International Journal of Sensors Wireless Communications and Control - Online First

These days, communication is governed by scaled-down devices with extremely high data transfer rates. The days of data transfer rates in megabytes are long gone and the systems are touching data speeds of hundreds of gigabytes. To cater to this requirement, the evolution of metasurface-based antennas working in the THz frequency range is gaining pace.

Metasurface layers in antennas have unique mechanical and electrical properties that make them feasible. Graphene is a preferred metasurface material when designing antennas. In this paper, two novel designs of graphene-based metasurface antennas are proposed. These patch antennas with graphene metasurface layers have two configurations for slotted patches. Both graphene and gold-slotted patches are tested in this paper, and results are presented and validated for various performance parameters like return loss, peak gain, peak directivity, and radiation efficiency.

The design of the proposed Graphene Patch Antennas (GPA) is done in HFSS, and once the design is complete, the data is collected for variations in antenna parameters and is further processed via machine learning for better convergence in terms of return loss using ensemble learning. Optimal tuning of antenna components like substrate length, patch length, slot width, etc., is done using two regressor algorithms viz Histogram-based Gradient Boosting Regression Tree (HBDT) and Random Forest Tree (RFT) in machine learning. The radiation efficiencies of the two designs presented in this work are 89.04%, 97.54%, an average radiation efficiency of of 93.29%. The bandwidth of the two antenna designs is 8.63 THz and 8.69 THz, respectively.

Experimental results indicate that the proposed designs are achieved better bandwidth, and radiation efficiencies compared to state-of-art designs.

As a communication paradigm that bridges the gap between virtual and physical spaces, the Internet of Things (IoT) has quickly gained popularity in recent years. In order to supplement the provisions made by security protocols, a network-based intrusion detection system (IDS) has emerged as a standard component of network architecture. IDSs monitor and detect cyber threats continuously throughout the network lifetime.

The main contribution is the development of a two-level neural network model that optimizes the number of neurons and features in the hidden layers, achieving superior accuracy in anomaly detection. Its include the utilization of deep anomaly detection (DAD) to protect networks from unknown threats without requiring rule modifications. The research also emphasizes the importance of feature extraction and selection, employing parallel deep models for segmenting attacks. The proposed supervised technique enables simultaneous feature selection using parallel models, enhancing the accuracy of IDS designs. A novel hybrid technique combining Deep Auto Encoders (DAEs) and DeepNets is proposed for intrusion detection.

The research present a data preprocessing stage, converting symbolic and quantitative data into real-valued vectors and normalizing them. Feature extraction involves utilizing DAEs to learn concise representations of datasets. The deep network architecture, including multilayer perceptrons and activation functions, is employed for feature extraction and classification. The proposed approach utilizes a DeepNets in the final stage in order to improve the rate of 97% accurate outputs and make it possible to achieve greatly accelerated execution durations.

When measured against the performances of other approaches to the extraction of features, the performance of the deep network platform is superior.