Abstract

Cloud computing has completely changed IT (information technology) by providing IT resources as services on the internet. However, certain types of attacks, such as interdependency attacks, impede its wide adoption. With the latter, an attacker who succeeds in compromising the VM of a user can traverse the hypervisor to launch an attack on the VM(s) of other users on the same hypervisor. Unfortunately, we note a lack of secure and performant allocation policies against this problem. Existing policies focus on security but ignore other factors, including workload balance and energy consumption, which are vital for commercial cloud platforms. In this context, we propose different allocation policies for choosing the datacenter server to which we allocate a new virtual machine. These policies aim to minimize the interdependence of different users’ VMs while keeping the system performant regarding workload balance and/or power consumption. By default, our allocation policies treat all legitimate users as attackers and host their virtual machines according to their efficiency and coverage. We first design a secure and balanced solution that increases workload balance to prevent the servers from being overused. Afterward, we propose an algorithm that addresses security, power consumption, and workload balance objectives simultaneously. Based on our simulation results, our solutions perform better than existing algorithms regarding security, workload balance, and power consumption. The balanced solution reduces the chance of an attacker to zero and increases workload balance linearly. In other words, the workload balance is between [5,35] , and it utilizes slightly more hosts than existing proposals, with gains between [2,8] . Although our final proposal is less secure than previous algorithms, it performs better, so it has a good workload balance ( [5,30] ) and consumes less energy.

Introduction

Cloud computing is one of the most remarkable advances in IT in the last two decades. It offers resource consumption on demand, a flexible environment, and easy to use. These facilities make it widely adopted by the customers. However, in cloud computing, the hypervisor allows multiple virtual machines (VMs) of different users to run simultaneously on the same physical server. Ideally, each of these users’ virtual machines should operate in isolation to maintain optimal security conditions. Unfortunately, perfect logical isolation has not been achieved in practice, leaving attackers with the possibility of launching attacks such as interdependency attacks, etc. [1], [2], [3], [4], [5], [6], [7]. With the interdependency attack, a malicious user who has compromised the VM(s) of a user i can traverse the hypervisor to launch an attack on the VMs of another user j≠i on the same hypervisor.

Hence, to tackle this issue, most of the proposals tried to satisfy security and/or performance constraints by using optimization methods such as heuristic algorithms [8], [9], [10], game theory approaches [3], [11], [12], [13], the multi-objectives optimization [14], [15], [16] since the problem is NP-hard [14]. However, we note the absence of a secure and performant virtual machine allocation technique against the interdependency problem. For instance, in [16], the authors proposed a secure solution against the interdependency attack that minimizes both attacker’s efficiency and coverage, which respectively represent the probability of success of the attacker and the probability that the virtual machine of a legitimate user will be compromised. Nevertheless, this solution overlooked essential performance constraints related to minimizing power consumption and maximizing workload balance in the datacenter. These two performance constraints are very important for commercial cloud platforms. The first one motivates a provider to allocate a lot of VMs to fewer servers to reduce the cost of energy consumption and the emission of carbon dioxide (CO2). Significant energy consumption leads to high energy costs among providers. On the other hand, maximizing the workload balance spreads users’ virtual machines among the servers to prevent the hosts from being over-utilized. To accommodate these two performance constraints essential for commercial cloud platforms, we propose extending [16]. We address the interdependency problem and the interdependency attack interchangeably throughout the paper. We also refer to the performance by workload balance and power consumption.

Conclusion

In this paper, we develop the first secure and performant solution against the interdependency attack between cloud users sharing the same hypervisor. It focuses on minimizing security metrics while considering power consumption and workload balance. This approach considers all legitimate users as attackers who attempt to hack the host’s hypervisor and gain unauthorized privileges on the VMs it contains. Specifically, we define a secure allocation policy that maximizes workload balance (SALAEC-B) and a secure and performant allocation policy that simultaneously optimizes security, workload balance, and power consumption (SPALAEC). We also show that these solutions are optimal with polynomial complexities synonymous with “feasible” and “efficient”. In addition, results from the simulation show that SALAEC-B is secure and balanced, and it performs better than its counterpart in the related work, PSSF-Balanced [14]. Finally, SPALAEC is also secure against the interdependency attack while being efficient regarding workload balance and power consumption. Furthermore, our VM allocation policies prevent the negative impact that can be caused by the failure of one of the servers, unlike in PSSF-LEAST [15] and PSSF-Balanced [14]. Our allocation policies do not consider the possibility of migrating a virtual machine from one server to another. We propose an energy-aware approach with virtual machine migration as future work to deal with the high energy consumption in cloud computing and service level agreements (SLAs) violations. In addition, during allocation, our algorithms use the amount of security investment, which does not change over time (i). Our future work may investigate (i) and other simulation scenarios with many users.

