Drive-Test-Based LTE Handover Dataset for Cellular Mobility Studies in Urban Bangladesh

This dataset, titled "Drive-Test-Based LTE Handover Dataset for Cellular Mobility Studies in Urban Bangladesh," captures real-world Long Term Evolution (LTE) mobility and handover behavior in a densely deployed urban environment. Collected through structured drive tests in Dhaka, Bangladesh, the dataset includes timestamped measurements of key radio parameters such as Reference Signal Received Power (RSRP), Reference Signal Received Quality (RSRQ), Carrier-to-Interference-plus-Noise Ratio (CINR), UE velocity, and Cell IDs. It offers both raw and preprocessed formats, making it valuable for research in handover optimization, mobility prediction, and radio network performance evaluation.
The data were collected over three separate drive test sessions using XCAL-M software, Samsung Galaxy S10 (Exynos), and a GPS-enabled Arduino Uno. Measurements were taken along a 13 km urban route between Le Meridien Dhaka and BRAC University, capturing diverse mobility conditions during peak evening hours. Each dataset includes Layer 3 signaling events, handover triggers, and detailed network statistics.
The raw dataset contains unfiltered CSV logs with varying column structures and lengths across different sessions. The processed dataset has been cleaned, interpolated, and aligned for machine learning readiness. A Time-to-Trigger (TTT) logic of 320 ms was applied to refine the handover labeling, enabling precise model training. Diagnostic files in .txt and .drm formats are also included for users who wish to reconstruct the original test environment.

Key features include:
Real LTE network data from a high-density urban area in Bangladesh
Handover-labeled samples with timestamped signal quality and velocity data
Structured folder organization for reproducibility: raw CSVs, processed ML-ready data, measurement reports, and binary logs
Useful for handover prediction, mobility-aware optimization, and quality-of-service research

This is one of the first publicly available LTE mobility datasets from Bangladesh, addressing the lack of open regional cellular datasets and supporting further advancements in wireless communication and machine learning for mobility management.