Das, Debasis

With the advent of Blockchain and the Internet of Things (IoT), the Smart Grid is a rapidly growing technology in decentralized energy distribution and trading. However, this advancement came with some serious cyber security challenges and attacks, such as single-point failure due to a centralized architecture of smart grids, slow transaction processing, emerging cybersecurity threats, double-spending, fork, and fault tolerance. We propose a comprehensive framework for the smart grid called SmartGrid-NG to solve all these issues. Instead of using blockchain as a blackbox plugin tool, we also propose a reputation-based blockchain protocol called GridChain to increase the performance of blockchain-based smart grid systems. The security analysis illustrates that the SmartGrid-NG withstands the attacks mentioned above. The performance analysis also states that the consensus delay is reduced to 80%, throughput is increased up to 60%, and computation overhead and energy consumption are reduced up to 70%.

The rapid advancement in deep learning-based object detection methods has made them a prevalent choice for real-time applications. Families of object detectors, including one-stage detectors, two-stage detectors, and region-based CNN networks, offer superior performance in accurately detecting objects. Despite their high accuracy, the complex design and black-box functionality of these models are not directly transferable in aerial imagery. Also, raise questions among users regarding the transparency of the algorithm in locating objects. Consequently, to demystify the decision process of these models, there is a need for Explainable AI (XAI) tools. XAI enables an understanding of the significance of each pixel in an image, shedding light on the contributions that lead to the model's final output. In this context, this work will present an efficient, explainable, multi-scale vehicle detection network from high resolution aerial imagery, named as EVDNet. The EVDNet model has trained with two publicly available aerial image benchmark dataset DOTA and VEDAI. To enhance interpretability, we leverage XAI method using GradCam. The experimental results not only showcase the effectiveness and performance of the EVDNet model but also provide valuable insights into the object detection process. This research contributes to bridging the gap between complex object detection models and user understanding, offering a more transparent and interpretable approach to high-resolution aerial imagery analysis.

As remote sensing capabilities continue to advance, there is a growing interest in leveraging computer vision and natural language processing for enhanced interpretation of remote sensing scenes. This paper explores the integration of textual information with images to augment traditional disaster classification methods. Our approach utilizes a predefined vision-language model to generate descriptive captions for images, fostering a more nuanced understanding of the remote sensing data. Next, we seamlessly integrate the generated textual information with image data through multimodal training, employing a multimodal deep learning method for disaster classification. The system categorizes input data into predefined disaster categories, presenting a comprehensive and accurate approach to emergency response system development. Experimental evaluations conducted on the AIDER dataset (Aerial Image Database for Emergency Response applications) showcase the efficacy of our approach, demonstrating improved accuracy compare to unimodal approach and reliability in disaster classification. This research contributes to the advancement of intelligent emergency response systems by harnessing the synergy between vision and language in the context of remote sensing.

Deep learning (DL) based cognitive radio networks (CRN) serve as a potential solution to the dilemma posed by spectrum limits and the rising demand for vehicular ad hoc networks (VANETs) routing services. However, the unpre-dictability of VANET restricts the generalization potential of DL-based techniques. Variations in traffic volume, road topologies, and radio propagation characteristics affect the training data significantly. Therefore, in this paper, we propose RF -CVN, a recurrent reinforcement learning (RRL) technique to sense the spectrum and discover a trustworthy path between the source and the destination using belief transmission (i.e., channel conditions, interference levels, and vehicle locations). We first devise a deep recurrent Q network for a multi-channel access scheme for unlicensed users to use available channels. The RRL allows the Q function to learn hidden states in partial observation or highly time-correlated network sensing cases. Later, the trust values are used to gain a more nuanced understanding of the network state, thereby enhancing the efficiency and reliability of the routing process. In this work, we argue that trust should be an integral part of the routing process and, therefore, design a trust mechanism to select a path. The trust mechanism aims to detect those spectrums that over-utilize or under-utilize their channel capacity during the local training. The outcomes of our simulations indicate that our RF -CVN routing method outperforms traditional routing systems based on cognitive radio-based vehicular ad hoc networks in terms of network performance and spectrum sensing efficiency.

Stress is a significant societal issue, as it is the cause of many health problems and huge economic losses for companies. Detecting stress in computer users is technically challenging and, however, of the utmost importance in the workplace, especially now that remote working scenarios are becoming omnipresent. Owing to heightened competitiveness within the sector, companies are increasingly seeking enhanced efficiency and extended work hours from their personnel. However, with the existing deadline stress and all, they also face a work-life balance problem. To prevent stress from becoming chronic and provoking irreversible damage, it is necessary to detect it in its early stages. In this paper, we are proposing a multimodal system to detect the seven emotions of employees at the workplace. This study presents a novel COMBINED-STRESS model for detecting and analyzing stress. In our model, we try to imbibe this approach by fusing the 3 modalities, i.e., stress review data, audio, and face data, and predicting an output regarding the mental and stress health of the patient. In the audio model, we achieved an accuracy of 95.26% on training and 87% on the validation set using Bi-LSTM, whereas in the facial emotion detection model, we achieved an accuracy of 80% using the ViT and Bi-LSTM models combined. In the third model, we have used sentiment analysis and the PSS stress questionnaire to detect the stress percentage. At the end, we have combined all three models using the weighted combination of all three and predicted the final stress score or percentage of the employee.

