Outdoor-to-Indoor Performance Analysis of a Commercial Deployment of 5G mmWave

Abstract

While millimeter wave (mmWave) channel modeling and propagation studies using channel sounders have been carried out for many years, the performance of commercially deployed 5G mmWave cellular networks has only recently begun to be thoroughly evaluated, mostly in outdoor environments. A recent measurement study [1] predicted outdoor-to-indoor (OtI) mmWave downlink throughputs of 500 Mbps - 2.5 Gbps based on measurements using channel sounders, not with measurements on deployed networks and consumer devices. In this paper, we report the first detailed OtI measurements of commercially deployed 5G mmWave using consumer handsets in a location in Chicago where a Verizon 5G mmWave base-station (BS) is deployed across the street about 25m from an university dormitory. Our detailed indoor measurements of uplink (UL) and downlink (DL) throughput and latency contradict the results in [1] and demonstrate that OtI 5G mmWave reception is extremely variable: maximum DL throughput of about 1.8 Gbps is obtained only in a very small number of locations where the user equipment (UE) is line-of-sight (LoS) to the BS through an open window. In general, the 5G mmWave connection performed better than low-band 5G in terms of DL throughput. However for UL throughput and latency, the UE performed better when connected to low-band 5G under non-LoS (NLoS) conditions compared to 5G mmWave. Furthermore, when the windows are shut, i.e., there is no Verizon 5G mmWave reception indoors, we observed better OtI DL throughput from mid-band 5G deployed by T-Mobile compared to Verizon 5G NR in the low band. Thus on overall, there is only an extremely small advantage in performance from OtI 5G mmWave reception compared to low and mid-band 5G.