Cellular network wireless positioning method combining pso and ss-elm

Abstract

The invention discloses a cellular network wireless positioning method combining PSO and SS-ELM. The realization of the present invention includes using labeled training data and unlabeled training data to train the output layer weight parameter β of SS-ELM, and introducing PSO in the process of training SS-ELM to automatically perform hyperparameters of SS-ELM Optimization, PSO's fitness value calculation function uses labeled training data and unlabeled training data to optimize and screen SS‑ELM during training to obtain the optimal SS‑ELM parameters as a regression model for online positioning services. The implementation of the patent of the present invention is implemented in two parts: line separation and online. The invention reduces the dependence of cellular network positioning based on RSS fingerprint data on labeled RSS fingerprint data, reduces the cost of manual data collection, and reduces the workload of manually adjusting parameters in the algorithm training process.

Description

technical field [0001] The invention belongs to a cellular network wireless positioning method in the field of pattern recognition and computing intelligence, and specifically relates to particle swarm optimization (Particle Swarm Optimization, PSO) and semi-supervised extreme learning machine (Semi-Supervised Extreme Learning Machine, SS-ELM). Background technique [0002] The highly developed cellular network system and cellular network signals covering the world make the cellular network system the most widely used mobile communication system, and the popularity of smart phones makes the positioning technology based on the cellular network system an important outdoor positioning technology. Especially when the satellite positioning system, such as the global positioning system (GPS), is not available, the smart phone can only rely on the cellular network system for outdoor positioning. At the same time, with the development of the Internet of Things technology, more and m...

AI Technical Summary

Problems solved by technology

But SS-ELM is right Norm and manifold regular constraints are very sensitive. In practical applications, there is no systematic theory to guide us to optimize the hyperparameters of the constraints. The training process of SS-ELM itself only trains and optimizes the output layer weight β, and there is no System theory guides the optimization of SS-ELM hyperparameters, so the optimization of SS-ELM hyperparameters can only be repeated experiments by experienced staff according to specific business scenarios

Images