Abstract This research focuses on the development of a synthetic traffic data generator implementing two different statistical models identified as long-range and shortrange dependence. Normal network traffic shows the long-range dependence feature by its traffic burstiness on different time scales. It is believed that anomalous traffic such as worm or denial-of-service can change the distribution to short-range dependence. In order to model long-range dependence of normal network traffic, Fractional Gaussian Noise (FGN) with the Hurst parameter greater than 0.5 has been implemented, while anomalous traffic is represented by FGN with Hurst parameter less than or equal to 0.5. In addition, anomalous traffic has also been modeled by an AutoRegressive model (AR). As a result, a synthetic network traffic data generator tool has been developed where the network traffic data can be generated easily based on pre-determined parameters.