Abstract

Live migration of virtual machines (VMs) is a technique that moves active VMs between different physical hosts without losing any running states. Although it is desirable for administrators that the live migration is completed as quickly as possible, the pre-copy-based live migration, widely used in modern hypervisors, does not satisfy this demand on the current trend that VMs on which running applications are performance-critical such as database management systems (DBMSes) have quite large memory. DMigrate , presented in this paper, shortens the time for live-migrating VMs with even large memory DBMSes. To quickly produce the running state of the migrating VMs on the destination, DMigrate performs regular memory transfers while simultaneously constructing the DBMS’s buffer-pool by fetching the data items from the shared storage. We prototyped DMigrate on MySQL 5.7.30, QEMU 5.1.0, and Linux 4.18.20. The experimental results show that the migration time of the prototype is up to 1.71 × and 1.71 × shorter under workloads, including sysbench and TPC-C, than the default pre-copy and post-copy schemes, respectively.

Introduction

Live migration of virtual machines (VMs) is a technique that moves active VMs between different physical hosts without losing any running states. Within a local area network (LAN), the main task of the live migration is to transfer VM memory pages to the destination while the VM disks are assumed to be stored in the shared storage. Live migration is helpful to intra-datacenter administrations including load balancing [1], [2], power saving [3], [4], and effective software upgrades [5], [6]. It is used in real-world data centers [7].

Although it is desirable for administrators that the live migration is completed as quickly as possible, the pre-copy-based live migration [8], widely used in modern hypervisors [9], [10], [11], does not satisfy this demand on the current trend that VMs on which memory-intensive applications such as database management systems (DBMSes) are running have quite large memory. For example, Amazon RDS offers VMs whose memory sizes are up to 3904 GiB [12]. Pre-copy transfers updated pages of a VM from the source to destination iteratively until the number of dirty pages is below a threshold, suspends the VM, and then passes the control to the destination after sending all dirty pages and CPU states such as register values. The migration of such VMs requires a long time due to the large number of memory page transfers. In addition, a larger VM memory size makes each iteration longer and could result in the VM spawning dirty pages that have to be transferred in the next iteration.

Conclusion

The live migration of DBMS-running VMs is non-trivial due to their tremendously large memory footprint, disturbing live migration-based administration in datacenters. This paper presented DMigrate that shortens the time for migrating DBMS-running VMs. To produce the running state of the migrating VMs on the destination, DMigrate performs regular memory transfers while simultaneously constructing the DBMS’s buffer-pool by fetching the data items from the shared storage. We prototyped DMigrate and conducted several experiments. The experimental results show that our prototype successfully shortens migration times of the pre-copy and post-copy schemes, and the prototype is effective under our synthetic migration scenarios, namely, VM eviction, batch migration, and migration in congestion.

Abstract

     Cloud computing platforms constructed on Trusted Computing Technology are commonly referred to as trusted cloud computing. The core component of this technology is known as Trusted Platform Module (TPM). Its specifications are given by the consortium known as Trusted Computing Group (TCG). Though TPM is an immovable module and is fixed inside any computer, yet its credentials can be shared with other devices. This paper presents the accessibility model for Virtual Machine (VM) of a distributed trust protocol which is based on TPM, through which consumer’s credentials are shared to other trusted devices. These credentials essentially describe the integrity of VM.

Introduction

     Cloud computing is a new way of outsourcing IT services as an online service. Although varying to the field of distributed, cluster or grid computing the use of the Internet in cloud computing separates it from the cluster and the grid computation. As per the terminology of cloud computing implies, the word cloud refers to the Internet that rawhides the basic enigmas of cloud computing users and conveys the seamless thinking of accepting Information Communication Technology (ICT) computation facilities from the cloud service provider (CSP). The domain of cloud computing includes three components namely Cloud Infrastructure, Cloud Provider and Cloud Consumer. Cloud computing services are used in other forms of conveyance such as Software as a Service, Platform as a Service and infrastructure as a service. According the definition of SaaS, code and databases furnished to several clients. In PaaS, software development platform is provided. IaaS allows the consumer to produce a complete and fully functional virtual machine with network capability, data storage devices, CPU or GPU processing, run time memory and almost all other resources needed for integration.

