Downloads & Free Reading Options - Results

Description: Abstract The rapid evolution of cyber threats, coupled with the advent of quantum computing, necessitates a fundamental rethinking of web security architectures. This dissertation introduces Darkelf Browser, a pioneering high-security web browser explicitly designed for users operating in high-threat environments such as investigative journalists, political dissidents, security researchers, and penetration testers. Traditional browsers, even those emphasizing privacy like Tor, are increasingly vulnerable to sophisticated attacks, forensic recovery, and quantum cryptographic breaking. Darkelf integrates lattice-based post-quantum cryptography—specifically, Kyber768 and Kyber1024 algorithms—into core communication protocols to ensure resilience against quantum adversaries. Its architecture emphasizes ephemeral memory-resident session keys, preventing forensic recovery of session data. The browser employs a supervised machine learning pipeline utilizing the RandomForestClassifier to detect malicious JavaScript in real time, safeguarding users from drive-by malware and script-based exploits. Complementing cryptography and threat detection are robust anti-forensic strategies, including encrypted RAM logs, runtime detection of debugging and forensic tools, disabling system swap files, and storing cryptographic keys exclusively in volatile memory (e.g., TPM hardware or in-memory regions). These features collectively create a hardened, stealthy, and adaptive browser framework. This work critically compares Darkelf's capabilities with legacy anonymity systems like Tor, emphasizing improvements in quantum resilience, forensic resistance, and usability in high-risk contexts. Ethical considerations surrounding dual-use technologies are discussed, and the entire project is released under an LGPL license, promoting transparency and responsible deployment. License: GNU Lesser General Public License (LGPL) 3.0