Research On Efficient Sampling System Based On Multiband Signals

In the space borne electronic reconnaissance, there are missions include intercepting and receiving signals of multi carrier frequencies. Modern radio frequency technologies enable modulation of narrowband signals by high carrier frequencies, therefore multiband signals consist of several relatively narrowband transmissions spread a wide spectrum range. For operation of converting such analog signal to digital form, the sampling rate may be exceedingly high under Nyquist sampling theory, which leads to unbearable high power dissipation. Based on limitations of power dissipation and digital resource, this paper focuses on the digitalization of multiband signal, and research works are summarized as below:Firstly, the efficient multiband signal sampling and reconstruction schemes are investigated. Comparing with the existing sampling methods, periodically non-uniform sampling (PNS) is choosen as efficient sampling method in this thesis. By analyzing the optimal average sampling rate, the way of getting optimal system sampling rate is derived. Reconstruction of signal from PNS can be achieved without distortion under the assumption that the signal is perfectly band-limited. However, most signals in practice could not meet such requirement, thus the sampling parameters play an important role on the quality of reconstructed signal. An optimization method based on minimization of the2-norm of reconstruction filter coefficients is proposed, which can diminish the impact of noise on signal reconstruction. Numerical simulations verify the effectiveness of the proposed method and relative analysis.Secondly, multiband signal spectrum sensing and monitoring are studied. Since PNS requires the information of spectrum structure of sampled signal, frequency sweep spectrum sensing scheme based on energy detector is proposed, which fulfills the requirement of low power dissipation and low complexity of spectrum sensing implemetation. Then to overcome the problem of time-varying of spectrum structure, a spectrum monitoring scheme and an associated optimal method are presented. By constructing monitoring vector, spectrum monitoring can by achieved at the PNS sampling rate. Simulation results exhibit the superiority of the proposed spectrum structure sensing and monitoring scheme. Thirdly, implementation of efficient sampling system is examined. The efficient sampling system based on multiband signals is designed and implemented. Related tests and analyses are performed. Test results show the achieved system has met the application requirement, and confirms the validation of related researches as well.