     Cloud services can be used likewise we use utilities in our daily life i.e. pay what you use. IaaS model is a fundamental layer for the cloud environment. At this layer computing resources are designed for virtualization purposes, in order to achieve widespread distribution and deployment. Due to high deployment, there is a high risk of security attacks and it is very difficult for a cloud buyer to rely on such a multitenant site. Cloud computing and IaaS security issues such as data security and privacy, trust, multiple tenants, damage, and data leaks, malicious VM etc are shown in [2] and [3]. Caring these issues it is very difficult for the consumer to rely on the infrastructure provided to the consumer. User stores data in bought machines (VM) swung on the provider facet for processing and other purposes. It is important to examine the cloud equipment for privateness and integrity before, after and at some stage in VM implementation. The literature reviewed for this motive conveys the clean impact that there may be a loss of such structures that permit the user to discover a VM controlled with integrity and privateness. In the IaaS framework all environmental safety measures are done via way of means of the provider, so the user ought to depend now no longer best on his or her infrastructure however additionally on his or her personal countermeasures

Conclusion

     Cloud computing is an internet-centric technology for outsourcing IT services such as software, platform, and infrastructure. In this research work accessibility model for the guest VM is proposed. In the proposed model integrity values of guest VM are shared and synchronized with more than one authenticated client devices to access the guest VM from everywhere and anywhere.

Abstract

The use of computers and the internet has spread rapidly over the course of the past few decades. Every day, more and more people are coming to rely heavily on the internet. When it comes to the field of information security, the subject of security is one that is becoming an increasingly important focus. It is vital to design a powerful intrusion detection system in order to prevent computer hackers and other intruders from effectively getting into computer networks or systems. This can be accomplished by: (IDS). The danger and attack detection capabilities of the computer system are built into the intrusion detection system. Abuse has occurred and can be used to identify invasions when there is a deviation between a preset pattern of intrusion and an observed pattern of intrusion. An intrusion detection system (IDS) is a piece of hardware (or software) that is used to generate reports for a Management Station as well as monitor network and/or system activities for unethical behaviour or policy violations. In the current study, an approach known as machine learning is suggested as a possible paradigm for the development of a network intrusion detection system. The results of the experiment show that the strategy that was suggested improves the capability of intrusion detection.

Introduction

Over the past few years, there has been an increase in the usage of computer systems to make the lives of consumers easier and more convenient. When people try to takeadvantage of the amazing capabilities and processing capacity of computer systems, however, security has been one of the most significant problems in the field of computer science. This is because attackers frequently try to break into systems and act maliciously, such as stealing vital information from a corporation, rendering the systems useless, or even destroying the systems. Internal attacks, such as pharming, distributed denial-of-service (DDoS), eavesdropping, and spear-phishing attempts, are often among the most difficult to identify of all well-known attacks. This is due to the fact that firewalls and intrusion detection systems (also known as IDSs) often guard against attacks from the outside. At this time, the majority of systems authenticate users by analysing a login pattern consisting of the user ID and password. As a result of this, we have proposed in this study a security solution that we have dubbed the Internal Intrusion Detection and Protection System (IIDPS) [1]. This solution recognises hostile or malicious behaviour carried out against a system at the System call level. IIDPS uses data mining and forensic profiling techniques in order to mine system call patterns, also known as SC-patterns, which are the longest system call sequences (SC-sequences) that have repeatedly appeared numerous times in a user’s log file for the user. SC-patterns can be used to identify malicious activity. The user’s computer usage history is used to compile the user’s forensic features, which are then defined as a SCpattern that commonly appears in the user’s own submitted SCsequences but is rarely utilised by other users. This information is gleaned from the user’s computer.

Conclusion

As a consequence of this, within this work, we suggest asecurity system that we refer to as the Internal Intrusion Detection and Protection System (IIDPS). This system is capable of identifying hostile behaviour that is aimed towardsa system at the SC level. The IIDPS minessystem call patterns, also known as SC-patterns, which are defined as the longest system call sequences (System Call-sequences) that have repeatedly appeared numerous times in a user’s log file for the user. These are the operations carried out by the user, such as sending a file, updating a file, or viewing a file, and they are validated by an administrator. The user’s computer usage history is used to compile the user’s forensic features, which are then defined as a SC-pattern that commonly appearsin the user’s own submitted SC-sequences but is rarely utilised by other users. This information is gleaned from the user’s computer.

