Volumen 19 (2019): Edición 1 (March 2019)

Nowadays, being in digital era the data generated by various applications are increasing drastically both row-wise and column wise; this creates a bottleneck for analytics and also increases the burden of machine learning algorithms that work for pattern recognition. This cause of dimensionality can be handled through reduction techniques. The Dimensionality Reduction (DR) can be handled in two ways namely Feature Selection (FS) and Feature Extraction (FE). This paper focuses on a survey of feature selection methods, from this extensive survey we can conclude that most of the FS methods use static data. However, after the emergence of IoT and web-based applications, the data are generated dynamically and grow in a fast rate, so it is likely to have noisy data, it also hinders the performance of the algorithm. With the increase in the size of the data set, the scalability of the FS methods becomes jeopardized. So the existing DR algorithms do not address the issues with the dynamic data. Using FS methods not only reduces the burden of the data but also avoids overfitting of the model.

Smart Grid (SG) is a major electricity trend expected to replace traditional electricity systems. SG has faster response to electricity malfunctions and improved utilization of consumed power, and it has two-way communication between providers and consumers. However, SG is vulnerable to attacks and requires robust authentication techniques to provide secure authenticity for its components. This paper analyses previous literature, comprising 27 papers on the status of SG authentication techniques, main components, and kinds of attacks. This paper also highlights the main requirements and challenges for developing authentication approaches for the SG system. This can serve as useful guidance for the development and deployment of authentication techniques for SG systems and helps practitioners select authentication approaches applicable to system needs.

Security is important in cloud data storage while using the cloud services provided by the service provider in the cloud. Most of the research works have been designed for a secure cloud data storage. However, cloud users still have security issues with their outsourced data. In order to overcome such limitations, a Dynamic Bloom Filter Hashing based Cloud Data Storage (DBFH-CDS) Technique is proposed. The main goal of DBFH-CDS Technique is to improve confidentiality and security of data storage in a cloud environment. The proposed Technique is implemented using data fragmentation model and Bloom filter. The DBFH-CDS Technique uses data fragmentation model for fragmenting the large cloud datasets. After that, Bloom Filter is employed in DBFH-CDS Technique for storing the fragmented sensitive data along with higher security. The DBFH-CDS Technique ensures high data confidentiality and security for cloud data storage with the help of Bloom Filter. The performance of proposed DBFH-CDS Technique is measured in terms of Execution time and Data retrieval efficiency. The experimental results show that the DBFH-CDS Technique is able to improve the cloud data storage security with minimum space complexity as compared to state-of-the-art-works.

Achieving secured data transmission is not always an easy job. Secret data sharing requires confidentiality and Undetectability. Steganography is preferred than cryptography to achieve undetectability. Steganography hides the secret data inside the other file such as text, audio, video, so that the existence of the secret data is completely hidden. Recent research focuses much on utilizing Voice over Internet Protocol (VoIP) calls as a carrier for data hiding. VoIP calls are much preferred among internet users for its wide availability, dynamic time limit and low cost. This paper focuses on data hiding methods that uses VoIP as a carrier. The paper also analyzes the performance of the algorithms using the three metrics undetectability, bandwidth and robustness.

The main goal of the present research is to classify images of plants to species with deep learning. We used convolutional neural network architectures for feature learning and fully connected layers with logsoftmax output for classification. Pretrained models on ImageNet were used, and transfer learning was applied. In the current research image sets published in the scope of the PlantCLEF 2015 challenge were used. The proposed system surpasses the results of all top competitors of the challenge by 8% and 7% at observation and image levels, respectively. Our secondary goal was to satisfy the users’ needs in content-based image retrieval to give relevant hits during species search task. We optimized the length of the returned lists in order to maximize MAP (Mean Average Precision), which is critical to the performance of image retrieval. Thus, we achieved more than 50% improvement of MAP in the test set compared to the baseline.

In movie domain, finding the appropriate movie to watch is a challenging task. This paper proposes a recommender system that suggests movies in cinema that fit the user’s available time, location, mood and emotions. Conducted experiments for evaluation showed that the proposed method outperforms the other baselines.

The paper offers new application of a Multi-channel Forward Scatter Radar (MFSR), which uses GPS signals for detection of air targets on their GPS radio shadows. The multi-channel GPS MFSR detector consists of several channels, which process information from several satellites simultaneously. The phenomena of diffraction in the near area is used for shadow target detection. The target is considered to be detected, if it is detected at least in one of detector channels. Two experiments have been made to verify the proposed detection algorithm. The results obtained show that the proposed multi-channel detection algorithm can be successfully used for detection of low-flying air targets at very short distances or the near area of diffraction. Such targets are undetectable in GPS bistatic radar.

Security has been a primary concern in almost all areas of computing and for the devices that are low on computing power it becomes more important. In this paper, a new class of computing device termed as Low Computing Power Device (LCPD) has been defined conceptually. The paper brings out common attributes, security requirements and security challenges of all kinds of low computing power devices in one place so that common security solutions for these can be designed and implemented rather than doing this for each individual device type. A survey of existing recent security solutions for different LCPDs hasve been presented here. This paper has also provided possible security solutions for LCPDs which include identification of countermeasures against different threats and attacks on these devices, and choosing appropriate cryptographic mechanism for implementing the countermeasures efficiently.

Estimating the speed and position using measurable electrical parameters would allow establishment of sensorless control systems for brushless DC motors, without the need to use expensive sensors for the rotor position and speed. When the motor is running, it heats up and the stator resistance rises. This heat-dependent change needs to be reflected in the observer, as it would produce an error in rated speed and position. An adaptive algorithm can compensate for the change of resistance as a disturbing effect of the motor heating. The adaptive algorithm for estimating the resistance is synthesized using the function of Lyapunov. This article is useful for estimation of brushless electric motor speed and position with observer. It contains simulations with an adaptive observer of resistance for sensorless estimation of speed and position in brushless DC motor through measurement of voltage and current.

This paper introduces an application of an Ant Colony Optimization algorithm to optimize the parameters in the design of a type of nonlinear robust control algorithm based on coefficient diagram method and backstepping strategy with nonlinear observer for the electrohydraulic servo system with supply pressure under the conditions of uncertainty and the action of external disturbance. Based on this model, a systematic analysis and design algorithm is developed to deal with stabilization and angular displacement tracking, one feature of this work is employing the nonlinear observer to achieve the asymptotic stability with state estimations. Finally, numerical simulations are given to demonstrate the usefulness and advantages of the proposed optimization method.

With the continuous increase of international oil prices, more and more shipping companies look for new solutions to the ever present question: How to reduce operational fuel consumption and decrease air pollution. Ship route planning is an indispensable part of the ship navigation process. In the modern world, the passage planning aspect of navigation is shifting. No longer do we see mariners drawing course lines on a paper chart. No longer do they calculate distances with compasses. Elaborate algorithms on various digital devices perform all these tasks. Algorithms plot the optimum tracks on digital charts and algorithms can decide how to avoid collision situations. Nowadays charter companies do not rely solely on the experienced navigators on board their vessels to decide the best route. Instead, this task is outsourced ashore to routing and weather-routing enterprises. The algorithms used by those enterprises are continuously evolving and getting better and better. They are coming popular because of another reason – more and more the shipping society support the newly idea for using crewless ships. However, are they up to the task to eliminate the human element in passage planning? In this article, we are going to review some of the weak points of the algorithms in use.