Asian Journal of Convergence in Technology
Not a member yet
868 research outputs found
Sort by
Secure Internet of Things (IoT) Networks: Study the challenges and develop solutions for securing IoT networks, including authentication, access control, and data protection
Get PDFThe present study focuses on the security issues that arise in the context of IoT networks and puts forward potential remedies in the areas of authentication, access control, and data safeguarding. Through a comprehensive analysis of the extant scholarly works, we ascertain deficiencies and construct a framework incorporating sophisticated procedures and cryptographic techniques. The model in question guarantees data confidentiality and integrity while considering limitations in available resources. The efficacy of network security enhancement is demonstrated through the use of simulations. The present study offers valuable contributions to the domain of IoT security by furnishing pragmatic perspectives for safeguarding IoT networks and promoting confidence in the burgeoning IoT milieu
A Review Paper on DDoS Detection Using Machine Learning
Get PDFThere is almost no place in the world today that is not connected to the internet. One of the most widely used technologies is the IOT, which allows millions of devices to be connected through the internet. As this technology grows, DoS/DDoS attacks are the most common and dangerous threats. DDoS attacks are becoming more complex and they are becoming almost impossible to detect. Distributed denial of service is a subclass of denial of service. In the order to prevent DDoS attacks, many types of research have been conducted. Machine learning and deep learning are commonly used to prevent DDoS attacks. This paper describes different attack types, such as layer attacks. In this paper, we carried out a comparative analysis of machine learning algorithms to discover and classify DDoS attacks. A study of the effectiveness of detecting DoS/DDoS attacks in networks has been conducted
Social Networking Sites Fake Profiles Detection Using Machine Learning Techniques
Get PDFIn the present paper, we offer a model that might be applied to identify if an account is real or false. It is unnecessary to manually examine each account because our model, which uses a support vector machine as a classification technique, can simultaneously process an extensive accounts dataset. We are concerned with the community of fake accounts, and our issue is classification and clustering. We employ artificial neural networks (ANN) and machine learning (ML) to assess the likelihood that a Facebook friend request is genuine or not. The existence of bots and phoney profiles is another risk factor for personal data being collected for illicit purposes. Bots are computer programmers that can compile data about users without their knowledge. Web scraping is the term for this activity. The fact that this behaviour is legal makes it worse. Bots can be disguised or appear as false friend requests to access private information on a social networking site. Still, there is a 7% false positive rate in which our system fails to identify a fake profile correctly
Control of Self-Excited Induction generator based wind turbine using by IM controller
Get PDFThis paper focuses on the electrical generation part of a wind energy conversion system. After a brief introduction of the induction machine, the electrical generator used in this paper, a detailed analysis of the induction machine operated in stand-alone mode is presented. This paper shows the effect of magnetic saturation during self- excitation process in an isolated three phase induction generator, for a given capacitance and rotor speed value. When the steady state condition of a self-excited induction generator (SEIG) is attained, an increase of load causes a decrease in the magnitude of generated voltage and its frequency. The dynamic model of the SEIG system is developed using dq variables in stationary reference frame. For the validation of mathematical modelling, model of Matlab Simulink is developed and performed on three phase SEIG machine with the internal model controller (IMC). Simulation results of the self-excited induction generator driven by the variable speed wind turbine are presented in the last section of this paper. The process of voltage build up and the effect of saturation characteristics are also explained
Accelerated Low Power AI System for Indian Sign Language Recognition
Get PDFDeep Convolutional Neural Network (CNN) based methods have become more powerful for wide variety of applications particularly in Natural Language Processing and Computer vision. Nevertheless, the CNN-based methods are computational expensive and more resource-hungry, and hence are becoming difficult to implement on battery operated devices like smart phones, AR/VR glasses, Autonomous Robots etc. Also with the increasing complexity of deep learning models like ResNet-50, there is a growing demand for efficient hardware accelerators to handle the computational workload. In this paper, we present the design and implementation of a neural network accelerator tailored for ResNet-50 on the ZCU102 platform using Field-Programmable Gate Arrays (FPGAs) which offers and customizable solution to address this challenge. We systematically investigate the design choices and optimization strategies for deploying custom built ResNet-50 network trained for Indian Sign language translation of 76 gestures enacted and build in our labs for Doctor patient interface on FPGA-based accelerators. In order to enhance operational speed, we have employed various techniques, including parallelism and pipelining, leveraging Depthwise Separable Convolution. Furthermore, we have implemented hierarchical memory allocation for different offsets using threads. Additionally, we have utilized weight and data quantization to optimize operational speed while minimizing resource consumption, thus achieving low power consumption while maintaining acceptable levels of inference accuracy. We, evaluated our accelerated FPGA model against CPU interms of various performance metrics viz: frames per second (fps), Memory allocations, LUTs, DSPs and Block RAMs used. Our findings underscore the superiority of FPGA-based accelerators, as evidenced by achieving a frame rate of 2.7fps on the Xilinx Ultra Scale ZCU102 platform with int8 quantization, compared to 0.8fps for Single precision. In contrast, the CPU achieved a frame rate of 0.6fps. Notably, we observed a minimal accuracy variation of only 1.37% with int8 quantization, while no accuracy variation was observed for Single precision. Our implementation utilized 16 convolution threads and 4 FC threads operating at 200 MHz for single precision, whereas for int8, we employed 25 convolution threads and 16 FC threads operating at 250 MHz
