Unknown cyber-attacks have appeared constantly. Several anomaly detection techniques based on semi-supervised learning have been proposed to detect these unknown cyber-attacks. Among them, the Denoising Auto-Encoder (DAE) scheme performs better than others in accuracy but is not good enough in precision. This paper proposes a novel two-stage deep learning structure for network flow anomaly detection by combining the models of Gate Recurrent Unit (GRU) and DAE. By using supervised anomaly detection with a selection mechanism to assist semi-supervised anomaly detection, the precision and accuracy of the anomaly detection system are improved. In the proposed structure, we first use the GRU model to analyze the network flow and then take the outcome from the Softmax function as a confidence score. When the score is more than or equal to the predefined confidence threshold, the GRU model outputs the flow as a positive result, no matter the flow is classified as normal or abnormal. When the score is less than the confidence threshold, GRU model outputs the flow as a negative result and passes the flow to DAE model for flow classification. DAE then determines a reconstruction error threshold by learning the pattern of normal flows. Accordingly, the flow is normal or abnormal depending on whether it is under or over the reconstruction error threshold. A comparative experiment is performed using NSL-KDD dataset as benchmark. The results revealed that the precision using the proposed scheme is 0.83% better than DAE. The accuracy using the proposed approach is 90.21%, which is better than Random Forest, Naïve Bayes, One-Dimensional Convolutional Neural Network, two-stage Auto-Encoder, etc. In addition, the proposed approach is also applied to the environment of software defined network (SDN). By adopting our approach in SDN environment, the precision and F-measure are significantly improved.

This paper proposes an energy-aware mode-selection mechanism for device-to-device (D2D) resource allocation in the coexisting 5G network environment of D2D users and traditional cellular users by taking into consideration the mechanisms of mode selection, transmit power, and spectrum resource allocation simultaneously. In the proposed approach, all spectrum resources are fully shared by traditional cellular users and D2D users. We first determine the communication mode of all users after considering the transmission opportunity of traditional cellular users and analyzing the interference degree of different D2D communication modes. Next, we divide all users into three tiers according to the distances between users and the base station to solve the interference problem generated by excessing transmit power in uplink transmission. Afterward, based on the different communication modes, corresponding spectrum resource allocation mechanisms using the Hungarian algorithm are developed. We simulate a single cell environment using Python language and perform several simulations for different pairs of traditional cellular users and D2D users. The simulation results reveal that the proposed approach outperforms the proposed approach by Hou et al. in system throughput. With power control, the energy efficiency of the proposed approach could be enhanced by 20%.

In 802.11 wireless networks, a complete handoff procedure for a mobile node requires access point (AP) selection, AP switch, call admission control (CAC), IP address re-allocation, and network re-configuration. Most current handoff schemes deal only with either AP selection or IP address re-allocation. In this paper, an integrated handoff procedure is proposed. First, AP selection is accomplished by choosing an AP with the lowest channel utilization and smaller number of associated users. The information about load of each AP is reported through modified beacon frames. In the case of adopting load-based AP selection, the average throughput can be increased up to 56%, as opposed to pure SNR-based AP selection. Next, both CAC and IP address pre-fetch are performed simultaneously through the simplified DHCP procedure. Specifically, efficient limited fractional guard channel policy (ELFGCP) is proposed for the CAC phase. By adopting ELFGCP, the failure probability can be reduced as much as 45% from conventional LFGCP. Finally, the simulation results demonstrate the applicability of the integrated approach, and the overall disconnection time due to handoff can be reduced from 2.9 to 0.004 s using traditional handoff procedures.