Compressive sampling based radar receiver
暂无分享,去创建一个
A typical radar system transmits a wideband pulse (linear chirp, coded pulse, pseudonoise-PN sequence etc.) and then correlates the received signal with that same pulse in a matched filter (effecting pulse compression). A traditional radar receiver consists of either an analog pulse compression system followed by a high-rate analog-to-digital converter or a high-rate A/D converter followed by pulse compression in a digital computer; both approaches are complicated and expensive. Achieving adequate A/D conversion of a wideband PN/chirp radar signal (which is compressed into a short duration pulse by the matched filter) requires both a high sampling frequency and a large dynamic range. Currently available A/D conversion technology is a limiting factor in the design of ultra wideband (high resolution) radar systems, because in many cases the required performance is either beyond what is technologically possible or too expensive. For these reasons, HW and SW solutions have been adopted in order to employ more effectively today's technology taking advantage of certain physical characteristics of the signals, such as the sparsity. In this paper, we introduce a new kind of radar receiver based on the Compressed Sampling theory - CS, through which is possible acquire and process signals, breaking the constraint of sampling imposed by Nyquist/Shannon criterion. In CS, an incoherent linear projection is used to acquire an efficient representation of a compressible signal directly using just a few measurements. The signal is then reconstructed by solving an inverse problem either through a linear program or a greedy pursuit. In that regard, the literature provides several alternatives, such as: Random Demodulation, Periodic Non-Uniform Sampling-PNS; among the more recent and successful stands the so-called Modulated Wideband Converter, MWC. The MWC extends conventional I/Q demodulation to multiband inputs with unknown carriers, and as such it also provides a scalable solution that decouples undesired RF-ADC dependencies. The MWC combines the advantages of RF demodulation and the blind recovery ideas, and allows sampling and reconstruction without requiring knowledge of the band locations. To bypass analog bandwidth issues in the ADCs, an RF frontend mixes the input with periodic waveforms. This operation imitates the effect of delayed undersarnplinq, specifically folding the spectrum to baseband with different weights for each frequency interval. In contrast to undersampling (or PNS) aliasing is realized by RF components rather than by taking advantage of the ADC circuitry. In this way, bandwidth requirements are shifted from ADC devices to RF mixing circuitries. The key idea is that periodic mixing serves another goalboth the sampling and reconstruction stages do not require knowledge of the carrier positions. In scenarios in which the carrier frequencies are known to the receiver MWC could play a crucial role to simplify the analog receiving chain of system radar, since it allows to eliminate the local oscillator STALO and the analog mixer from the chain, ensuring the possibility of simultaneously acquiring multiband siqnals, current radar doesn't support this operation because it needs to re-tune STALO oscillation frequency for new acquisition. Tuning speed measures the length of time required for the Local Oscillator to change from one center frequency to another within a specified accuracy level. In typical systems, when tuning from one frequency to another, the LO usually slightly overshoots the desired frequency and then settles to the desired frequency within a certain time period, usually in the order of tens μs for a VCO-based LO; when using a YlG-based LO, it needs tens of ms. In case of spectrum blind acquisitions, this technique allows reconstruction of the received signal by sampling at aminimum rate compared to the traditional rules of Nyquist/Shannon, without loss of accuracy; this meansreduction of the analog-digital converter cost and manage a smaller number of data, thanks to a more sophisticated processing.