Resource allocation scheme for MDC multicast in CRNs with imperfect channel information

In this paper, resource allocation problem with imperfect channel information in cognitive radio networks (CRNs) is studied concerning the multiple description coding (MDC) multicast transmission. The traditional unicast model is extended to MDC multicast in CRNs, which aims to maximize the total received rate of all cognitive radio (CR) users. Primarily, a new auxiliary variable, named as normalized channel power gain in this paper, is introduced to substitute the transmission rate as the optimization variable. Then a two-stage method is proposed to conduct the resource allocation: first the multicast group (MG) selection and the normalized channel power gain setting, second the optimal power allocation. Meanwhile, as only estimated channel gain for the interference channel gain is obtained, the primary users´ interference can not be accurately estimated when we carry out the power allocation. Consequently, we suitably enlarge the estimated channel gain for the purpose of interference control. It has been verified in the simulation results that the proposed scheme can improve the system performance apparently in terms of both throughput maximization and interference control.