This paper proposes power allocation algorithm in orthogonal frequency division multiplexing (OFDM) based cognitive radio (CR) system using decode and forward (DF) relay. It is reported in the literature that both classical i.e. uniform power loading scheme and water filling methods that allocate power on OFDM subcarriers based on the channel gain are not effective for cognitive radio network (CRN) as the schemes introduce large interference to primary user (PU). To this aim, the present work proposes a simple, computationally efficient yet effective power loading scheme that maximizes the transmission capacity of CR while keeping the interference introduced to the PU below acceptable limits. To meet the objective, proposed power allocation to subcarriers not only considers channel gains but also the relative distances between CR subcarriers and PU band. This capacity performance is further improved significantly using DF relay, where our research direction focuses on optimal sharing of power between CR and relay using maximal ratio combiner (MRC) at destination receiver. Performance for the proposed power allocation schemes are compared with OFDM based hybrid power allocation as well as joint subcarrier and power adaptation methods reported in the literature. Numerical results show that the relative gain on CR's capacity are 3.7 times and 2.3 times, respectively compared to the optimal and hybrid OFDM power allocation schemes, while for interference introduced to PU is set at 3 times 10-6 (in watt). Finally, a few related research problems are also highlighted as future scope of works.