ABSTRACT

Electroencephalography (EEG) plays a vital role in detecting how brain responses to different stimulus. In this paper, we propose a novel Shallow-Deep Attention-based Network (SDANet) to classify the correct auditory stimulus evoking the EEG signal. It adopts the Attention-based Correlation Module (ACM) to discover the connection between auditory speech and EEG from global aspect, and the Shallow-Deep Similarity Classification Module (SDSCM) to decide the classification result via the embeddings learned from the shallow and deep layers. Moreover, various training strategies and data augmentation are used to boost the model robustness. Experiments are conducted on the dataset provided by Auditory EEG challenge (ICASSP Signal Processing Grand Challenge 2023). Results show that the proposed model has a significant gain over the baseline on the match-mismatch track.

INTRODUCTION

ICASSP 2023 Auditory EEG Challenge is designed to explore the relationship between auditory stimulus and evoked EEG signal. In this paper, we mainly focus on the first (match-mismatch) task. Traditional methods [1] adopt the linear model to fit the feature transform from the stimulus to the EEG signal. Recently, deep learning based methods have been proposed to improve the performance of relating speech with EEG signal[2, 3], where the feature transform modules are replaced by long context model, such as long short-term memory (LSTM), stacked dilated Convolutional blocks, etc.

CONCLUSION

The objective of the study is to learn the relationship between auditory stimulus and EEG. We propose a SDANet model with ACM and SDSCM to increase the crossing linkages rather than treating them separately. Results show that our model improves the classification accuracy effectively.

Abstract

Nowadays, Java programming is used in a variety of application systems as a highly portable object-oriented programming language. To assist its self-studies by novice students, we have developed the Java programming learning assistant system (JPLAS), and implemented the personal answer platform on Node.js. JPLAS offers several types of exercise problems at different learning levels, including the grammar-concept understanding problem (GUP), the value trace problem (VTP), the mistake correction problem (MCP), the element fill-in-blank problem (EFP), the code completion problem (CCP), and the phase fill-in-blank problem (PFP). Any answer is automatically marked by string matching with the correct one on the platform. In this paper, we present an implementation of the six problem types in a Java programming course in Okayama University. We generated 109 problem instances by following its curriculum and assigned them to 58 third-year students as homework before the final examination. Their solution results reveal the difficulty difference among the problem types and confirm the validity in the Java programming course.

Introduction

Java is a powerful, general-purpose object-oriented programming language. Java was created in 1995, and is running on more than three billion devices worldwide as one of the most popular programming languages. Nowadays, Java is used in a variety of application systems, such as web applications, mobile applications, desktop applications, games, IoT application systems, and cloud service systems, as a highly portable object-oriented programming language. Thus, many universities and professional schools are offering the courses to cultivate Java programming engineers.

To assist self-studies of Java programming by novice students, we have developed the Java programming learning assistant system (JPLAS) and implemented the personal answer platform on Node.js that will be distributed to students on Docker [1].

Evaluation

In this section, we present the implementation results of the 109 instances to 58 third-year students taking a Java programming course in Okayama University.

A. Solution Performances by Problem Types

First, we analyze differences of solution performances of the students by the six problem types. Table II shows the number of students who submitted answer files, the average correct answer rate, and the average number of answer submission times for each instance by them for each problem type. From this table, it can be observed that GUP, VTP, MCP, and EFP are easier for these novice students. The average correct answer rate and the average number of answer submission times are similar in them. CCP and PFP are harder for them, where both the correct rate and the number of submission times become worse. The average correct answer rate of two CCP instances at ID = 8 and 10, and one PFP instance at ID = 12 is less than 90%. According to Table I, they are on override, overload, structure, inheritance, abstract class, and interface, which can be difficult for novices to understand. The hint function should be implemented for them, which will be in future works.

Abstract

Nowadays many researchers are developing various distributed and decentralized frameworks for federated learning algorithms. However, development of such a framework targeting smart Internet of Things in edge systems is still an open challenge. In this paper, we present our solution to that challenge called Python Testbed for Federated Learning Algorithms. The solution is written in pure Python, and it supports both centralized and decentralized algorithms. The usage of the presented solution is both validated and illustrated by three simple algorithm examples.

