Key points:

First generation millimeter-wave devices using the IEEE 802.11ad WLAN standard never saw widespread deployment, since most current applications work equally well with the data rates offered by sub-6 GHz IEEE 802.11ac. However, more bandwidth-hungry applications are on the horizon and there is little doubt that millimeter-wave networks will be required in the not too distant future to support them. The design of 802.11ad devices also provided important insights into the feasibility and practical issues of integrating millimeter-wave in consumer devices. With the first millimeter-wave 5G deployments under way, we are now at the very important point where researchers gain insights into the real-world problems and performance of operational millimeter-wave mobile networks. As a consequence, it is very important to carry out measurement studies and ideally make the data available to other researchers. An additional discussion point were testbeds, where at a number of universities and research labs powerful millimeter-wave prototypes are being developed. Given the complexity and effort such designs require, it would be highly desirable to converge on one or more “standard” millimeter-wave platforms for the research community to use and extend (similar to what has happened with USRP/GnuRadio and WARP for 802.11).

Two major challenges remain for the practical deployment of mmWave networks. The first is how to realize pervasive coverage at low deployment cost. Due to the relatively short coverage of mmWave base stations, the network needs to be ultra dense, which in turn entails huge capital and operational costs. The second is how to best coordinate the higher layer network protocols with the mmWave PHY layer, so as to maximize the end-to-end network throughput. The highly dynamic mmWave channel may disturb the adaptation mechanisms of higher layer protocols such as TCP and real-time video delivery, which reduces the utilization of channel capacity.