Array super-resolution direction-of-arrival estimation method

Abstract

The invention discloses an array super-resolution direction-of-arrival estimation method, which comprises the following steps of: based on information in a received signal autocorrelation matrix, determining a 1-degree interval where a signal angle is located through a first part by adopting a two-section deep neural network structure, and specifically estimating the signal angle on a grid with higher resolution through a second part. The two parts are combined to realize accurate estimation of the incident angle of the signal, including off-network signals in general meaning, and the resolution of 0.01 degree can be achieved. According to the method, a two-section network architecture is innovatively used, the problems of excessive neural network parameters and overlong training time are effectively avoided while ultrahigh-resolution direction-of-arrival estimation is realized, and the method is simple in calculation and quick in response and meets the requirements of practical application.

Description

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine

Login to View More

AI Technical Summary

Benefits of technology

This technology uses traditional deep learning techniques for estimating distance between points or lines within an image frame. It involves dividing each pixel into smaller sections called segments - one section at 0° (angle) and another section at 30°(fraction). These segmented areas represent different angles along their edges. By comparing these values with previous measurements taken during this process, we can determine how far they were before being captured accurately compared against previously collected measurement results. Overall, it provides technical benefits such as high accuracy, low computational complexity, efficient processing time, and easy implementation.

Problems solved by technology

This patents describes how modern techniques like machine learning help estimate unknown angles from radio frequency (RF), millimeter wave (mmWave). However, current approaches require complicated calculations involving numerous mathematical steps such as matrix multiplication, convolution operation, etc., making them difficult to use effectively without overfitting any specific numerical values into these computational tools.

Images