Introduction

Federated learning was introduced by McMahan et al. [1] as a decentralized approach to model learning that leaves the training data distributed on the mobile devices and learns a shared model by aggregating locally computed updates. They presented FedAvg, a practical method for the federated learning of deep networks based on iterative model averaging, see Algorithm 1 FederatedAveraging in [1] on page 5. The main advantages of federated learning are: (i) it preserves local data privacy, (ii) it is robust to the unbalanced and nonindependent and identically distributed (non-IID) data distributions, and (iii) it reduces required communication rounds by 10–100x as compared to synchronized stochastic gradient descent (FedSgd).

McMahan’s seminal paper [1] inspired many researchers’ papers and in this limited space we mention just few of them. Immediately after [1], Bonawitz et al. [2] introduced an efficient secure aggregation protocol for federated learning, and Konecny et al. [3] presented algorithms for further decreasing communication costs. More recent papers are focused on data privacy [4, 5].

Conclusion

In this paper, we developed the federated learning framework targeting smart IoTs in edge systems called Python Testbed for Federated Learning Algorithms (PTB-FLA), with the primary intention to be used as a framework for developing FLAs on a single computer. The solution is written in pure Python, and it supports both centralized and decentralized algorithms. The PTB-FLA usage is both validated and illustrated by three simple algorithm examples.

The main PTB-FLA advantages are the following: (i) it keeps the application footprint small so to fit to smart IoTs and (ii) it keeps the installation as simple as possible (with no external dependencies).

Abstract

Electroencephalography (EEG), which tracks the brain waves that contain the brain’s neural activity, plays an essential role in detecting human motion and treating neurological diseases. In the Artificial Intelligence (AI) era, deep learning algorithms are widely used in human action recognition and classification. Various convolutional neural networks that process this signal are also being born. This paper provides a detailed survey of the application of deep learning to EEG signals and outlines the research process when classifying EEG signals. At the same time, this paper reviews the relevant research on the classification of human action EEG signals in recent years. Human motion signals usually use different deep learning algorithms and convolutional neural network architectures in the EEG signal analysis task. This article will discuss the advantages and challenges of each method in other studies. Finally, the paper discusses future directions for deep learning-based EEG signal classification.

Abstract

In recent times, the energy calibration methods are universally expanding with the goal of effectuating, reliably operating, and managing the utility system. The growing demand for power in the current environment has necessitated the mandatory installation of energy meters, as well as the development of new methods for calibrating meter readings and governing the effective use of energy resources. AMR (Automatic Meter Reading system) is one such modernization. This employs analogue or digital energy meters with the assistance of smart meters. Currently, energy scaling is done by hand, which is a timeconsuming process in the world of day-to-day networking demand and also requires skilled labour. The concept of AMR Systems is to overcome complexities in the rapidly growing field of energy management. This article proposes a smart energy meter based on IoT to detect the power theft. The proposed model consists of Arduino UNO, ESP8266, AC713 current sensors, and so on. The AC713 senses current usage with the help of the ESP32, which is then passed to the IoT platform. Though AMR is a very effective method, it costs the proxy of existing energy meters by SEM (Smart energy meters), which is highly inefficient. As a result, the proposed method focuses on detecting the power theft caused by public tampering. The proposed model is programmed by using a BLYNK software and simulated in PROTEUS software. The proposed system is then validated by using the simulated results.

INTRODUCTION

People are currently exposed to a medium that provides extreme ultimatums. The major crisis in today’s society is the energy crisis. The proposed solution provides an appropriate solution to control this issue. To deal with this, various methods are now available. One of these methods is to reduce domestic or household power consumption. The traditional method of meter reading requires a large number of hired hands to visit the habitat in person and take readings offline from the meter by checking each one individually. This procedure necessitates two-way cooperation as it is necessary for an authority to take readings in a house only when an individual is present in the house in order to avoid insecurities. This causes a lag in the private working lives of ordinary people because they must be present in person during this procedure. Furthermore, it is a laborious procedure that leads to chaos when a hired-hand is mistaken for any criminal activities. “Monitoring of Energy Tampering using IoT” is a novel application that has been developed to control power tampering from anywhere in the world. This system has been modernized so that the data is updated every 10 to 12 seconds. This system is fundamentally designed to reduce power theft, which contributes to the reduction of commercial losses; direct tampering can be reduced by using this system. The proposed solution is cost-effective and provides significant assistance to the electricity authority. It is also a time-saving procedure for both authorities and general public.

