On the capacity of the multiantenna gaussian cognitive interference channel

Stefano Rini, Andrea Goldsmith

Research output: Contribution to journalArticlepeer-review

8 Scopus citations

Abstract

The capacity of the multiantenna Gaussian cognitive interference channel is studied. The cognitive interference channel is a variation of the classical two-users interference channel in which one of the transmitters, the cognitive transmitter, is also provided with the message of the second transmitter, the primary transmitter. We study the capacity of the multiple-input multiple-output Gaussian model, that is the channel in which the inputs are vectors and the outputs are obtained as linear combinations of the channel inputs plus an additive complex Gaussian noise. This channel models a wireless scenario in which transmitters and receivers have multiple antennas. For this channel, we derive capacity to within an additive gap, that is we show that inner and outer bounds to capacity lie to within a constant distance of each other. The gap between the inner and outer bounds depends on the number of antennas at the cognitive receiver and both bounds can be easily evaluated by considering jointly Gaussian inputs. We also derive capacity to within a constant multiplicative factor of two, that is we show that the ratio between inner and outer bound is at most two. The additive gap well-characterizes the capacity at high SNR, while the multiplicative gap is useful at low SNR. We also derive the exact capacity for a subset of the strong interference regime: in this subset, the primary transmitter can decode the cognitive message without loss of optimality. This new capacity result extends and generalizes previously known capacity results, in particular, the capacity in the very strong interference and the primary decodes cognitive regimes.

Original languageEnglish (US)
Article number6985749
Pages (from-to)2252-2267
Number of pages16
JournalIEEE Journal on Selected Areas in Communications
Volume32
Issue number11
DOIs
StatePublished - Nov 1 2014
Externally publishedYes

All Science Journal Classification (ASJC) codes

  • Computer Networks and Communications
  • Electrical and Electronic Engineering

Keywords

  • Capacity
  • Capacity to within a constant gap
  • Cognitive interference channel
  • Interference pre-cancellation
  • Superposition coding

Fingerprint Dive into the research topics of 'On the capacity of the multiantenna gaussian cognitive interference channel'. Together they form a unique fingerprint.

Cite this