V
Initializing
Papers Dataset Projects Profiles
VIXERO TECHNOLOGY ENTERPRISE · RESEARCH DIVISION

We don't read bytes.
We measure the physics of evasion.

Independent research at the intersection of adversarial machine learning, non-Euclidean deep learning, and encrypted network defense. Operating from Kuching, Sarawak. Building toward thermodynamic, payload-independent perimeter security.

Principal Investigator Vickson Ferrel
Affiliation UNIMAS FCSIT
Status NVIDIA Inception Member
Established 2026
/ 01 — Publications

Papers & Preprints

Peer-reviewed work and active manuscripts
from the Vixero research pipeline.
Published April 3, 2026 arXiv preprint
cs.CR / cs.LG
CC BY 4.0
arXiv:2604.02149

AEGIS: Adversarial Entropy-Guided Immune System

Thermodynamic State Space Models for Zero-Day Network Evasion Detection

As TLS 1.3 limits the effectiveness of traditional Deep Packet Inspection, content-reading classifiers like ET-BERT remain structurally vulnerable to byte-level adversarial morphing. We introduce a Thermodynamic Variance-Guided Hyperbolic Liquid State Space Model (TVD-HL-SSM) that discards payload bytes entirely in favor of 6-dimensional continuous-time flow physics. By projecting temporal features into a non-Euclidean Poincaré manifold and measuring sequence-wide Shannon entropy, AEGIS detects adversarial tunneling without ever reading what's inside the packet.

0 F1 Score
0 TPR
0 FPR
0 Inference (RTX 4090)
0 Theoretical Ceiling
01
// IN DEVELOPMENT //
⊘ Embargoed 2026 — TBD Target venue:
peer-reviewed, top-tier
security conference
PROJECT // PELT

Successor Architecture to AEGIS

Active research · architecture details under embargo

Building on the empirical limits identified in AEGIS, PELT addresses the              failure mode through                   and a fully adaptive             mechanism. Preliminary internal results show                against the adversarial corpus that broke V3. Full architecture, ablations, and benchmarks will be disclosed at publication.

// No paper ships until it survives an adversarial gauntlet harder
// than the one that forged its predecessor.

02
Available April 11, 2026 Vixero whitepaper
WP-SEB series
Open access
WHITEPAPER // WP-SEB-001

Beyond the Grid

Cybersecurity Vulnerabilities in Legacy Fossil Fuel Infrastructure and the Case for Sustainable Transition

A grounded analysis of the security debt accumulating across legacy energy infrastructure: aging SCADA systems, unsegmented OT networks, and the structural incentives that have kept critical systems exposed for decades. Argues that the cybersecurity case for accelerating the energy transition is materially stronger than its economic case alone — and lays out the threat landscape with citations rather than vendor-naming.

03
Available April 11, 2026 Vixero whitepaper
WP-EDU series
Open access
WHITEPAPER // WP-EDU-001

Digital Divide Within the Campus

Network Infrastructure Failures and Their Impact on Academic Continuity in Malaysian Public Universities

A field-grounded examination of how chronic network infrastructure failures across Malaysian public universities erode academic continuity, research productivity, and student outcomes. Documents the everyday cost of unreliable campus connectivity — from interrupted coursework to stalled research pipelines — and argues that institutional digital infrastructure should be treated as core academic infrastructure, not as a peripheral utility. A grounded case for policy attention from someone writing it from inside the problem.

04
/ 02 — Open Science

The Corpus

AEGIS-Adversarial-Corpus
Public release · gated for verification
// Vix0007/AEGIS-Adversarial-Corpus

A 4-tier corpus, distilled from 400GB of raw network reality.

Trans-Pacific backbone captures, IoT botnet swarms, SHA256-verified zero-days, and proprietary cryptographic evasion traffic — fused into 908,037 sequences of 6-dimensional flow physics. Payload bytes are deliberately discarded during preprocessing, preserving privacy by construction.

Released as 10GB of PyTorch tensors at a 40:1 compression ratio over the source PCAPs. Sufficient to independently reproduce the thermodynamic variance metrics reported in the AEGIS paper, without reproducing the underlying evasion tooling.

🔓 OPEN SCIENCE · Hosted on Hugging Face · Anyone can request access
⌬ Request Access on Hugging Face
0
Sequences
0
Tensors (.pt)
0
Raw PCAP
0
Threat Tiers
TIER I
Planetary Baselines
MAWI WIDE · CIC-IDS-2017 · DoHBrw — natural stochastic decay
TIER II
Automated Swarm
IoT-23 · CTU-13 — Mirai, Torii, Okiru beaconing
TIER III
Zero-Day Vault
MTA · BCCC-Mal-NetMem — memory-resident rootkits, APT C2
TIER IV
Cryptographic Evasion
VLESS Reality · GhostBear · AMOI morphing
/ 03 — Engineering

Active Projects

Production systems and research
infrastructure under active development.
Production · V3 Active
AEGIS
Adversarial Entropy-Guided Immune System

A line-rate, payload-independent network defense system. Operates on flow thermodynamics rather than packet contents. Currently targeting BlueField-3 DPU compilation via TensorRT for sub-millisecond inline deployment in next-generation data centers.

Mamba-3 Hyperbolic eBPF / XDP CUDA 13 TileLang JIT
Active · Research Tool
EDITH
Autonomous research and operations agent

A general-purpose agent stack supporting the full Vixero research and engineering pipeline — academic monitoring, infrastructure ops, blue-team triage, and red-team automation. Built on the OpenClaw framework. Designed with red teamers as the primary future audience: an autonomous operator that can be pointed at a target surface and asked the right questions.

Agentic Red Team Blue Team OpenClaw
/ 04 — Identifiers

Researcher Profiles

Verified academic and
engineering identifiers.