Gridless DOA Estimator for 1.5-Bit Sparse Massive MIMO Systems Based on Covariance Matrix Estimation.

Area of Science:

• Wireless communication systems
• Signal processing

Background:

• Massive MIMO systems utilize low-bit ADCs and sparse arrays to reduce hardware costs.
• 1.5-bit quantization offers a balance between complexity and performance but introduces significant quantization loss, degrading DOA estimation.
• Existing methods struggle with accuracy in 1.5-bit sparse massive MIMO systems.

Purpose of the Study:

• To enhance the accuracy of Direction-of-Arrival (DOA) estimation in 1.5-bit sparse massive MIMO systems.
• To develop a robust covariance matrix estimation method tailored for 1.5-bit quantized sparse arrays.

Main Methods:

• A new covariance matrix estimation method is proposed, leveraging the Toeplitz property of sparse arrays.
• The method transforms the problem into a non-convex optimization, relaxed into a convex problem via semidefinite programming.
• DOAs are recovered using subspace-based methods after covariance estimation.

Main Results:

• The proposed method significantly improves DOA estimation accuracy compared to 1.5B-MUSIC and 1-bit covariance-fitting baselines.
• Performance is competitive with methods using unquantized data, particularly on coprime arrays in low-snapshot conditions.
• Demonstrates superior performance in challenging low-snapshot scenarios.

Conclusions:

• The developed covariance estimation technique effectively mitigates quantization loss in 1.5-bit sparse massive MIMO systems.
• This advancement enables more accurate DOA estimation, crucial for optimizing massive MIMO performance.
• The method provides a practical solution for cost-effective massive MIMO hardware with improved signal processing capabilities.