In its broadest sense, 5G is envisaged as an infrastructure spanning all wireless networking needs which at one end of the scale can provide high throughput, low latency connections for ultra-high definition video and virtual reality applications and at the other end support connection of millions of low bandwidth low power machine-to-machine (M2M) devices and sensors. Achieving this breadth of application will require support for relatively low frequency spectrum (a few hundred MHz) to achieve coverage and in building penetration to the smallest low power M2M devices and millimetre wave spectrum perhaps up to 300GHz to achieve high speeds and communication using highly directional beams. It is likely that 5G will not fully replace existing technologies such as 4G and Wi-Fi. Instead 5G will be a mix of radio access technologies that encompasses aspects of and integrates with existing systems whilst introducing novel spectrum access and management methods.

The 5G performance vision

  • Support a 1000x data volume increase
  • Peak data rate (fixed) 10 to 50Gbps
  • Peak data rate (mobile) 5Gbps
  • Edgeless cells. No cell boundary
  • 1Gbps everywhere
  • Latency 10ms

Achieving the 5G performance vision will not be easy. Significant developments will be needed at all layers of the network stack. Consequently, we have examined the challenges that must be addressed to enable the physical and data link layers (Layer 1 & 2 of the OSI model) as these are concerned with the physical transmission media – in this case radio spectrum.

The major challenges are:

  • Spectrum allocation
  • Radio access technology
  • New coding and modulation scheme
  • MIMO & antenna design
  • Low cost device development

Achieving the 5G performance vision will not be easy.


Current 5G research considers the possible use of spectrum between 3GHz and 300GHz referring to this wide band as millimetre wave or mmWave. There is not yet agreement regarding the spectrum to be used for provision of the 5G performance layer but research to date has largely focused on spectrum between 23GHz to 90GHz for the following reasons:

  • Physical properties permit forming of narrow beams with a consequent reduction in fading, multi-path and interference
  • RF integrated circuit fabrication improvement enables low cost transceiver devices at frequencies that until recently were impossible in silicon. Silicon-based devices operating at 60GHz and around 80GHz are now relatively commonplace.
  • Small form factor multi-element antenna arrays are possible. Size decrease is proportional to wavelength.
  • Globally there are existing licensed bands in this range that are currently underutilised but have large chunks of contiguous spectrum. Additionally the oxygen absorption band (57-64GHz) and 24.05-24.25GHz are designated for unlicensed use.