CONCLUSION

In the digital age, the proposed solution focuses on the IoT’s connection and associating factor. In this proposed design, an IoT and Microcontroller-based meter reading system is framed to continuously govern the scale of power and detect electricity tampering and notify the appropriate authority. The design’s goals were met when the customer could access the energy meter from anywhere. The display of energy consumption units is shown in LCD displays. The benefits of SEM are that it saves time and eliminates the need to wait for customers to pay their bills. This enables energy providers to engage in activities against lawbreakers.

Abstract

Solar energy forecasting has seen tremendous growth by using weather and photovoltaic (PV) parameters. This study presents new approach that predicts solar energy production by using the scheduled, unscheduled maintenance activities and weather data. The dataset is obtained from the 1MW solar power plant of PDEU (our university), which has 12 structured columns and 1 unstructured column with manual text entries about different scheduled and unscheduled maintenance activities, and weather conditions on the daily basis. The unstructured column is used to create new features by using Hash-Map, flag words and stop words. The solar power generation forecasting is formulated as a vector auto regression (VAR) optimization problem and total power generation forecasting is presented with the results of four different cases. The results have shown that the root mean square percentage error (RMSPE) in total power generation forecasting is less than 10% for different lag (p) values. The vector auto regression can forecast the unscheduled maintenance activities like Grid failure, Inverter Failure, scheduled maintenance activity like module cleaning, weather activity like cloudy along with total power generation forecasting for effective and efficient management of solar power plants. The power generation decay is different for all the PV sets which show the variations in the impacts of weather, aging and maintenance on the solar power plant. This research work has proven that the peaks of total power generation forecasting and prediction can be tracked in a better way by using daily unscheduled, scheduled maintenance activities and weather conditions.

Introduction

Solar energy forecasting is an emerging area over the last decade by using historical time series data collected through a weather station (such as weather variables wind speed and direction, solar irradiance, and temperature). The scheduled and unscheduled maintenance activities are carried out frequently for efficient and effective management of solar power plant. Another important parameter that contributes slowly but effectively is the age of the Solar PV module. The degradation of PV modules is further impacted by the conditions such as High temperature, solar cell shedding, discoloration, and delamination, etc. India has been endowed with non-depleting renewable energy resources like solar, hydro, geothermal, and wind and it is currently ranked no. 3rd globally in the renewable energy market. According to MNRE sources, about 5,000 trillion kWh of energy is falling on India’s land mass annually, with the majority receiving between 4 and 7 kWh per square meter per day. This is going to be one of the fastest-developing industries to boost rural electrification as well as the economy. Currently, there are more than 35 solar power plants in India. These plants are managed and operated by various government organizations. The operational expenditure of MW to GW solar photovoltaic power plants imposes a huge cost. Artificial Intelligence based solutions such as forecasting and recommendation-based systems can provide fruitful insights to reduce such expenditure.

Conclusions and Future Work

The maintenance activities like “grid failure”, “inverter failure”, “module cleaning” and the weather parameter “cloudy” have huge impact on the solar power generation forecasting and prediction. The optimum value of lag p=12 results in lowest RMSPE, RMSE and MAE for different cases. The error is reduced due to inclusion of duration for maintenance activities in this model for case 2. The VAR model used in this work is capable of forecasting all the four features as well as power generation target variable simultaneously. The predicted power generation is able to follow the trends of original power generation when the thresholding is applied to the predicted values of maintenance features (activities). The results have shown that the root mean square percentage error (RMSPE) in total power generation forecasting is less than 10% for different lag (p) values for all cases. The vector auto regression can forecast the unscheduled maintenance activities like Grid failure, Inverter Failure, scheduled maintenance activity like module cleaning, weather activity like cloudy along with total power generation forecasting which result in effective and efficient management of solar power plant. The power generation decay is different for all the PV sets which show the variation in the impacts of weather, aging and maintenance on the solar power plant. The effects of aging, dust and other environmental factors can be considered in the model as a future scope to improve the solar power generation forecasting and prediction.

