Netaji Subhas Open University

Efficiency is essential in lightweight cryptography to ensure robust protection within constrained environments. This paper introduces a lightweight block cipher built on a Non-Linear Feedback Shift Register (NFSR), referred to as the NFSR-Based Block Cipher (NFBC). The cipher is designed for constrained environments where both efficiency and strong security are critical. NFBC operates as a Nonce-based Authenticated Encryption with Associated Data (NAEAD) scheme, providing confidentiality and authenticity within a single framework. The design achieves robust protection while maintaining minimal hardware overhead. NFBC operates on 128-bit data, key, and nonce sizes, fully aligning with National Institute of Standards and Technology (NIST) lightweight cryptography requirements. The design integrates a Non-Linear Feedback Shift Register (NFSR) for generating high-entropy round subkeys, dynamic chaotic substitution box (S-boxes) for strong confusion, and a Group Permutation (GRP) mechanism that ensures rapid diffusion with hardware-friendly efficiency. Security is rigorously evaluated across multiple dimensions. NFBC passes all 15 NIST Statistical Test Suite (STS) tests, achieves near-ideal avalanche and Bit Independence Criterion (BIC), and demonstrates consistently high nonlinearity across 200 independent chaotic S-box instances. The S-box achieves a maximum differential probability of 10/256 and maximum linear probability of (72/256)^{2}, confirming strong resistance to classical attacks. To strengthen theoretical guarantees, a provable branch-number analysis is introduced, establishing conservative lower bounds on active S-boxes (ge 45 across 20 rounds), with corresponding differential and linear probabilities that are vanishingly small. Formal AEAD security is ensured through an Encrypt-then-MAC (EtM) construction. Implementation on a Xilinx Artix-7 FPGA confirms practicality: NFBC achieves superior throughput-per-area compared to several NIST Round-2 lightweight candidates (SPOC, SPOOK, GIFT-COFB, ESTATE, SAEAES, Oribatida) and CAESAR finalists (Ascon-128, Ascon-small, CLOC-AES, SILC-AES). Estimated countermeasure overheads for masking and fault detection demonstrate feasibility against side-channel and fault attacks. These results confirm NFBC as a secure, efficient, and practically deployable lightweight cryptographic solution.

The rate of growth of cross-border insurance transactions has been very high in the last few years as a result of globalization of the financial market, the emergence of multinational insurance and the digital platform. Despite this growth, the global insurance sector continues to face high regulatory disintegration, which creates inconsistencies in licensing, solvency regulation, reporting and data protection. These disparities tend to raise the compliance costs and make different markets inaccessible to different insurers working in different jurisdictions. The objective of the current research was to determine the key regulatory impediments to intercountry insurance transactions, and also to test how the new global regulations, like the Insurance Capital Standard (ICS), Solvency II, and International Financial Reporting Standards (IFRS) 17, could be effective in facilitating regulatory convergence. It used a qualitative, doctrinal, and comparative research design to look at statutory frameworks, supervisory guidelines, international standards and judicial interpretations in major jurisdictions, such as the European Union, the United States, the United Kingdom, Singapore and some of the emerging markets. It was found that there were still existing gaps in the adoption and enforcement of new frameworks, with wide dispersion in standards of solvency, financial reporting practices, data protection and dispute resolution procedures. Compared jurisdictions like the EU and Singapore also showed a relatively higher alignment, but the United States and most of the developing markets had lower harmonization rates.

Cloud computing has grown due to internet advancements in recent years. The Internet, Cloud of Things (IoT, CoT), edge, and fog computing are integrated in modern cloud architecture. These devices generated a lot of client-requested data that must be handled quickly. Mobile Edge Computing (MEC) devices offloaded smartphone tasks to nearby edge servers to speed up execution. Task offloading has become increasingly crucial for accelerating execution, minimizing latency, and conserving energy. Task execution and efficient resource use enhance the Quality of Service (QoS). So, researchers have proposed heuristic and rule-based dynamic offloading policies. However, they typically fail to balance exploring multiple solution spaces and using high-quality solutions, resulting in imbalanced task allocation convergence. In this paper, a machine learningbased method is proposed that uses the Random Forest (RF) classifier to choose Differential Evolution (DE) and Particle Swarm Optimization (PSO). The classifier was trained using several metrics derived from different job sizes. It used the best possible metaheuristic method for the incoming job size. The proposed methodology yielded promising results when compared to several current strategies.

The transition towards autonomous electric vehicles offers a sustainable and eco-friendly solution to pollution but introduces the possibility of cyber threats due to their intense reliance on interconnected systems. This paper proposes an integrated security system ensuring confidentiality, integrity, and authentication for vehicle-to-vehicle(V2V) communication. It employs a combination of Falcon and Kyber two post-quantum cryptographic algorithms to maintain Autonomous Electric Vehicles(EVs) authentication and integrity respectively. Both of these algorithms are resistant to quantum attacks and viable for IoT devices. In addition, a Graph Neural Network-based intrusion detection system is introduced in another layer to monitor charging stations and identify network attacks. Promising experimental outcomes concerning communication rate, communication time, energy consumption, throughput, and packet delivery ratio of the Falcon and Kyber algorithm witnessed effective V2V communication. Similarly, the graph neural network model achieves a 98.7% accuracy while maintaining the overall Quality of service and marking a significant step in secure communication for autonomous EVs.