Semi-supervised learning positioning method for distributed large-scale multi-antenna system
A technology of semi-supervised learning and multi-antenna system, which is applied in the field of semi-supervised learning and positioning of distributed large-scale multi-antenna systems, and the field of expectation maximization algorithm can solve the problems of low positioning accuracy and high sampling cost, so as to reduce sampling cost and ensure The effect of uniqueness and high positioning accuracy
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0035] like figure 1 As shown, a semi-supervised learning localization method for distributed large-scale multi-antenna systems includes the following steps:
[0036] S1: Generate two different data sets, including training data set and coordinate membership degree set;
[0037] S2: Use the semi-supervised learning algorithm to estimate the GMM parameters of the Gaussian mixture model, and initialize the GMM according to the training data set;
[0038] S3: Iteratively estimate the GMM parameters based on the semi-supervised expected maximum EM algorithm, and complete the training of the GMM;
[0039] S4: According to the trained GMM and the coordinate membership degree set, complete the estimation of the location information corresponding to the target received signal strength RSS data.
[0040] In the specific implementation process, such as figure 2 As shown, first of all, in order to reduce the sampling cost and improve the practical usability of the algorithm, the presen...
Embodiment 2
[0117] More specifically, on the basis of Example 1, such as Figure 6 As shown, taking the RRH number M=30 as an example, two schematic diagrams of different antenna distributions are given. The present invention shows that under different antenna distribution situations, the percentage of RRH number (M), marked data to total training data (p L ), the number of classifications (L), the signal-to-noise ratio (Signal-to-Noise Ratio, SNR), the distance between the user and the RRH, etc. affect the positioning performance. Several traditional positioning schemes are compared, including:
[0118] Supervised learning positioning schemes, such as KNN [5], Multi-Layer Perception Regression (MLPR) [20], Bayesian Ridge Regression (BRR) [21], Gradient Boosting Regression (Gradient Boosting Regression) , GBR)[22], linear regression (Linear Regression, LR)[23], etc.;
[0119] Semi-supervised learning positioning schemes, such as semi-supervised K-Means (Semi-supervised K-Means, S-K-Mea...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More