DeepCell: A Ubiquitous Accurate Provider-side Cellular-based Localization

TL;DR

DeepCell tackles the problem of ubiquitous outdoor localization on low-end phones by shifting fingerprinting to the cellular provider side. It constructs a rich, label-enhanced fingerprint from unlabeled provider data through GPS-aligned synchronization, Gaussian Process-based spatial augmentation, and a virtual-grid training scheme, then trains a deep multinomial classifier to map RSS signatures to grid cells and refine positions. The approach achieves a median localization error of $29\ \mathrm{m}$ in a realistic urban testbed, outperforming the state-of-the-art DeepLoc by more than $75.4\%$ and extending to low-end devices without extra energy cost. This provider-centered framework enables scalable, robust localization across networks and devices, offering practical ubiquity for location-based services.

Abstract

Although outdoor localization is already available to the general public and businesses through the wide spread use of the GPS, it is not supported by low-end phones, requires a direct line of sight to satellites and can drain phone battery quickly. The current fingerprinting solutions can provide high-accuracy localization but are based on the client side. This limits their ubiquitous deployment and accuracy. In this paper, we introduce DeepCell: a provider-side fingerprinting localization system that can provide high accuracy localization for any cell phone. To build its fingerprint, DeepCell leverages the unlabeled cellular measurements recorded by the cellular provider while opportunistically synchronizing with selected client devices to get location labels. The fingerprint is then used to train a deep neural network model that is harnessed for localization. To achieve this goal, DeepCell need to address a number of challenges including using unlabeled data from the provider side, handling noise and sparsity, scaling the data to large areas, and finally providing enough data that is required for training deep models without overhead. Evaluation of DeepCell in a typical realistic environment shows that it can achieve a consistent median accuracy of 29m. This accuracy outperforms the state-of-the-art client-based cellular-based systems by more than 75.4%. In addition, the same accuracy is extended to low-end phones.

Figures (11)

• Figure 1: DeepCell system architecture.
• Figure 2: The gridding approach. The area of interest is superimposed by equally-sized square cells. Red points present the training samples.
• Figure 3: Network structure. The input is the RSS coming from different cell towers in the environment with a bit corresponds to each cell tower indicates wheather or not the cell-tower is in the active cells (features). The output is the probability distribution for different grid cells in the area of interest. Grey-shaded neurons represent examples of nodes that have been temporary dropped-off to increase the model robustness and avoid over-training.
• Figure 4: Our testbed.
• Figure 5: Effect of different fingerprint models on the localization accuracy.