Reconstruction of Frequency Hopping Communication Signals via Analog to Information Converter

Frequency Hopping (FH) signal is widely used in military filed especially in military communication for its good anti-jamming performance and multiple access capability. With the increase of FH signal's band width, it makes a huge pressure on traditional Analog to Digital Converter. Theory of compressed sensing (CS) enables a successful reconstruction of sparse signal by sampling at a sub-Nyquist sample rate. In this paper, we apply CS into frequency hopping communication signal's reconstruction and explore the available improvement. We adopt the Random Demodulator scheme, which is a popular architecture in Analog to Information Converter (AIC), to obtain the digital sequence and realize a precise reconstruction with high possibility at a sub-Nyquist sampling rate. Then some improvements are made on the construction of observation matrix to enhance the adaptation in sampling. Besides, K-SVD algorithm is deployed to train an overcomplete dictionary which is used as sparse matrix. Simulations are performed to verify the reconstruction performance of this scheme.