Non-Orthogonal Multiple Access: Instead of modelling the ideal case for multiple users using the channels in perfectly separable (i.e. orthogonal) ways, the system design exploits the fact that non-perfect information recovery can still lead to a higher sum rate.

NOMA access schemes might be enabled by things like Successive Interfence Cancellation, or time-domain distribution of information using coded multiple access followed by detection schemes that one would typically expect in channel decoders.

They typically trade-off

  • theoretical per-user rate vs
  • sum rate in a noisy environment vs
  • synchronization requirements vs
  • synchronization overhead vs
  • computational complexity vs
  • latency bounds vs
  • robustness,

where in orthogonal access schemes (classical TDMA, FDMA, CDMA) a few of these pairwise trade-offs are already set.

The topic gained interest during the 5G design phase; contributing factors for that might have been availability of compute resources at least at base stations, MIMO systems and their inherent ability to gain separation-aiding information for multi-user access becoming the norm, and the desire to integrate many low-intensity/bounded latency M2M devices into cellular network infrastructure, where requests for resources are infeasible.