mmWave Change Everything
Is it better to have a high-gain or low-gain antenna?
The plan to introduce cellular services in frequency bands >6 GHz is driving an abrupt and unprecedented change in how devices and systems have to be designed, operated and tested.
The Friis propagation equation predicts losses at mmWave frequencies:
- To overcome these losses and provide a realistic link budget, it is necessary to use high-gain antennas comprised of multiple elements at both ends of the link
- High-gain antennas create narrow beamwidth signals
- Radio propagation at mmWave is very different:
- Very sparse and spatially dynamic, unlike rich multipath with Rayleigh fading
- The simplest use of large antenna arrays at the base station is beam steering – create narrow beams within the cell to direct signals to specific locations, possibly with reflections involved
- The steering only needs to know the direction of the user. Beamforming requires precise real-time channel state information (CSI)
- Beamforming requires full digital control of the amplitude and phase of every antenna element while beam steering can be done using simple analog phase shifters
- In a predominantly line of sight channel with several users in different locations, beamforming would simultaneously generate a beam towards each user much like beam steering
- The benefits of beamforming become more apparent as the channel becomes more scattered, which is when simpler beam steering is less effective
5G Operation at mmWave Frequecies
- mmWave has great potential (spectrum!)
- mmWave signals do not bend around corners (diffract) and are easily blocked orattenuated
- mmWave signals do bounce (reflect) readily giving rise to local scattering(multipath)
- mmWave signals act more like light rays, so can be directed using special antennas
- Path loss through the air is much greater at mmWave than at LTE bands
- Changing from 1 GHz to 28 GHz path loss increases by 28 dB over 1 m
- Cables are lossy and expensive, so most testing will be done over the air