In underlay spectrum sharing cognitive radio (CR), the secondary users (SUs) share the spectrum with the primary users (PUs) in such a way that the interference caused by the SUs should be less than the interference temperature (IT) threshold. Interference Alignment (IA) is considered as a novel interference management technique for CR networks (CRN) which eliminates the interference at the PUs and also the interference among the SUs. However, most of the iterative solutions for IA aimed at removing the interference between the users, does not take into account the QoS (Quality of Service) of the PUs and SUs. Also the achievable rate of PUs by employing IA is less than that of the maximum achievable rate in a MIMO (Multiple input multiple output) system with no interference. In this paper we first present an Antenna Selection (AS) scheme to significantly improve the QoS of PUs and then we have proposed an adaptive power allocation (PA) scheme which aims at optimum PA in view of the proportional fairness among SUs. Numerical results are derived for comparing the proposed schemes in terms of outage performance of PUs and SUs. Simulation results show the effectiveness of the proposed methods in terms of SNR.