Anomalous or unnormal multimedia medical devices are to yield anomaly network traffic and affect the diagnosis about medical issues. How to find anomaly network traffic is significantly important for normal applications of multimedia medical devices. This paper studies traffic anomaly detection problem in large-scale communication networks with multimedia medical devices. We employ empirical mode decomposition method and wavelet packet transform to propose an accurate detection method to capture it. Firstly, we use the wavelet packet transform to pre-treat network traffic. Network traffic is decomposed into multiple narrowband signals exhibiting more detailed features of network traffic. Secondly, the empirical mode decomposition method is utilized to divide these narrowband signals into the intrinsic mode function at different scales, in time and time-frequency domains. We calculate the spectral kurtosis value of the intrinsic mode function at these different scales to remove false components of the empirical mode decomposition. As a result, we can obtain new time and time-frequency signals which highlight the hidden nature of anomaly network traffic. Thirdly, we perform the reconstruction of empirical mode decompositions and wavelet packet transforms for the above time and time-frequency signals to attain a series of new time signals. Then we can find and diagnose abnormal network traffic. Simulation results show that our method is effective and promising.

Traffic matrix (TM) is a key input of traffic engineering and network management. However, it is significantly difficult to attain TM directly, and so TM estimation is so far an interesting topic. Though many methods of TM estimation are proposed, TM is generally unavailable in the large-scale IP backbone networks and is difficult to be estimated accurately. This paper proposes a novel method of TM estimation in large-scale IP backbone networks, which is based on the generalized regression neural network (GRNN), called GRNN TM estimation (GRNNTME) method. Firstly, building on top of GRNN, we present a multi-input and multi-output model of large-scale TM estimation. Because of the powerful capability of learning and generalizing of GRNN, the output of our model can sufficiently capture the spatio-temporal correlations of TM. This ensures that the estimation of TM can accurately be attained. And then GRNNTME uses the procedure of data posttreating further to make the output of our model closer to real value. Finally, we use the real data from the Abilene Network to validate GRNNTME. Simulation results show that GRNNTME can perform well the accurate and fast estimation of TM, track its dynamics, and holds the stronger robustness and lower estimation errors.

Anomalous traffic often has a significant impact on network activities and lead to the severe damage to our networks because they usually are involved with network faults and network attacks. How to detect effectively network traffic anomalies is a challenge for network operators and researchers. This paper proposes a novel method for detecting traffic anomalies in a network, based on continuous wavelet transform. Firstly, continuous wavelet transforms are performed for network traffic in several scales. We then use multi-scale analysis theory to extract traffic characteristics. And these characteristics in different scales are further analyzed and an appropriate detection threshold can be obtained. Consequently, we can make the exact anomaly detection. Simulation results show that our approach is effective and feasible.