A new signal processing in antenna arrays

Background

Phased array antennas are highly desirable for many wireless applications, but their cost, complexity, and power consumption have limited adoption. RF beamforming needs a phase shifter behind every antenna element, which adds cost, loss, and design challenges. Digital beamforming is highly flexible but needs an ADC for every element, which makes it expensive and power-hungry. Improvements such as better algorithms, more efficient amplifiers, or lower-resolution ADCs have brought only incremental gains, often at the expense of signal-to-noise ratio and dynamic range. Architectures that use fewer phase shifters tend to suffer from limited scan range, high loss, or added complexity.
This invention presents a signal processing architecture for phased arrays that eliminates RF phase shifters and reduces the ADC count to one.

Description of the invention

All antenna signals are combined without phase shifting, down converted, and digitized by this single ADC. A digital signal processing unit, implemented using a trained neural network, then reconstructs the individual antenna element signals from the combined digital stream, enabling flexible digital beamforming. This approach significantly lowers hardware cost, complexity, and power consumption compared to traditional methods. To ensure consistent performance across a wide field of view, a custom antenna array with a flat radiation pattern has been designed and integrated.

Advantages

This technology makes phased arrays simpler, cheaper, and more power-efficient. It combines all antenna signals, digitizes them with a single ADC, and steers beams through digital processing. With no phase shifters and only one ADC, it gives flexible beamforming at low cost and power for communications, radar, and beyond. 

Potential applications

• Satcom terminals; Stationary or moving stations
• Radar and Advanced Sensing Systems
• High-Resolution Imaging and Security Screening
• Advanced Audio Systems