[Paper Review] CaSCADE: Compressed Carrier and DOA Estimation

Shahar Stein, Or Yair+2 more|arXiv (Cornell University)|Apr 10, 2016

Sparse and Compressive Sensing Techniques31 references3 citations

TL;DR

CaSCADE proposes a sub-Nyquist sampling and reconstruction system for joint carrier frequency and direction-of-arrival (DOA) estimation using an L-shaped array of uniform linear arrays (ULAs), enabling efficient spectrum sensing in wideband multiband signals. It outperforms the MWC in recovery accuracy and complexity while resolving the pairing issue between frequencies and DOAs via ESPRIT and compressed sensing.

ABSTRACT

Spectrum sensing and direction of arrival (DOA) estimation have been thoroughly investigated, both separately and as a joint task. Estimating the support of a set of signals and their DOAs is crucial to many signal processing applications, such as Cognitive Radio (CR). A challenging scenario, faced by CRs, is that of multiband signals, composed of several narrowband transmissions spread over a wide spectrum each with unknown carrier frequencies and DOAs. The Nyquist rate of such signals is high and constitutes a bottleneck both in the analog and digital domains. To alleviate the sampling rate issue, several sub-Nyquist sampling methods, such as multicoset sampling or the modulated wideband converter (MWC), have been proposed in the context of spectrum sensing. In this work, we first suggest an alternative sub-Nyquist sampling and signal reconstruction method to the MWC, based on a uniform linear array (ULA). We then extend our approach to joint spectrum sensing and DOA estimation and propose the CompreSsed CArrier and DOA Estimation (CaSCADE) system, composed of an L-shaped array with two ULAs. In both cases, we derive perfect recovery conditions of the signal parameters (carrier frequencies and DOAs if relevant) and the signal itself and provide two reconstruction algorithms, one based on the ESPRIT method and the second on compressed sensing techniques. Both our joint carriers and DOAs recovery algorithms overcome the well-known pairing issue between the two parameters. Simulations demonstrate that our alternative spectrum sensing system outperforms the MWC in terms of recovery error and design complexity and show joint carrier frequencies and DOAs from our CaSCADE system's sub-Nyquist samples.

Motivation & Objective

• Address the high sampling rate bottleneck in wideband multiband signal sensing, especially for cognitive radio (CR) applications.
• Overcome the limitations of existing sub-Nyquist methods like the MWC, which do not support DOA estimation.
• Enable joint estimation of unknown carrier frequencies and DOAs from compressed sub-Nyquist samples without requiring Nyquist-rate sampling.
• Resolve the pairing problem between carrier frequencies and DOAs in multiband signals using novel reconstruction algorithms.
• Design a practical system with reduced hardware complexity and improved robustness in low SNR environments.

Proposed method

• Proposes a ULA-based sub-Nyquist sampling front-end equivalent to one MWC channel per sensor, enabling compressed sensing of wideband signals.
• Extends the ULA design to an L-shaped array composed of two orthogonal ULAs sharing a common sensor at the origin to enable 2D DOA estimation.
• Derives sufficient conditions for perfect recovery of carrier frequencies and DOAs based on the spark of the sensing matrix and signal sparsity.
• Develops two reconstruction algorithms: one based on ESPRIT for analytical parameter estimation and another using compressed sensing (e.g., OMP) for sparse vector recovery.
• Uses trilinear decomposition via PARAFAC to estimate signal parameters from cross-correlation matrices in the joint estimation framework.
• Applies a bilinear model to relate sub-Nyquist samples to the original signal parameters, enabling joint frequency and DOA recovery.

Experimental results

Research questions

• RQ1 Can a sub-Nyquist sampling system based on a ULA achieve better performance than the MWC in carrier frequency estimation for multiband signals?
• RQ2 How can joint carrier frequency and DOA estimation be achieved from sub-Nyquist samples without the pairing ambiguity between parameters?
• RQ3 What is the minimal number of sensors required for reliable recovery of M transmissions in both spectrum sensing and joint DOA estimation scenarios?
• RQ4 How does the proposed CaSCADE system perform in low SNR conditions compared to existing methods like MWC and PARAFAC-based approaches?
• RQ5 Can the combination of ESPRIT and compressed sensing techniques provide robust and accurate recovery of signal parameters from compressed measurements?

Key findings

• The CaSCADE system achieves lower recovery error than the MWC in both carrier frequency and DOA estimation, particularly in low SNR regimes.
• The minimal number of sensors required for perfect recovery is 2M for spectrum sensing and 2M+1 for joint estimation, with M+1 sensors sufficient with high probability.
• The ESPRIT-based algorithm outperforms the compressed sensing approach in terms of accuracy and robustness, especially at low SNR.
• The PARAFAC-based method provides a viable alternative but requires higher computational load and is less accurate than the ESPRIT and CS-based approaches.
• The system maintains stable performance across varying numbers of sensors and SNR levels, with reconstruction error decreasing as sensor count increases.
• The proposed method successfully resolves the pairing issue between carrier frequencies and DOAs, avoiding iterative joint diagonalization required by traditional JAFE methods.