Use of Machine Learning Applications for Speech Impaired People
Get PDFWhile sign language is very important to deaf-mute people to communicate both with normal people and with themselves; it is still getting little attention from the normal people. We, as normal people, tend to ignore the importance of sign language unless there are loved ones who are deaf-mute. One of the solutions to communicate with the deaf-mute people is by using the services of sign language interpreters. But the usage of sign language interpreters can be costly. A cheap alternative to replace the interpreters is the use of a model that can automatically translate their actions into words. This paper covers the development of such a model which helps to automatically detect our actions in real time using the Mediapipe model and translate the same into respective text format
Detection of Network Layer Attacks in Wireless Sensor Network
Get PDFThe Wireless Sensor Network (WSN) technology is being used in a huge number of monitoring applications. It consists of a large number of sensor nodes with limited battery life. These sensor devices are deployed randomly in a sensor zone to collect the data. But these are threatened and attacked by several malicious behaviors caused by some nodes, which result in security attacks. Several security attacks occur in different layers of the wireless sensor network. Due to these attacks, confidential information can be stolen by attackers or unauthorized users, which can cause several problems for authorized users. Cyber-attacks by sending large data packets that deplete computer network service resources by using multiple computers when attacking are called wormhole and Sybil attacks. It is important to identify these attacks to prevent further damage. To overcome these problems, we use a prediction module that consists of various machine learning algorithms to find the best-performing algorithm. we use XGBoost, Adaboost, Random Forest, and KNN algorithms. To train these algorithms, we have used the WHASA dataset which contains 10 different attacks of the VANET environment and benign (normal) class. By using these algorithms classification of attacks can be done which occur on the computer network service that is " normal " access or access under " attack " by Wormhole and Sybil attack as an output
Dynamic Pricing Strategies and Demand Forecasting in Inventory Models for Deteriorating Items: A Theoretical Framework
Get PDFThis study presents a theoretical framework for dynamic pricing strategies and demand forecasting in inventory models tailored for deteriorating items. Drawing from an extensive inventory models literature review, particularly focused on deteriorating items, our research study addresses the need for robust inventory control systems. The methodology integrates insights from various studies, including Sharma et al.'s examination of time-based uniform pricing models for deteriorating items and their analysis of trapezoidal demand rates in inventory models. Additionally, models proposed by Sharma and Bansal regarding time-dependent demand and fractional backlogging are considered. The theoretical framework also incorporates insights from Sharma's investigation into fixed deterioration rates with limited backlogging. The study leverages Sharma's surveys on developing inventory models and his overviews of inventory management literature. Findings underscore the significance of dynamic pricing strategies and demand forecasting in mitigating inventory deterioration, emphasizing the necessity for tailored approaches to deteriorating items. Implications of the theoretical framework are discussed, highlighting avenues for enhancing inventory control practices in various industries
Bucket Filling Algorithm
Get PDFThis research is about a bucket feeling algorithm. Bucket feeling algorithms are very important in our life. We use these algorithms for storing technology. There are some kind of bucket feeling algorithms. But these algorithms have some problems. To solve the problems we have proposed this algorithm. In this algorithm we tried to split this bucket into some packet. And we have tried to fulfil the individual packets and by fulfilling the packet, we can fill the bucket. In this algorithm, the minimum size of the packet should be the maximum size of the entity. By this, we can easily and efficiently can fill the buckets
Cooling System for HV Switching Personnel in Arc Flash Suit
Get PDF High Voltage switching is very dangerous and has caused many fatalities or damage to equipment. Two of the most common sources of HV switchgear accidents are a lack of following safety protocols and an insufficient understanding of the HV electrical systems; just as car drivers can avoid accidents with proper understanding of driving, so can HV personnel. But because HV switching accidents do happen, Switching Personnel are required to wear Arc Flash suits. But the experience of Switching Personnel in equatorial Malaysia is that they sweat profusely even after a minute of wearing such a suit. The reason is because these suits are designed and manufactured in temperate countries where warm armor like this will be welcomed. To solve this, in this research an air conditioning system was built to enable blowing cool air within the Arc Flash suits via a pipe going up from the bottom of the suit to the chest level. The robust car air conditioner system was driven by a 2 HP motor and a long flexible pipe will enable the switching personnel to reach many HV switchgears. One such system should be permanently placed in each substation