Signal source positioning method and system for matrix completion by adaptive noise estimation

Abstract

The invention discloses a signal source positioning method and system for matrix completion by adaptive noise estimation. The method comprises the following steps: acquiring sensor data output by a sensor in an observation area; performing rasterization on the observation area; constructing a rasterized matrix; performing matrix completion on the rasterized matrix, in the matrix completion process, adopting local polynomial fitting to estimate the signal intensity from the signal source at the central point of the grid, and estimating noise used for matrix completion according to the signal intensity from the signal source at the central point of the grid; and performing vector decomposition on the complemented rasterized matrix to obtain a fingerprint vector, and performing calculation by using the fingerprint vector to obtain a position coordinate of the signal source. Due to the fact that the local polynomial fitting method is adopted to estimate the noise in the whole matrix complementing process, more reasonable noise is calculated to be used for matrix complementing, the error of matrix complementing is reduced, and the accuracy of signal source positioning is improved.

Description

technical field [0001] The present invention relates to the technical field of wireless communication, in particular to a method and system for locating signal sources using adaptive noise estimation for matrix completion. Background technique [0002] In recent years, marine research has been paid more and more attention, and underwater communication is very important to marine research. Whether it is military or economic, underwater communication, as a field of wireless communication, plays an important role. Positioning is an important technology in underwater communication. With the deepening of marine research, the requirements for the accuracy of underwater positioning are getting higher and higher. Due to the instability of ocean currents, as well as the influence of water temperature and tides, a parameterized model cannot be used in the ocean to help locate the signal source, and a positioning method that does not depend on the parameterized model is required. In a...

AI Technical Summary

Problems solved by technology

[0003] The weighted centroid location method is a commonly used signal source location method. Although this method has the characteristics of low complexity and less hardware resource consumption, it is very dependent on the selection of parameters and is greatly affected by the transmission environment. It is limited by the variety of signals. Path propagation, non-line-of-sight and other situations, resulting in low positioning accuracy of this method

Now, the method of machine learning is very popular, but this method needs to train the model, it will cost more resources, and it needs to use a large amount of data during training, but in many cases it is impossible to collect a large amount of data

Compared with the above two methods, the matrix-based signal source location method does not depend on parameter selection, and the data demand is small, which has certain advantages, but it still brings some errors in the process of matrix completion. Therefore, there are It is necessary to take measures to reduce the error in the matrix completion process

Images