage of high data rates in multipath fading environments. Recently, researching on OFDM cooperative communication has been a hotspot. However, researching on OFDM non-cooperative reception is on the scratch line. For national information security, it is important to do it. In the dissertation, the key technologies in OFDM non-cooperative reception are studied. The primary works and innovations are as follows:(1) Propose OFDM non-cooperative reception mathematical model and receiver block diagram, based on the analysis of the particularity of OFDM non-cooperative reception. It is pointed out that there are three key problems should be solved, i.e., modulation recognition, parameters estimation and synchronization.(2) Propose blind modulation classification algorithm for OFDM systems. The algorithm is divided into two parts, i.e., recognition in passband and classification in baseband. In passband recognition part, time series normal test is used to distinguish the multicarrier signals from the single carrier signals; in baseband part, a novel blind modulation classification algorithm is derived from the finite alphabet property of information symbols and the equivalent parallel model of OFDM systems. The performances of the proposed algorithm are tested with Monte-Carlo simulations in AWGN channel and multipath channel respectively. It is found that the proposed algorithm works well in low signal-to-noise ratio (SNR) without parameter information.(3) It is proved that OFDM signal is cyclicstationary unconditionally based on the model of pulse shaping after IDFT. It is different from the conclusion of paper [34] . Based on the model of pulse shaping in parallel subchannels, paper [34] considered that OFDM signal is cyclicstationary in the case of that signal has time-frequency guard regions, pulse shaping (not a rectangular pulse), and subcarrier weighting. In fact, the model in this paper is accepted by many practicality standards. Spectral correlation density function (SCDF) of passband OFDM signal is deroved in the dissertation. In the derivation procedure, this paper also proposes a simple method to derive SCDF. With the method, the SCDFs of MPSK signal are identical with results in classical papers. Simulation results support the theoretical analysis nicely.(4) Based on the second order cyclicstationary, this paper proposes two algorithms to estimate chip width and carrier frequency of OFDM signal with no parameter information needed. One is joint estimation algorithm of carrier frequency and chip width based on SCDF, the other is chip width estimation algorithm based on cyclic correlation. The two algorithms suit for low SNR signals, because they work with nonzero cyclic frequencies on which noise or interference has little influence. Cramé-Rao lower bound (CRLB) of joint estimation is derived. Computer simulation results show that the proposed algorithms have good performance in AWGN channel and multipath channel when the input signal-to-noise ratio is OdB or even worse. So the proposed algorithms will work well for signal reconnaissance.(5) Implement the proposed blind chip width estimation algorithm with DSP processor. Second order cyclicstationary is a useful tool for blind signal processing, but it is difficult to implement in hardware for heavy computation. Now, most algorithms based on second order cyclicstationary are in the phase of computer simulation. This paper makes a positive attempt to implement the proposed blind chip width estimation algorithm with TI's DSP processor TMS320C6416.