Mouad Jaouhari, Julio Montecinos, and Lokman Sboui   
This work introduces a Hybrid Multi-hop Direct-to-satellite IoT (HMDI) architecture that achieves near real-time connectivity in remote areas by replacing absent inter-satellite links (ISLs) with ground-based IoT relays. The proposed 3Red-D algorithm combines three matrix reduction steps with a Dijkstra-based routing method, reducing computational complexity by 98% compared to brute-force while maintaining 100% accuracy in finding the lowest-delay path.

Nourhene Aloui, Sandra Flores-Mejia, and Lokman Sboui  
We developed an operational pipeline using satellite imagery and the Random Forest machine learning model to provide cost-effective and scalable detection of AMATU weed infestations in Quebec soybean fields. The pipeline delivers probability heatmaps to guide targeted agricultural interventions, supporting the reduction of herbicide use and promoting sustainable farming.

Benjamin Lepourtois, Adji Toure, Vanessa Ayotte, Sonia Seck, Judith Boissonneault, and Lokman Sboui
This project introduces AutoBib, an automated solution developed to streamline the generation of comprehensive bibliometric reports and analyze scientific collaborations. Developed with the ÉTS library, AutoBib integrates data extraction from Scopus and SciVal using Python and automates report formatting in MS Excel and MS Word, significantly reducing the time and labor required for bibliometric analysis. 

Mohamed Karaa, Raed Bahria, Hakim Ghazzai, and Lokman Sboui
We develop a semantic compression framework for low-bandwidth video surveillance, extracting abstract representations using state-of-the-art object detection, tracking and captioning models.

Karnika Biswas, Hakim Ghazzai, Abdullah Khanfor, and Lokman Sboui  
We propose a novel link allocation method to provide continuous in-flight connectivity for Urban Air Mobility (UAM) vehicles. The method dynamically switches between cost-efficient terrestrial cellular networks and Low Earth Orbit (LEO) satellite networks based on real-time metrics like signal strength, congestion, and flight trajectory. Numerical results confirm the algorithm minimizes data loss while ensuring the optimal selection of above-threshold data rates. 

Nourhane Sboui, Raed Bahria, Hakim Ghazzai, Sameh Najeh and Lokman Sboui
We introduce a comprehensive no-reference deep image quality assessment (NR-DIQA) framework designed to effectively manage LEO satellite images for Earth observation (EO) by identifying and filtering distorted and anomalous images before transmission.

Mohamed Karaa, Hakim Ghazzai, and Lokman Sboui
We present an image dataset of snow-covered urban roads captured by traffic cameras during the season. We discuss the dataset acquisition methodology, as well a benchmark problem for automating its annotation with snow levels, and its potential applications.

Mohamed Karaa, Hakim Ghazzai, and Lokman Sboui
An AI-based annotation system for a dataset of snow-covered road images into four snow level categories. The annotation enables the training of supervised classification models.

Sundos Mojahed, Rejean Drouin, Lokman Sboui
ODACE is a novel platform that automates the manual mobile phone certification process using Appium and ADP to verify mobile phone functions, significantly reducing the time and effort required for certification. 

Saad Abobakr, Mahmud Alosta, Mohamed Amine Abdelkefi, Amine El Kaouachi, Lokman Sboui
ODACE is a novel platform that automates the manual mobile phone certification process using Appium and ADP to verify mobile phone functions, significantly reducing the time and effort required for certification. 

Allafi Omran, Mohamed Cheriet, Lokman Sboui
A novel scheme to enhance network resilience and efficiency in post-disaster and crowded cellular networks by integrating unmanned aerial vehicles (UAVs) and satellites.

Lokman Sboui, Hakim Ghazzai, Zouheir Rezki, Mohamed-Slim Alouini
A study of the achievable rate of a 5G scenario where a UAV relay is extending the wireless network and is serving both primary and secondary users in a cognitive radio framework.

Advisor: Prof. Mohamed-Slim Alouini, King Abdullah University of Science and Technology (KAUST)
My doctoral research addressed the exponential increase in global data traffic through the development of 5G frameworks and cognitive radio (CR) concepts. The study aimed to mitigate spectrum scarcity while enhancing performance in terms of reliability, scalability, and energy efficiency.