Vehicular communication, underpinned by IEEE 802.11p/WAVE-based Vehicle Ad-hoc Networks (VANETs), is instrumental in the seamless functioning of intra-vehicle exchanges. However, a comprehensive assessment of these systems reveals suboptimal efficiencies at the data layer, specifically regarding default broadcast intervals. Such inefficiencies lead to escalated packet collisions and subpar utilization of the delay time counter - factors that undermine the synergistic interplay between Active Safety Systems (ASS), such as Adaptive Cruise Control (ACC), and their passive safety counterparts. To address these intricacies, this research proposes an innovative mathematical framework tailored for the IEEE 802.11p MAC layer. We propose a model that elucidates the intricate dynamics of the delay time counter and offers refined broadcast intervals buttressed by robust algorithmic strategies. Empirical evaluations, conducted in meticulously simulated vehicular environments, validate the prowess of the proposed paradigm, highlighting a decline in packet collision instances. Quantitative findings from this research evince a notable decrease in packet collision rates and a commensurate enhancement in communication reliability, pivotal for advanced vehicular systems. Such technical augmentations directly elevate the operational reliability of cutting-edge safety mechanisms, exemplified by systems like the Toyota Pre-Crash Safety System.

Know Your Customer (KYC) is the authentication and verification process of the user carried out in the financial sector by institutions such as banks, insurance companies, fintech companies, etc., and other sectors such as smart healthcare, real estate, telecommunication, online gaming and gambling companies before engaging in any financial activities. The KYC process is made mandatory for Financial Institutions (FIs) by governments worldwide to keep a check on terror financing and money laundering activities. However, the traditional physical KYC process has limitations such as high operational costs, large time consumption, privacy and security issues, repetitive processes across multiple institutions and user inconvenience. Hence, the FIs are looking for alternatives to the physical KYC. The digitization of the KYC process, known as e-KYC, was explored but had problems such as a single point of failure, repetition of KYC process across multiple FIs and lack of security. Blockchain-based e-KYC has become popular these days and caught the eye of FIs due to inherent properties of the blockchain such as decentralization, transparency, and immutable ledger. Moreover, blockchain-based solution is highly secure and cost-efficient. This paper proposes a blockchain-based e-KYC platform where users can register themselves and get their video KYC done after ether payment. Once successfully verified, the user will receive a KYC key. The user will submit this key to the banks or other FIs in which he wants to register. The FIs can verify whether the user has undergone the KYC process by looking in the blockchain using the KYC key. Our solution requires the user to submit his documents only at the e-KYC platform and not at different FIs for registration, thus preserving his privacy and, at the same time, ensuring the user's authenticity.

Finance is the backbone of any organization or government. The government's financial health relies heavily on managing its FinTech applications. In the current FinTech system, user financial information is stored centrally. In the centralized system, the information security risks are high. Thus, a decentralized system is required for better security, trust and safety management in the FinTech information system. Blockchain is a decentralized technology that can solve the issues of traditional FinTech applications like banking. In the blockchain, consensus protocols are responsible for maintaining a consistent copy of the blockchain at each node of the blockchain network. However, these protocols are not suitable for regular currency transactions due to high latency and low throughput. Therefore, to achieve reliability, low latency, and high throughput, we have introduced a new protocol called FinBlock. In FinBlock, the pipeline concept is introduced in blockchain to increase the transaction throughput. Additionally, the number of message broadcasts is also reduced, which further improves the latency. Moreover, FinBlock achieves a speedup of 3 with respect to traditional practical byzantine fault tolerance (PBFT) consensus protocol. The result showed that FinBlock achieved better transaction processing time, transaction throughput, and message count than the traditional PBFT, even with hundreds of nodes in the network with reduced message complexity.

With the rapid advancement in wireless communications and network technology, the vehicular ad-hoc network has become an interesting topic for researchers due to intelligent transportation systems on the road. However, the communications are performed through wireless mediums and involve vehicles, which have limited storage space and computation capability. Therefore, secure and lightweight schemes are required for authentication and communication purposes. In this paper, we propose an identity-based communication method using bilinear pairings to provide computationally better and secure communication between vehicles and infrastructure (V2V and V2I) in VANETs. This scheme is secure against various security attacks, namely, man-in-the-middle, forgery, impersonation, replay, etc. The computation and communication costs also have been improved considerably compared to other related schemes.

Smart application has been focused more to solve real-world problems like smart home, smart street lighting system etc., but the concept of data transmission between devices require precise attention to preserve security with better performance output in device communication. The blockchain is a decentralized computation and information sharing platform to enable different players to verify on the fly. This paper presents a hyper-ledger fabric-based blockchain architecture using a one-way hash function to employ distributed data storage, to enhance the security of the smart applications, and to improve performance in the computation. The proposed protocol resists to different attacks, i.e., modification, impersonation, replay, man-in-the-middle, stolen verifies, device injection, false reputation, and appending. Moreover, performance results are comparatively better rather than relevant vehicular communication protocols.