Abstract

Undoubtedly, the evolution of Generative AI (GenAI) models has been the highlight of digital transformation in the year 2022. As the different GenAI models like ChatGPT and Google Bard continue to foster their complexity and capability, it’s critical to understand its consequences from a cybersecurity perspective. Several instances recently have demonstrated the use of GenAI tools in both the defensive and offensive side of cybersecurity, and focusing on the social, ethical and privacy implications this technology possesses. This research paper highlights the limitations, challenges, potential risks, and opportunities of GenAI in the domain of cybersecurity and privacy. The work presents the vulnerabilities of ChatGPT, which can be exploited by malicious users to exfiltrate malicious information bypassing the ethical constraints on the model. This paper demonstrates successful example attacks like Jailbreaks, reverse psychology, and prompt injection attacks on the ChatGPT. The paper also investigates how cyber offenders can use the GenAI tools in developing cyber attacks, and explore the scenarios where ChatGPT can be used by adversaries to create social engineering attacks, phishing attacks, automated hacking, attack payload generation, malware creation, and polymorphic malware. This paper then examines defense techniques and uses GenAI tools to improve security measures, including cyber defense automation, reporting, threat intelligence, secure code generation and detection, attack identification, developing ethical guidelines, incidence response plans, and malware detection. We will also discuss the social, legal, and ethical implications of ChatGPT. In conclusion, the paper highlights open challenges and future directions to make this GenAI secure, safe, trustworthy, and ethical as the community understands its cybersecurity impacts.

Introduction

The evolution of Artificial Intelligence (AI) and Machine Learning (ML) has led the digital transformation in the last decade. AI and ML have achieved significant breakthroughs starting from supervised learning and rapidly advancing with the development of unsupervised, semi-supervised, reinforcement, and deep learning. The latest frontier of AI technology has arrived as Generative AI [1]. Generative AI models are developed using deep neural networks to learn the pattern and structure of big training corpus to generate similar new content [2]. Generative AI (GenAI) technology can generate different forms of content like text, images, sound, animation, source code, and other forms of data. The launch of ChatGPT [3] (Generative Pre-trained Transformer), a powerful new generative AI tool by OpenAI in November 2022, has disrupted the entire community of AI/ML technology [4]. ChatGPT has demonstrated the power of generative AI to reach the general public, revolutionizing how people perceive AI/ML. At this time, the tech industry is in a race to develop the most sophisticated Large Language Models (LLMs) that can create a human-like conversation, the result of which is Microsoft’s GPT model [5], Google’s Bard [6], and Meta’s LLaMa [7]. GenAI has become a common tool on the internet within the past year. With ChatGPT reaching 100 million users within two months of release, suggesting that people who have access to the internet have either used GenAI or know someone who has [8]. Figure 1 demonstrates the working of an AI-powered chatbot where a user initiates requests, and after analysis using Natural Language Processing (NLP), is given a real-time response by the chatbot. This response is analyzed again to provide a better user experience in the proceeding conversation.

Conclusion

GenAI driven ChatGPT and other LLM tools have made significant impact on the society. We, as humans, have embraced it openly and are using them in different ingenious ways to craft images, write text or create music. Evidently, it is nearly impossible to find a domain where this technology has not infringed and developed use-cases. Needless to mention, cybersecurity is no different, where GenAI has made significant impacts how cybersecurity posture of an organization will evolve with the power and threat ChatGPT (and other LLM tools) offers. This paper attempts to systematically research and present the challenges, limitations and opportunities GenAI offers in cybersecurity space. Using ChatGPT as our primary tool, we first demonstrate how it can be attacked to bypass its ethical and privacy safeguards using reverse psychology and jailbreak techniques. This paper then reflects different cyber attacks that can be created and unleashed using ChatGPT, demonstrating GenAI use in cyber offense. Thereafter, this article also experiment various cyber defense mechanisims supported by ChatGPT, followed by discussion on social, legal and ethical concerns of GenAI. We also highlight the key distinguishing features of two dominant LLM tools ChatGPT and Googe Bard demonstrating their capabilities in terms of cybersecurity. Finally, the paper illustrates several open challenges and research problems pertinent to cybersecurity and performance of GenAI tools. We envision this work will simulate more research and develop novel ways to unleash the potential of GenAI in cybersecurity.

Abstract

In recent times, there has been a drastic increase in the number of road accidents and their severity. These accidents mostly happen due to human error. Some common human errors like distracted driving, drowsy driving, driving at high speeds and delayed reaction times of the drivers result in fatal accidents that cause loss of life and damage to property. These factors can be eliminated by means of self-driving cars. A driverless car (or self-driving car or autonomous car) is a vehicle that can travel from one point to another by self-navigating without any human intervention by using sensors, cameras, radar, GPS, and AI (Artificial Intelligence). We will be implementing a Driverless Car using Deep Learning, OpenCV, DonkeyCar, TensorFlow and C++/Python.

