Internet users are increases, distributed denial of service (DDoS) attack present a very serious threat to the stability of the internet. The DDoS attack, which is consuming all of the computing or communication resources necessary for the service, is known very difficult to protect. The threat posed by network attacks on large network, such as the internet, demands effective detection method. Therefore, an intrusion detection system on large network is need to efficient real-time detection. In this paper, we propose the entropy-based detection mechanism against DDoS attacks in order to guarantee the transmission of normal traffic and prevent the flood of abnormal traffic. The OPNET simulation results show that our ideas can provide enough services in DDoS attack.