A Class of Three-User Multiple Access Cognitive Radio Channels: An Achievable Rate Region

We consider three-user multiple access cognitive radio channels and obtain achievable rates for reliable communication over these channels. A synopsis of the channel model is as follows: before establishing communication with the receiver, one of the transmitters (named primary) splits its message-set into two parts and shares them among the remaining two transmitters (named secondary) via a noiseless channel of infinite capacity. We thus have a cooperative scenario with each secondary transmitter having noncausal knowledge of a part of the primary´´s message. First, we precisely define the channel model and then derive a set of achievable rates for this channel by employing a combination of two well-known techniques: superposition coding introduced for broadcast channels and Gel´f and-Pinkser´s binning principle used for coding channels with random parameters. Lastly, we consider the Gaussian model of this channel to numerically evaluate and plot the rate region. Results demonstrate that, by sharing its message in parts, the primary user achieves higher rates, while the two secondary users continue to transmit at the same rate as that of the classical three-user multiple access channel without message-sharing between the transmitters.