Introduction

There has been a large spike in the number of automobile accidents these days. According to the U.S. Department of Transportation (USDOT), it has been estimated that about 20,160 people have died in road accidents in the first six months of 2021 in the U.S. alone. In India, data from the National Crime Records Bureau (NCRB) shows that in 2020, an average of 37 people died per 100 car crashes which results in a total of about 1.3 lakh deaths. Also, we need to note that these numbers include the period of lockdown due to the Covid-19 pandemic, during which there was minimal traffic movement.The major causes of these catastrophic accidents include distracted driving, drowsy and drunk driving. These factors lead to increased human error, which in turn increases the risk of having an accident. We can overcome the dangerous consequences of human error using autonomous vehicles or driverless cars. Most leading automobile manufacturers across the globe have started research and development of semi-autonomous and fully autonomous cars.

Conclusion & Future Scope

Driver errors while driving are the most prominent cause of traffic accidents. Driverless cars have the following benefits related to road transport: reduce stress due to driving, reduce driver cost, provide mobility of non-drivers, increased safety, and decrease in traffic congestion. Even a person with disabilities can make use of autonomous vehicles the same as any other individual and can even customize the vehicles for more convenience. A self driving car not only increases safety, but it also improves efficiency and productivity of the people who spend most of their time stuck in traffic and trying to reach their work places. Donkey is a self-driving car platform which uses a high-level Python library to attain autonomous driving. We have trained the Donkeycar to drive on its own based on driving style and path. This uses a supervised learning technique often referred to as behavioral cloning. Next, we carry out steering and throttle calibration. The reason for calibrating the car is to make it drive consistently. Later, the car is driven using a web controller/ physical joystick controller. As the driving car becomes reliable, we used Keras to train a neural network to drive like a human.

Abstract

Developed by OpenAI, ChatGPT (Conditional Generative Pre-trained Transformer) is an artificial intelligence technology that is fine-tuned using supervised machine learning and reinforcement learning techniques, allowing a computer to generate natural language conversation fully autonomously. ChatGPT is built on the transformer architecture and trained on millions of conversations from various sources. The system combines the power of pre-trained deep learning models with a programmability layer to provide a strong base for generating natural language conversations. In this study, after reviewing the existing literature, we examine the applications, opportunities, and threats of ChatGPT in 10 main domains, providing detailed examples for the business and industry as well as education. We also conducted an experimental study, checking the effectiveness and comparing the performances of GPT-3.5 and GPT-4, and found that the latter performs significantly better. Despite its exceptional ability to generate natural-sounding responses, the authors believe that ChatGPT does not possess the same level of understanding, empathy, and creativity as a human and cannot fully replace them in most situations.

Introduction

Did you know that you could be talking to a machine without even knowing it? You might just be doing that with ChatGPT, an advanced natural language processing model developed by OpenAI, a research company co-founded by Elon Musk, Sam Altman, Greg Brockman, Ilya Sutskever, and John Schulman [1]. ChatGPT is an advanced natural language processing (NLP) model that is trained on a vast amount of data [2], including billions of web pages and documents, making it capable of generating human-like text responses to prompts [3]. Since its launch, it has quickly become one of the fastest-growing consumer applications in history, with an estimated 100 million active users monthly. Its vast knowledge base and language processing capabilities have the potential to revolutionize the way people interact with technology, making it easier and more natural to communicate with machines. However, while ChatGPT has impressive language processing capabilities and is an exciting technology with a wide range of potential applications in various fields, it still has limitations and challenges, including bias and the occasional generation of nonsensical output, known as “hallucination” [4].

Conclusions

This paper examines the evolution of ChatGPT and studies its applications, opportunities, and threats with a focus on the business and industry and education domains. We conducted an experiment to evaluate the effectiveness of GPT-4 in an educational setting. While ChatGPT has revolutionized natural language processing and has the potential to save time and money by automating routine tasks, it is essential to be mindful of the potential threats and take steps to mitigate them, as it can produce misleading results and biases, raise ethical concerns, and be misused. That is why some countries, such as Italy, have banned the use of this AI [25,26], and responsible use of ChatGPT must be a top priority to ensure that everyone benefits from its use. While these models have shown impressive capabilities, we believe it is unlikely that they will be able to fully replace humans or individuals in all tasks and situations, and the jobs that are most likely to be affected are the ones that require routine and repetitive tasks [27]. Machines may struggle to replicate human intuition, emotion, creativity, and intelligence, which are essential for many tasks; however, they can assist humans in developing new ideas and insights based on existing data and may be able to create entirely new concepts with human input and guidance.