Gibbs Haversine Reinforcement Learning Based Handover For 5g Enabled Seamless Mobility in Wireless Network

Abstract

Listen

Translate article

Seamless Mobility (SM) is crucial for bringing about better Quality of Service (QoS) like minimum handover latency with maximum throughput in 5G networks. In this work a method called, Jenkin Impulse Response Filtering and Reinforcement Learning-based Gibbs Haversine Distribution (JIRF-RLGHD) for optimal selection of target cells for the handover process to ensure seamless mobility is designed. The JIRF-RLGHD method is split into two sections. They are predicting the signal quality of both serving and adjacent wireless nodes using the Box Jenkin Impulse Response Filtering model. The second task remains in applying Reinforcement Learning-based Gibbs Haversine Distribution for optimal selection of target cells for handover to ensure seamless mobility in a wireless network. The overall proposed method was simulated on a Python programming interface language. The simulation results reveal that the JIRF-RLGHD method offers a greater delivery rate, and handover success with lesser handover latency at minimal packet loss rate. Numerical results show that the JIRF-RLGHD method performs better in terms of data delivery rate by 18%, and handover latency by 33% compared to existing methods.