Abstract

Abstract
Featured Application Short-range millimetre wave communications. The design of a uniformly spaced 1 x 4 linear antenna array using epitaxial layers of benzocyclobutene over an InP substrate is demonstrated. The array elements are conjugately matched with a uni-travelling carrier photodiode at the input. The phased array is optimised to counteract mutual coupling effects by introducing metal strips with isolated ground planes for each radiating element. The proposed antenna array can provide a gain of 10 dBi with a gain variation of +/- 3 dB. The array operates over a bandwidth of 10 GHz (295-305 GHz) with a wide scanning angle of 100 degrees in the broadside.

Abstract
In this paper, we investigate modification of transmission spectra of long-period fiber grating structures with an acoustic shock front propagating along the fiber. We simulate transmission through inhomogeneous long-period fiber gratings, pi-shift and reflective pi-shift gratings deformed by an acoustic shock front. Coupled mode equations describing interaction of co-propagating modes in a long-period fiber grating structures with inhomogeneous deformation are used for the simulation. Two types of apodization are considered for the grating modulation amplitude, such as uniform and raised-cosine. We demonstrate how the transmission spectrum is produced by interference between the core and cladding modes coupled at several parts of the gratings having different periods. For the pi-shift long-period fiber grating having split spectral notch, the gap between the two dips becomes several times wider in the grating with the acoustic wave front than the gap in the unstrained grating. The behavior of reflective long-period fiber gratings depends on the magnitude of the phase shift near the reflective surface: an additional dip is formed in the 0-shift grating and the short-wavelength dip disappears in the pi-shift grating.

Abstract
The basic idea of this work is to develop and test a control system for a solid-state transformer, interconnecting two distribution grids with also having the possibility of generating an isolated microgrid from the medium voltage grid. The system is modular and is easily adaptable to any power and voltage level, with different controllers for each subsystem. The system is assessed, through several MATLAB/Simulink simulations, for various operating points. The system is bidirectional and resilient to failures, being able to mitigate network anomalies, namely voltage and harmonic sags. When operating as an isolated microgrid, it can feed linear, non-linear, or unbalanced loads. © 2021 IEEE.

Abstract
Anomaly detection in electricity consumption data is one of the most important methods to identify anomalous events in buildings and electric assets, such as energy theft, metering defect, cyber attacks and technical losses. In this paper, a novel deep learning based approach is presented to detect anomalies in electricity consumption data one hour ahead of time. We address this challenge in two stages. First, we build an Long Short-Term Memory (LSTM) based neural network model to predict the next hour sample. Second, we use another LSTM autoencoder to learn the features of normal consumption. The output of the first stage is used as an input to the LSTM autoencoder. The LSTM autoencoder will learn the features of normal consumption and the input will be similar to output when applied. For anomalies, the input and output will be significantly different. The Exponential Moving Average (EMA) is used as a threshold and two types of anomalies are distinguished, local and global anomalies. Several weather features are considered in this study, such as pressure, cloud cover, humidity, temperature, wind direction and wind speed in addition to temporal and lag features. A feature selection method to find the optimal features that achieve good results is also implemented. We validate the proposed approach by comparing the detected anomalous consumption and the normal consumption within the same period, and the results demonstrate a drastic increase in electricity consumption during the anomalous periods. The results also show that the temporal and lag features have improved the efficiency and performance of the proposed method.

Abstract
Networking equipment that connects households to an operator network, such as home gateways and routers, are major victims of cyber-attacks, being exposed to a number of threats, from misappropriation of user accounts by malicious agents to access to personal information and data, threatening users’ privacy and security. The exposure surface to threats is even wider when the growing ecosystem of Internet-of-Things devices is considered. Thus, it is beneficial for the operator and customer that a security service is provided to protect this ecosystem. The service should be tailored to the particular needs and Internet usage profile of the customer network. For this purpose, Machine Learning methods can be explored to learn typical behaviours and identify anomalies. In this paper, we present preliminary insights into the architecture and mechanisms of a security service offered by an Internet Service Provider. We focus on Distributed Denial-of-Service kind of attacks and define the system requirements. Finally, we analyse the trade-offs of distributing the service between operator equipment deployed at the customer premises and cloud-hosted servers.