What will 5G be? What it will not be is an incremental/nadvance on 4G. The previous four generations of/ncellular technology have each been a major paradigm shift/nthat has broken backwards compatibility. And indeed, 5G will/nneed to be a paradigm shift that includes very high carrier/nfrequencies with massive bandwidths, extreme base station and/ndevice densities and unprecedented numbers of antennas. But/nunlike the previous four generations, it will also be highly/nintegrative: tying any new 5G air interface and spectrum together/nwith LTE and WiFi to provide universal high-rate coverage and/na seamless user experience. To support this, the core network/nwill also have to reach unprecedented levels of flexibility and/nintelligence, spectrum regulation will need to be rethought and/nimproved, and energy and cost efficiencies will become even more/ncritical considerations. This paper discusses all of these topics,/nidentifying key challenges for future research and preliminary/n5G standardization activities, while providing a comprehensive/noverview of the current literature, and in particular of the papers/nappearing in this special issue.