Accurate Seamless Vertical Handover Prediction Using Peephole LSTM based on Light-GBM Algorithm in Heterogeneous Cellular Networks

Present and future mobile networks represent a combination of wireless radio access technologies (2G to 6G) and beyond, all coexist. Seamlessness Vertical Handover (VH) decision-making is still a challenging issue in heterogeneous cellular networks due to the dynamic conditions of networks, different demands on QoS, and the latency of the handover process. Maintaining a very high accuracy VH decision requires considering several network parameters. There is a trade-off between the gain of the VH accuracy and the corresponding latency in the computational complexity of the decision-making algorithms. This paper proposes a lightweight VH prediction DL strategy based on the Light-GBM feature selection and Peephole Long Short-Term Memory (PLSTM) prediction algorithm. For dense networks with large datasets and high-dimensional data, Peephole LSTM and fast feature selection, namely the Light-Gradient Boosting Machine (LightGBM), can reduce the computing complexity while preserving prediction accuracy and excellent performance levels. The proposed algorithm is evaluated using three case study scenarios using different feature selection thresholds. The performance evaluation is achieved by training and testing the proposed model, which shows an improvement using the proposed Light-GBM and Peephole LSTM in terms of reducing the number of features by 64.28% and enhancing the VH accuracy prediction by 43.81% in Root Mean Squared Error (RMSE) and reducing the VH decision time of up to 51%. Furthermore, a network simulation using the proposed VH prediction algorithm shows an enhancement in overall network performance regarding the number of successful VHs rate is 87%. Consequently, the data throughput is significantly enhanced.