Research on Measurement Relaxation and Predictive Handover Based on LSTM Networks

Ziqi Dai, Gengshuo Liu, Weidong Gao, Xu Zhao, Kaisa Zhang, Yuchen Guo

Abstract: In this study, we explore the application of 5G RedCap technology in smart factory scenarios, focusing on how inspection vehicles with fixed trajectories can maintain low power consumption while ensuring service reliability. The development of 5G technology, along with denser base station deployments and more frequent handovers, increases the likelihood of handover failures, making it challenging to achieve low power consumption while ensuring system reliability for RedCap. This paper investigates the use of Long Short-Term Memory (LSTM) networks to predict Reference Signal Received Power (RSRP) and optimize measurement relaxation algorithms and handover decisions based on these predictions. Specifically, we consider scenarios where low-speed or stationary user equipment start moving upon receiving service requests and need to exit the measurement relaxation state in a timely manner before handover. By predicting future RSRP values, we propose an algorithm to precisely control the timing of exiting the measurement relaxation state and make effective handover decisions based on the predicted data. The results show that this method significantly improves network performance and reliability while reducing energy consumption, providing substantial practical value for the management and optimization of 5G RedCap networks.

External IDs:dblp:conf/icct/DaiL0ZZG24