Predictions for 2018
2018 will see continued growth in mmWave 5G development activity. Despite the fact that the decision on which frequency bands will be allocated for 5G will be not be confirmed until the 2019 World Radio Conference in (WRC-19), we have already undertaken custom MMIC and module design work in most of the main candidate mmWave 5G bands (FCC licensed bands at 28 GHz, 37 GHz, 39 GHz, and the 26 GHz 5G “Pioneer Band” in Europe). It is becoming increasingly apparent that a single worldwide mmWave band for 5G is unlikely to be adopted and multiple bands are likely to be progressed.
Some of the early applications for mmWave 5G, such as fixed wireless access, are likely to use point to point or point to multi-point links. However, non-line of sight mmWave links will ultimately be required for many of the planned 5G applications. Various approaches are currently being trialed but it is likely that some form of electronic beam steering, either phased array or switched antennas, will be utilized. There is still much debate about the number of beams and Tx/Rx channels. Whilst some are proposing user terminals with many channels/elements, I feel a more modest number of channels (say 4 or 8) is more likely.
GaN technology also continues to go from strength to strength. LDMOS was previously the exclusive semiconductor technology choice for cellular base station PAs, but commercially competitive GaN transistors are now available and the technology is gaining ground from LDMOS, with a significant percentage of base stations already using GaN PAs. GaN has a higher power density than LDMOS and can provide a higher Power Added Efficiency (PAE), particularly at higher RF frequencies. I believe that GaN will continue to take market share from LDMOS in the cellular base station PA market and will become the preferred choice for the higher frequency cellular bands.
[Originally published on on EE World Online]