This paper proposes a resource allocation and decoding strategy at the secondary systems for underlay and interweave multi-carrier cooperative cognitive radio. The objective is to maximize the sum rate of both primary and secondary systems, taking into account the interference threshold constraint and the fact that the primary receiver always considers interference as noise. The decoding strategy at the secondary receivers is either Successive Interference Cancellation or treating interference as noise, depending on their channel gains. The sum rate and secondary rates are highly improved, compared with other resource allocation algorithms. The proposed algorithm benefits from multi-user diversity when the number of secondary cells increases.