Model-based Learning Network for 3-D Localization in mmWave Communications

TL;DR

This paper introduces a hybrid model-based neural network approach for 3D localization in mmWave communications, combining geometric estimators with neural networks to improve accuracy, robustness, and efficiency over existing methods.

Contribution

It develops a novel WLS estimator using hybrid measurements and integrates neural networks to learn higher-order errors, enhancing localization performance.

Findings

Outperforms state-of-the-art localization methods in accuracy.

Achieves CRLB with the proposed estimator.

Demonstrates robustness and reduced computation time.

Abstract

This study considers the joint location and velocity estimation of UE and scatterers in a three-dimensional mmWave CRAN architecture. Several existing works have achieved satisfactory results with neural networks (NNs) for localization. However, the black box NN localization method has limited performance and relies on a prohibitive amount of training data. Thus, we propose a model-based learning network for localization by combining NNs with geometric models. Specifically, we first develop an unbiased WLS estimator by utilizing hybrid delay/angular measurements, which determine the location and velocity of the UE in only one estimator, and can obtain the location and velocity of scatterers further. The proposed estimator can achieve the CRLB and outperforms state-of-the-art methods. Second, we establish a NN-assisted localization method (NN-WLS) by replacing the linear approximations in the proposed WLS localization model with NNs to learn higher-order error components, thereby enhancing the performance of the estimator. The solution possesses the powerful learning ability of the NN and the robustness of the proposed geometric model. Moreover, the ensemble learning is applied to improve the localization accuracy further. Comprehensive simulations show that the proposed NN-WLS is superior to the benchmark methods in terms of localization accuracy, robustness, and required time resources.