Publication Date

Spring 2025

Degree Type

Master's Project

Degree Name

Master of Science in Computer Science (MSCS)

Department

Computer Science

First Advisor

Navrati Saxena

Second Advisor

Robert Chun

Third Advisor

Pooja Shyamsundar

Keywords

Handover Prediction, 5G, RNN’s, LSTM, Bi-LSTM, T-SMOTE

Abstract

As users move across network cells in 5G, it is critical to maintain seamless connectivity through efficient and fast handovers. However, as 5G networks have a very dense deployment of cells and higher carrier frequencies, handovers are often more frequent and challenging, leading to failures or the ping-pong effect. In this research, we are going to use Artificial Intelligence (AI) techniques to enable predicting handover(HO) events proactively as opposed to reactively, aiming to reduce HO failures and unnecessary handovers. We develop Long-Short-Term Memory (LSTM) and Bidirectional LSTM (Bi-LSTM) models to forecast future signal measurements, predict handover trigger points, and compare them with classical approaches like Random Forest(RF) and K-Nearest Neighbors(KNN). In our proposed approach, we also use oversampling methods such as the synthetic minority oversampling technique (SMOTE) and temporal SMOTE (T-SMOTE) to address class imbalance in HO event data. The implementation results demonstrate that deep learning models (LSTM and Bi-LSTM) achieve higher accuracy and F1 scores of rare handover events compared to KNN and RF. In addition, the Bi-LSTM model slightly outperforms the unidirectional LSTM, highlighting the benefit of capturing temporal patterns in both directions. These findings suggest that AI-driven handover prediction can significantly improve preparedness to receive handovers in 5G networks.

Download

Available for download on Wednesday, May 20, 2026

Share

COinS