The Inplementation Of Software Defined Radio And Study Of Related Techniques

In this paper, fundamentals of Software-Defined Radio, OFDM transmission technologies and signal classification and SNR estimation techniques are studied. Firstly, the basic principles of Software-Defined Radio are introduced, then a universal Software-Defined Radio hardware platform"DSP+FPGA+ASIC"is introduced. Basic transmission technologies includingπ/4-QPSK, GMSK, QAM4 and OFDM are realized on the platform. Secondly, modulation identification techniques are studied. A joint signal classification and SNR estimation algorithm based on kurtosis is proposed. The proposed algorithm could perform directly in Intermediate Frequency. Moreover, it is robust to Doppler frequency shift. Simulation results show that the proposed kurtosis based algorithm could obtain a probability of detection of OFDM signal up to 100% even in very low SNR (i.e. SNR = 0dB). Finally, SNR estimation techniques are studied. A SNR estimation algorithm (K2) based on 2-order kurtosis is proposed. The proposed K2 algorithm is applicable to frequency selective multipath channels as well as AWGN channels. The ideal SNR estimation MSE value of K2 is presented. The CRB of SNR estimation algorithms in frequency selective multipath channels is proposed. Furthermore, On the basis of K2, a SNR estimation algorithm (KP) using P-order kurtosis is proposed. The proposed KP algorithm is more convenient and easier to implement than the moment based algorithms. The MSE performance of the KP algorithm is analyzed.