Research Article
Receiver-side Opportunism in Cognitive Networks
@INPROCEEDINGS{10.4108/icst.crowncom.2011.245868, author={Natasha Devroye and Petar Popovski}, title={Receiver-side Opportunism in Cognitive Networks}, proceedings={6th International ICST Conference on Cognitive Radio Oriented Wireless Networks and Communications}, publisher={IEEE}, proceedings_a={CROWNCOM}, year={2012}, month={5}, keywords={opportunistic interference cancelation cognitive network}, doi={10.4108/icst.crowncom.2011.245868} }
- Natasha Devroye
Petar Popovski
Year: 2012
Receiver-side Opportunism in Cognitive Networks
CROWNCOM
IEEE
DOI: 10.4108/icst.crowncom.2011.245868
Abstract
Cognitive radios may increase spectral efficiency by filling in spectral gaps, transmitting under the interference-temperature, or by exploiting transmitter-side information - all methods which rely on transmitter-side cognition. In this work we shift our focus to receiver-side cognition by extending work on opportunistic interference cancellation (OIC) to multi-user scenarios. We consider a single primary transmitter-receiver link which communicates simultaneously with a group of cognitive (secondary) users that form one of three classical multi-user channels: 1) a multiple access channel (MAC), 2) interference channel (IC), or 3) a broadcast channel (BC). When these cognitive users are permitted to transmit subject to peak interference at the primary receiver, we illustrate the benefit of having the primary share its transmission rate and codebook with the cognitive receivers. With these codebooks, when the channel conditions permit, the cognitive receivers decode both the primary and the intended cognitive message, which boosts the secondary rates as compared to treating the primary signal as noise. The primary users must not change their encoders/decoders and remain oblivious to the secondary operation. We obtain achievable rate regions for the secondary MAC, IC and BC in which the cognitive receivers opportunistically cancel the primary interference to achieve higher rates at the cost of codebook knowledge and, perhaps, increased decoding complexity.