# A Decentralized Fusion Scheme for 5G Multi-BS Positioning

**Authors:** Sharief Saleh, Qamar Bader, Mohamed Elhabiby, Aboelmagd Noureldin  
**arXiv:** [2303.04192](https://arxiv.org/abs/2303.04192)  
**Date:** 2023-03-07  
**Summary page:** https://www.summarizepaper.com/en/paper/2303.04192v1/

## Key points

- The demand for precise positioning and navigation solutions has increased due to the growth of highly-intelligent IoT applications, context-aware services, and location-based services (LBS).
- Safety-critical applications such as self-driving cars require higher levels of autonomy that compel centimeter-level absolute and relative positioning.
- 5G networks are expected to provide high precision positioning estimation utilizing mmWave signals in urban and downtown areas.
- In such areas, 5G base stations (BSs) will be densely deployed, allowing for line-of-sight communications between the user equipment (UE) and multiple BSs at the same time.
- An unscented Kalman filter (UKF) is commonly believed to eradicate linearization errors but fails to do so completely.
- A de-centralized integration between 5G BSs would fully avoid linearization errors and enhance the positioning performance significantly by fusing position measurements instead of directly fusing range and angle measurements.
- The proposed de-centralized KF method was evaluated in a quasi-real simulation setup provided by Siradel using a real trajectory in Downtown Toronto.
- The experiments compared the performance of de-centralized KF integration to that of centralized EKF and UKF integration schemes.
- It was shown that the proposed method outperformed both UKF and EKF implementations in multiple scenarios as it decreased RMS and maximum 2D positioning errors significantly, achieving decimeter-level accuracy for 90.3% of the time.
- The paper also provides an overview of previous works related to 5G positioning using optimal estimators and establishes the system model along with basics on 5G positioning.
- Various implementations of Kalman filters are exhibited while thoroughly analyzing centralized fusion schemes before proposing 5G decentralized integration as a strong contender.
- Overall, this paper presents a novel approach that avoids disadvantages associated with centralized integration schemes while providing better performance than traditional methods used for multi-BS fusion in 5G networks.

## Summary

The demand for precise positioning and navigation solutions has increased in recent years due to the growth of highly-intelligent IoT applications, context-aware services, and location-based services (LBS). Additionally, safety-critical applications such as self-driving cars require higher levels of autonomy that compel centimeter-level absolute and relative positioning. Fifth-generation (5G) networks are expected to provide high precision positioning estimation utilizing mmWave signals in urban and downtown areas. In such areas, 5G base stations (BSs) will be densely deployed, allowing for line-of-sight communications between the user equipment (UE) and multiple BSs at the same time. Traditionally, 5G multi-BS fusion is conducted via an extended Kalman filter (EKF), which directly utilizes range and angle measurements in a centralized integration scheme. However, such measurements have a non-linear relationship with the positioning states of the filter, giving rise to linearization errors. An unscented Kalman filter (UKF) is commonly believed to eradicate these errors but fails to do so completely. In this paper titled "A Decentralized Fusion Scheme for 5G Multi-BS Positioning," the authors argue that a de-centralized integration between 5G BSs would fully avoid linearization errors and enhance the positioning performance significantly. This is achieved by fusing position measurements instead of directly fusing range and angle measurements, which inherently leads to a linear measurement model by design. The proposed de-centralized KF method was evaluated in a quasi-real simulation setup provided by Siradel using a real trajectory in Downtown Toronto. The experiments compared the performance of de-centralized KF integration to that of centralized EKF and UKF integration schemes. It was shown that the proposed method outperformed both UKF and EKF implementations in multiple scenarios as it decreased RMS and maximum 2D positioning errors significantly, achieving decimeter-level accuracy for 90.3% of the time. The paper also provides an overview of previous works related to 5G positioning using optimal estimators and establishes the system model along with basics on 5G positioning. Various implementations of Kalman filters are exhibited while thoroughly analyzing centralized fusion schemes before proposing 5G decentralized integration as a strong contender. Overall, this paper presents a novel approach that avoids disadvantages associated with centralized integration schemes while providing better performance than traditional methods used for multi-BS fusion in 5G networks.