Novel Algorithm to Reduce Handover Failure Rate in 5G Networks

Abstract

Ultra-reliable low latency communication (URLLC) is a key feature in 5G which requires improved mobility performance and reliability. In future, the number of devices are going to increase many times in 5G compared to current 4G, so the number of mobility (handover) scenarios are bound to increase many folds, and without proper technologies it may induce more handover failures. According to tests done in North America it is observed that handover failure (HOF) rate is 7.6% in urban areas and 21.7% in downtown area while successful recovery from HOF is only 38% [9]. Also, it is observed that if user equipment (UE) faces radio link failure (RLF) which leads to HOF, then the service interruption time is more. Therefore, to ensure better quality of experience (QoE) in 5G NR (New Radio), it is important to have minimal interruption time, and high handover success rate. In this paper we propose a novel Machine Learning (ML) and beam measurement based advance handover (HO) algorithm. In this concept, HO is initiated in advance before UE runs into RLF to ensure less HOF. In our proposed algorithm, the Network parameters used to train the ML model is based on serving cell reference signal received power (RSRP), block error rate (BLER), Timing Advance (TA) and serving beam direction. The proposed idea performance with existing Handover mechanism shows reduction of HOF rate by 35%.