5G Throughput-prediction Technology for 28-GHz Channels Using Physical-space Information

Advances in wireless communications, such as the 5th-generation mobile communication system (5G), have enabled a wide variety of devices to be connected to wireless networks.In 6G, all physical entities will be connected to wireless networks, and their physical-space information, such as position and velocity, will be available for new mobile services.NTT's Innovative Optical and Wireless Network (IOWN) will accelerate to obtain the physical-space information from various sensors.Therefore, mobile traffic is growing rapidly toward 6G.The use of the millimeter-wave (mmWave) bands is promising to increase the capacity of mobile networks.However, the mmWave link quality (LQ) is strongly affected by surrounding objects.To stably use mmWave bands, an effective solution is to predict future LQ and adaptively control wireless communication.This article introduces 5G throughputprediction technology that is based on deep neural networks using physical-space information and an automated 5G measurement environment using humanoid robots for deep-learning evaluations.