Handover (HO) is a fundamental process in cellular networks (CNs) that ensures continuous connectivity as users move across different coverage areas. With the evolution of mobile communication technologies, the application of Machine Learning (ML) to optimize HO has emerged as a promising research avenue. However, the current literature is fragmented, spanning various cellular standards, ML approaches, datasets, and evaluation practices. This paper presents a comprehensive Systematic Literature Mapping (SLM) to consolidate and structure existing research on ML-based HO optimization. A total of 2,772 articles were identified through a carefully validated search strategy, from which 86 primary studies were selected based on strict inclusion and exclusion criteria. The analysis reveals a predominant focus on LTE (53.8%) and 5G (44.1%), a high reliance on simulated datasets (82.6%), and the frequent use of reinforcement learning (46.5%) for tasks such as handover prediction (45%), ping-pong reduction (22%), and base station selection (10%). Despite these advancements, critical gaps persist, including under-use of real-world data, limited exploration of legacy networks such as 3G, and insufficient methodological standardization. This study contributes a structured synthesis of the field, highlights research trends and deficiencies, and offers actionable insights to guide future work in ML-driven mobility management for evolving CNs.

The proposed work offers a concise review of the application of machine learning (ML) to cellular network handovers (HO) via the Systematic Mapping Study (SMS) methodology, emphasizing the problem areas and requirements. The key points include the paramount role of high-quality data, with meticulous data acquisition and preprocessing as vital steps in ML dataset construction. The article identifies prevalent parameters for HO enhancement and underscores the diversity of ML algorithms, aligning them with specific data input and tasks. This study establishes a robust basis for forthcoming research in applying machine learning to cellular network HOs.

This work analyzes handover (HO) efficiency in mobile networks using real-world data, addressing key challenges that affect connection stability, latency, and network load. Unsuccessful HOs—often caused by suboptimal parameter settings, network congestion, or rapid user movement—are identified as a major problem. To address these issues, the study evaluates key radio frequency indicators such as Received Signal Strength Indicator (RSSI), Reference Signal Received Quality (RSRQ), Signal to Interference plus Noise Ratio (SINR), and the HO ratio, along with mobility-related factors including user direction and the ping-pong effect. A set of key performance indicators (KPIs) is used to quantify performance: Handover Rate (HOR), Handover Ping-Pong (HOPP), and Unnecessary Handover (UHO).The evaluation, based on real-world network data, shows an average HOR of 79.9%, an HOPP rate of 72.45%, and 14.83% of HOs classified as unnecessary. These findings reveal the limitations of traditional static threshold-based strategies and emphasize the need for adaptive, data-driven optimization approaches.The results demonstrate that a comprehensive HO strategy integrating multiple real-time parameters is essential for efficient mobility management and improving overall network reliability and user experience. This study reinforces the importance of leveraging real operational data to refine and validate mobility algorithms in modern cellular systems.

Este trabajo analiza el proceso de handover en redes LTE (Long Term Evolution) en dos escenarios, detectando las características y el comportamiento de los parámetros RF (Radio Frequency) mediante modelos predictivos. Los dos escenarios recopilan datos de mediciones de los parámetros de RF, el primero en una zona urbana de la ciudad de Quito con la finalidad de analizar el comportamiento y establecer las características de una zona con handovers. El segundo escenario busca obtener un modelo predictivo de las zonas de handover fallidos; en este caso, se realizan recorridos en una zona rural de la ciudad de Quito y se analiza la falla de handovers con fallas en llamadas VoIP, ya que en este servicio es más notable la pérdida de conexión en LTE. Para la recolección de datos se utilizan herramientas de monitoreo instaladas en teléfonos celulares y para el análisis de datos y la obtención del modelo de Machine Learning se emplean R y RStudio. Los datos recopilados fueron limpiados y transformados para obtener un DataSet único por cada escenario, luego se dividió en un set de entrenamiento y test. El set de entrenamiento fue procesado mediante la técnica de árboles de decisión, que permitió obtener un modelo gráfico del comportamiento de los parámetros RF que generan handovers o handovers fallidos según el escenario. Finalmente, con el set de test se evaluaron los modelos definiendo matrices de confusión y se calculó el accuracy respectivo, teniendo 96.34% en el escenario 1 y 95.5% en el escenario 2.