Providing statistical quality-of-service guarantees in cognitive radio networks with cooperation

Cognitive radios (CRs) opportunistically sharing the wireless spectrum with the legacy primary system (PS) are promising to improve the spectrum utilization. However, this cognitive spectrum hole access makes the bandwidth of the cognitive radio network (CRN) formed by multiple CRs varies severely, which makes providing quality-of-service (QoS) guarantee over the CRN a very challenging task. Considering cooperative diversity with the merit of alleviating bandwidth variation, this paper proposes a cooperative scheduling algorithm for the CR in the CRN to select a relay path with the lowest packet dropping probability for packet transmissions. Under this scheme, the upper bounds of the mean packet dropping probability have been analytically derived and the corresponding admission control mechanism is further proposed. Analytical results show that the proposed scheme provides a better QoS guarantee than direct transmissions without cooperative relay for the CRN. Simulation results also confirm that traffic sources admitted into the system can absolutely meet their mean packet dropping probability requirements.