There is no denying that 2G, 3G, and 4G have widened the use of mobile devices that ultimately changed the way we live and work. As has happened with every generation of wireless communication, the bandwidth of 4G, LTE, and LTE-Advanced is running out, and we can't wait for 5G to happen. If 5G comes even close to the hype, it will go beyond simply providing more bandwidth for people to watch movies on the go. 5G could make its way into the overall network infrastructure in ways its predecessors haven't. It could provide the means to greatly improve signal quality and the number of connections available in a given location.

The 2017 International Microwave Symposium in Honolulu was the site of a two-day 5G summit. Here, RF/microwave/wireless engineers heard from industry and academia representatives on what the 5G physical layer might become. Even though 3GPP will have radio standards in place by the end of 2017, there are many other aspects of the technology that will need to be worked out. It will take a worldwide effort to make that happen. Here are some of the issues still on the table for the physical layer that were presented on Day 2 of the IMS 2017 5G Summit.

5G will likely first appear in fixed-access applications, in the form of front-haul home/business Internet access. Typical range should run between 200m and 300m. Furthermore, wireless could take over from copper on the backhaul, especially where there isn't fibre to the cell.

20170611_IMS_5G_01 Figure 1: How 5G will be used in the early stages of its implementation.

Power amplifiers are perhaps the most limiting factor in 5G implementation. They distort incoming signals by adding harmonics and thus harmonic distortion. They need high breakdown voltages. Digital circuits need to get denser because of chip size. A trade-off also comes with power versus board size, a real issue for handsets.

Although power amplifiers degrade signals, they're not the only culprits. For example, an A/D converter's effective number of bits (ENOB) drops as sample rates increase. With increasing data rates, a loss of resolution can also add distortion to the signal chain. An 8-bit ADC's ENOB can drop to between 5 and 6bits. That limits the levels of modulation. Currently wireless systems can work with 64QAM (quadrature amplitude modulation) at 1Gbit/s. That's pretty good, but it eventually won't be good enough. Another ADC issue stems from the fact that a system needs to keep its ADCs synchronised. The limiting factor there is the phase noise of a phase-locked loop (PLL). The synchronisation isn't an issue for 4G data rates.

Locations of arrays in a phone need to be considered. Where do you put them? And you should consider the effects of plastics and other metal in the design.

Massive MIMO (much greater than 8×8) antenna arrays provide the means for beamforming and special multiplexing. That lets systems concentrate signal strength where needed and minimises wasted energy that comes from broadcasting. More antennas in a phased array improve gain. But, Massive MIMO has problems in that antennas around the edges of arrays don’t provide as much power as those in the centre. The solution is to go to multiple smaller arrays acting as a single set.

20170611_IMS_5G_02 Figure 2: Power efficiency may be an issue if Massive MIMO is used.

Power efficiency can be as low as 2% efficient. Beamforming provides a direct signal where needed from the base station. Multiple beams will be needed to cover a large venue like a stadium, utilising hybrid analogue and digital beamforming. Spectrum sharing could result in making better use of available bandwidth. To make that possible, multicarrier modulation could be used.

20170611_IMS_5G_03 Figure 3: Generalised multicarrier scheme presented at IMS 2017.

Current carrier frequencies (below 6GHz) simply lack the bandwidth to carry the desired data rates. mmWave signals look attractive because of their higher bandwidth, but there are trade-offs on range, and blockages, resulting in greater transmission losses.

Even if 5G works as well as the hype would have it, networks still have to support the legacy technologies 2G, 3G, 4G, and LTE. That adds considerable complexity.

5G is a worldwide effort. Frequencies will differ from country to country, and will probably need more than one type or radio such as enhanced OFDM, filtered OFDM, and others.

Issues in testing all of this were also stressed.