Single-Channel DoA Estimation Based on Nonuniform Time-Modulated Array With Asynchronous Sampling
Direction-of-arrival (DoA) estimation is a technique to detect targets based on radio frequency (RF) signals or sound signals. It is essential for intelligent perception and precise navigation and can realize accurate target positioning, which has crucial applications in radar detection, biomedicine, aerospace technology, and other fields. However, the conventional methods require multiple RF channels to collect signals. Due to high hardware complexity, they are unsuitable for Internet of Things (IoT) devices. Meanwhile, existing single-channel methods do not fully consider accuracy and practicality.
This paper proposes a novel single-channel DoA estimation method based on a nonuniform time-modulated array (NTMA) to reduce system complexity while maintaining high performance to mitigate these issues. Considering that the conventional DoA estimation methods require the collection of multiple-channel data from multiple receiving channels, the core of this method is to isolate different antennas’ signals in the frequency spectrum so that the multiple-channel data can be extracted and recovered from a single channel. Being equivalent to frequency division multiplexing (FDM), it makes full use of spectrum resources to reduce hardware complexity.
The single-channel NTMA is connected with a modulation unit behind each antenna, and they are controlled by a field-programmable gate array (FPGA) to modulate the received signals periodically. Notably, the modulation frequencies between antennas are different and coprime, which ensures the signals received by different antennas are isolated in the frequency spectrum, enabling the recovery of the multiple-channel data from the single channel.
The researchers also try to improve the estimation accuracy and signal-to-noise ratio (SNR) by designing a 1-bit phase shifter, which functions as the modulation unit. The analysis results improved the signal’s SNR by about 6 dB and focused the incident power on the useful spectrum components. This enhancement improves the estimation distance and accuracy.
The proposed method is a new architecture of hardware and software collaboration. To verify its feasibility, numerical simulation and experimental verification are performed. The measured results show that this method has a mean error below 0.27° and a maximum absolute error of 0.8° when sources incident from 60°, which is highly accurate and available.
The ternary coprime NTMA is used to evaluate this method, which is not only to solve the angular ambiguity but to avoid synchronous sampling. This method enables the reduction of the sampling frequency and improves its practicality by comparing it with other relative methods. Considering synchronous sampling, the proposed method can be applied to arbitrary array geometries, such as uniform linear arrays and uniform rectangular arrays. By taking a 10×10 uniform rectangular array as an example, it can be expected to reduce the number of receiving channels from 100 to 1.
This method offers several advantages, including reduced hardware costs, lower power consumption, and simplified system integration. It is widely applicable across fields like audio processing, radar detection, and more. Additionally, it is cost-effective to deploy, making it ideal for projects with budget constraints or those requiring large-scale implementation.