arXiv subjects

Wayne Goodridge

Publications and source records attributed to Wayne Goodridge.

5 recordsLinked to original sources

arXiv2026

Design and empirical validation of a stock-Android software architecture for Wi-Fi Direct multi-group communication

Context: Stock Android exposes Wi-Fi Direct peer-to-peer APIs, but it does not provide application-transparent communication across multiple Wi-Fi Direct groups. For developers working on non-rooted devices, the main obstacle is architectural: interface-specific transport contexts, relay roles, and forwarding state must be coordinated entirely at application level. Objectives: This paper investigates whether multi-group Wi-Fi Direct communication can be realized as a stock-Android software architecture while preserving forwarding-state consistency and remaining compatible with Android 11 devices without rooting or operating-system modification. Methods: We design SWARNET, a layered artifact composed of a Flutter application layer, a Kotlin native networking layer, interface-bound P2P and legacy-Wi-Fi sockets, relay-state management, and subscription-based forwarding tables. We evaluate the implemented artifact on five stock Samsung Galaxy A10s smartphones across four single-group and multi-group scenarios using archived throughput and packet-loss measurements. Results: The artifact remained operational in all four scenarios. Peak receiver throughput observed from the archived curves was approximately 19.7~Mbit/s in 2d1g, 17.9~Mbit/s in 3d1g, 16.1~Mbit/s in 4d2g, and 16.0~Mbit/s in 5d3g. Packet loss increased with forwarding complexity, reaching about 19--20\% only in the highest-load region of the three-group case. Conclusion: The contribution is an implementable software architecture and a feasibility study showing that stock-Android multi-group Wi-Fi Direct communication can be engineered at application level on non-rooted devices. The results support architectural feasibility in a small static testbed; they do not establish broad resilience, scalability, or deployment readiness.

cs.NI↗

arXiv2024

Learning Sign Language Representation using CNN LSTM, 3DCNN, CNN RNN LSTM and CCN TD

Existing Sign Language Learning applications focus on the demonstration of the sign in the hope that the student will copy a sign correctly. In these cases, only a teacher can confirm that the sign was completed correctly, by reviewing a video captured manually. Sign Language Translation is a widely explored field in visual recognition. This paper seeks to explore the algorithms that will allow for real-time, video sign translation, and grading of sign language accuracy for new sign language users. This required algorithms capable of recognizing and processing spatial and temporal features. The aim of this paper is to evaluate and identify the best neural network algorithm that can facilitate a sign language tuition system of this nature. Modern popular algorithms including CNN and 3DCNN are compared on a dataset not yet explored, Trinidad and Tobago Sign Language as well as an American Sign Language dataset. The 3DCNN algorithm was found to be the best performing neural network algorithm from these systems with 91% accuracy in the TTSL dataset and 83% accuracy in the ASL dataset.

cs.LG↗

arXiv2024

Recovering WPA-3 Network Password by Bypassing the Simultaneous Authentication of Equals Handshake using Social Engineering Captive Portal

Wi-Fi Protected Access 3 (WPA3) is the accepted standard for next generation wireless security. WPA3 comes with exciting new features that allows for increased security of Wi-Fi networks. One such feature is the Simultaneous Authentication of Equals (SAE) which is a protocol whereby passphrases are hashed using a Password Authenticated Key Exchange with keys from both the Access Point and the Client making the password resistant to offline dictionary attacks. (Harkins, Dan. 2019) This objective of this research paper seeks to bypass WPA3-SAE to acquire network password via Man-in-the-Middle attack and Social Engineering. This method can prove to be useful given that majority of network attacks stem from social engineering. For this research we would be looking at the security of WPA3 personal transition mode and capture the network password via a captive portal. Breaching the WPA3 network can be possible by building on various security flaws that was disclosed on WPA3 in 2021. Due to the discovery of Dragonblood downgrade attacks disclosed in 2019, identified that WPA2/3Handshakes could be acquired. A Man in the Middle attack proposed set up is carried out by using race conditions to deauthentication WPA3 network and then using a Raspberry Pi to spawn a rouge WPA3 network. As such, the handshake acquired can then be utilized as to verify the password that would be entered in the captive portal of the rouge WPA3 network. This research identified that the Password was able to be recovered from Social Engineering Captive Portal when Protected Management Frames are not implemented. This research also indicates that some devices are not able to connect to a WPA 3 transition network which is contradicting the Wi-Fi Alliance claim that it is backwards compatible with WPA2.

cs.CR↗

arXiv2024

Swarm Intelligence-Driven Client Selection for Federated Learning in Cybersecurity applications

This study addresses a critical gap in the literature regarding the use of Swarm Intelligence Optimization (SI) algorithms for client selection in Federated Learning (FL), with a focus on cybersecurity applications. Existing research primarily explores optimization techniques for centralized machine learning, leaving the unique challenges of client diveristy, non-IID data distributions, and adversarial noise in decentralized FL largely unexamined. To bridge this gap, we evaluate nine SI algorithms-Grey Wolf Optimization (GWO), Particle Swarm Optimization (PSO), Cuckoo Search, Bat Algorithm, Bee Colony, Ant Colony Optimization, Fish Swarm, Glow Worm, and Intelligent Water Droplet-across four experimental scenarios: fixed client participation, dynamic participation patterns, hetergeneous non-IID data distributions, and adversarial noise conditions. Results indicate that GWO exhibits superior adaptability and robustness, achieving the highest accuracy, recall and F1-scoress across all configurations, while PSO and Cuckoo Search also demonstrate strong performance. These findings underscore the potential of SI algorithms to address decentralized and adversarial FL challenges, offereing scalable and resilient solutions for cybersecurity applications, including intrusion detection in IoT and large-scale networks.

cs.LG↗

arXiv2013

Using Bias Optimization for Reversible Data Hiding Using Image Interpolation

In this paper, we propose a reversible data hiding method in the spatial domain for compressed grayscale images. The proposed method embeds secret bits into a compressed thumbnail of the original image by using a novel interpolation method and the Neighbour Mean Interpolation (NMI) technique as scaling up to the original image occurs. Experimental results presented in this paper show that the proposed method has significantly improved embedding capacities over the approach proposed by Jung and Yoo.

cs.MM↗