Wireless sensing and communications have become an indispensable part of modern life.
Among them, the key technology of direction of arrival (DOA) estimation uses array signal processing technology to measure the angular direction of radio frequency signals, and has been widely used in civil and military fields.
Traditional DOA estimation methods, such as the Multiple Signal Classification (MUSIC) algorithm, require a large number of radio frequency circuits to receive multi-channel signals, downconvert and high-speed sampling, and then perform digital signal processing.
The high complexity of hardware and algorithms, as well as the massive amount of data, greatly increase the latency, power consumption and cost of traditional signal processing systems.
Therefore, there is an urgent need for the market to develop new computing paradigms that can replace electronic processors and process radio frequency signals more efficiently, achieving low-latency, high-performance and cost-effective DOA estimation.
As a new computing paradigm, optical computing has huge advantages in computing speed, throughput and energy efficiency, providing a breakthrough in overcoming the energy efficiency bottleneck of the von Neumann architecture.
In order to directly process radio frequency signals, diffractive neural networks have been built for large-scale space optical calculations.
These networks modulate electromagnetic waves at the speed of light and process the information they carry, enabling tasks such as object identification and wireless encoding/decoding.
However, the angular resolution of existing diffractive neural networks is still limited by the diffraction limit, and its application in advanced wireless sensing missions has yet to be explored.
In addition, using reconfigurable smart surfaces (RIS) to modulate spatial electromagnetic waves to build a next generation communication system still lacks angle sensing and computing capabilities.
According to foreign media reports, in a paper published in the journal “Light: Science & Applications”, a team of scientists led by Professor Lin Xing of the Department of Electronic Engineering of Tsinghua University and his colleagues developed a super-resolution diffractive neural network (S-DNN) for all-optical DOA estimation over a wide frequency range achieves angular resolution that exceeds the Rayleigh diffraction limit.
Photo source: The journal Light: Science & Applications.
By directly processing space electromagnetic (EM) waves, S-DNN can perform DOA estimation at the speed of light without the need for traditional RF circuits, ADC or digital signal processing.
, In addition, compared to the MUSIC algorithm, S-DNN achieves higher angular resolution and a more robust estimate of input noise, and only requires one snapshot.
The researchers also applied S-DNN’s DOA estimation function to provide RIS with user angle information, thereby achieving integrated sensing and communication with low latency and low power consumption.
, Return to the first electric network home page>,.
