UplinkNet: Practical Commercial 5G Standalone (SA) Uplink Throughput Prediction

TL;DR

Problem: uplink throughput prediction is critical for QoE but challenging in real-world 5G SA deployments where UL gains are limited. Approach: UplinkNet, a ConvLSTM-based, compact neural network that predictorUL throughput using only Android API–accessible RF parameters, trained on 40 real-world drive-test traces from Tokyo and Bangkok and evaluated on unseen data. Key findings: achieves an average RMSE of 5.22 Mbps and 98.9% accuracy for total data transferred, outperforming larger baselines even at markedly reduced parameter counts (as low as ~1k–2k parameters). Significance: enables practical, on-device uplink throughput prediction for UL-intensive applications and IoT on 5G SA networks, with potential extensions to uplink MIMO, carrier aggregation, and mmWave deployments.

Abstract

While 5G New Radio (NR) networks offer significant uplink throughput improvements, these gains are primarily realized when User Equipment (UE) connects to high-frequency millimeter wave (mmWave) bands. The growing demand for uplink-intensive applications, such as real-time UHD 4K/8K video streaming and Virtual Reality (VR)/Augmented Reality (AR) content, highlights the need for accurate uplink throughput prediction to optimize user Quality of Experience (QoE). In this paper, we introduce UplinkNet, a compact neural network designed to predict future uplink throughput using past throughput and RF parameters available through the Android API. With a model size limited to approximately 4,000 parameters, UplinkNet is suitable for IoT and low-power devices. The network was trained on real-world drive test data from commercial 5G Standalone (SA) networks in Tokyo, Japan, and Bangkok, Thailand, across various mobility conditions. To ensure practical implementation, the model uses only Android API data and was evaluated on unseen data against other models. Results show that UplinkNet achieves an average prediction accuracy of 98.9% and an RMSE of 5.22 Mbps, outperforming all other models while maintaining a compact size and low computational cost.

Figures (7)

• Figure 1: Prediction Accuracy on Evaluation Data
• Figure 2: Summary of Models Architecture
• Figure 3: 5G Uplink Throughput characteristics in some of the test scenarios.
• Figure 4: Overview of UplinkNet
• Figure 5: Predicted uplink throughput from our model compared to others when predicting Driving (Urban Bangkok).