Indian Institute of Information Technology, Pune

In digital forensic investigations, most modern storage media use the NTFS file system. This means that NTFS-level analysis is necessary for a thorough examination of evidence. In real-world forensic situations, user files are often deleted, partially overwritten, or damaged. However, NTFS internal system artifacts often keep important traces of filesystem activity. Artifacts like $MFT, $LogFile, $UsnJrnl, $Bitmap, and other NTFS metadata files keep structured and log-based evidence safe even when there are no file names, directory entries, or higher-level filesystem context. This paper presents NTFS-LogID, a machine learning-based framework for the identification and classification of NTFS internal filesystem artifacts directly from raw binary data. The suggested method divides NTFS artifacts into binary chunks of a certain size and then extracts a full set of features, including byte-level statistical descriptors, structural ratio features, NTFS-specific signature indicators, and structureaware characteristics. After that, supervised learning models are trained on these representations to tell different NTFS internal artifact classes apart. Experimental results demonstrate that NTFS-LogID accurately identifies NTFS internal artifacts under metadata-constrained conditions, highlighting its effectiveness and practical applicability in real-world digital forensic investigations where only raw disk data is available.

Traditionally, the file sharing was primarily facilitated through email. However, with recent development, modern instant messaging applications and their web versions have become the predominant means for sharing and downloading files of various formats. This shift has also corresponded to a notable increase in cybercrime, highlighting the critical importance of storing and analyzing digital evidence. Given the widespread use of the Windows operating system and its supported NTFS file system, understanding and interpreting its artifacts has become pivotal. The study includes the remnants for five web-based chat applications (WhatsApp Web, Facebook Messenger, Telegram, Slack, and Discord) accessed via popular web browsers, namely Mozilla Firefox, Microsoft Edge, Google Chrome, Opera, and Brave installed on Windows 10 desktop operating system. These essential components have provided the focal point for our comparative analysis of the MFT entries associated with different file formats from diverse instant messaging applications operating on distinct web browsers.

An efficient and energy-aware arithmetic circuit is crucial in modern digital systems, especially in signal processing, cryptography, and embedded computing. Multipliers, a key component of arithmetic logic units (ALUs), directly impact computational speed and power efficiency. Conventional multipliers based on irreversible logic lead to excessive power dissipation due to information loss. This paper addresses the above challenges by introducing three optimized reversible vedic multiplier designs that integrate BME, improved Peres, and hybrid gates to enhance performance. These architectures effectively reduce quantum cost (QC), garbage outputs (GO), and total reversible logic implementation cost (TRLIC), improving power efficiency. The designs are implemented and synthesized using Cadence 180 nm EDA tools and Xilinx ISE 14.7, ensuring feasibility for FPGA-based applications. The performance evaluation across 2× 2 , 4× 4 , 8× 8 , 16× 16 , and 32× 32 -bit multipliers demonstrates their scalability. The hybrid gate-based architecture is the most efficient choice for low-power, high-speed computing applications. It reduces the area by 14.1

The decentralized and pseudonymous nature of cryptocurrency has facilitated its extensive use in illicit activities, including money laundering, tax evasion, and ransomware. Limiting such activities requires a well-established forensic framework. However, a dedicated methodology for examining cryptocurrency wallets remains underdeveloped. This study presents a systematic forensic analysis of Electrum wallets installed on virtual machines running Windows 10, outlining the wallet taxonomy and meticulously listing all artifacts. This study primarily focuses on memory forensics, with most of the analysis devoted to memory-based artifacts extracted from five distinct memory dump scenarios. Artifacts extraction were performed using Volatility 3 plugins, in conjunction with Python-based analysis scripts, within a Kali Linux environment. Following the memory-based analysis, a limited disk examination was conducted after wallet inactivity or system shutdown to assess whether any residual Electrum artifacts persisted beyond memory. The research examines the artifacts retrievable from wallet files, both before and after backup, and compares these results with those obtained from other methods reported in the literature. The experimental outcomes demonstrate the impact of this research on the successful extraction of private keys, wallet addresses, extended public keys, wallet files, and transaction IDs. The extracted Electrum addresses and private keys provided access to critical wallet details, and unspent Bitcoin were successfully recovered using these keys, confirming the feasibility of forensic cryptocurrency recovery and revealing data of high evidentiary value to the digital forensic community.

The rapid rise of industrial automation has accelerated the deployment of autonomous mobile robots for material handling, inspection, and collaborative operations. Effective performance in dynamic environments demands path-planning algorithms that ensure safety, energy efficiency, and adaptability—balancing global optimality with adaptive reactivity. No single algorithmic solution fully satisfies these requirements, necessitating hybrid frameworks that integrate the global optimization capability of metaheuristics with the adaptive learning of reinforcement learning. To address these challenges, we introduce RL-PFWOA, a novel hybrid hierarchical framework featuring bidirectional feedback between a metaheuristic core and a reinforcement learning agent. Global exploration is achieved through a hybrid Pufferfish optimization (PFO) and whale optimization algorithm (WOA) strategy, where the WOA coefficient A dynamically alternates between encircling and spiral exploitation phases. A DDPG agent performs path refinement and adaptively modulates exploration via hybrid weights w_PFO and w_WOA . A Bayesian weighting mechanism further balances path length, energy consumption, and traversal time, ensuring adaptive multi-objective optimization and preventing premature convergence. Benchmarking on ten standard functions (Sphere, Rastrigin, Schwefel, etc.) demonstrates superior convergence and minimal cost values compared to metaheuristic (PSO, GA, ABC, WOA) and reinforcement learning (PPO, SAC) baselines. MATLAB/Simulink 2023a simulations validate practical performance: RL-PFWOA yields the shortest paths (55.0 m static, 58.1 m dynamic), lowest energy use (9.21 J static, 9.87 J dynamic), and minimal obstacle collisions (1.5–2.1 SD = ± 0.35 ), a 56