Research Projects

The 6G-XCEL project is a cutting-edge research initiative focusing on the integration and advancement of Artificial Intelligence (AI) within future 6G networks. It aims to tackle the challenges of utilizing AI in communication networks by developing a decentralized, AI-based network control framework across different network domains and physical layers. This framework will enhance security and sustainability in network operations, employing multi-party AI controls over compute accelerators to manage both radio and optical networks efficiently. By leveraging the latest AI algorithms and data compression techniques, 6G-XCEL will create reference use cases, data and model repositories, and a benchmarking platform for future AI/ML solutions in 6G networks. The project involves a significant collaboration between researchers in the EU and the US, aiming for global validation, adoption, and standardization of AI approaches in 6G networks.

SoftEdge is an SFI funded project that leverages SDN/NFV technologies to create fully-programmable intelligent wireless systems, and it also pushes services closer to demand by harnessing resources at the network edge. The intelligent control relies on the use of machine learning algorithms in the control plane to dynamically adjust the network operation to the environment and user demand. Additionally, the project studies the economics of edge systems and designs virtualization and pricing mechanisms that will optimize efficiency and performance. My involvement as the lead scientist ranges from project management, reporting, performing research activities to supervising other researchers working on these topics.

The project focuses on the use of machine learning to optimize the prediction, detection and respective interventions of rare events in industrial processes based on large data acquisition, big data fusion and big data analytics. The proposed general framework relies on an IoT network and aims at ultra-reliable detection / prevention of rare events related to a predetermined industrial physical process (modeled by a particular signal). The project consortium consisted of research teams from all around Europe. I was leading the research activities from the Trinity College Dublin side, and at the same time I was in charge of coordinating the research activities between our research group and the rest of the consortium. The work also involved project documentation writing, project management tasks and timely delivery of project milestones.

CreativeConnections is a startup project based on my work on machine learning that I co founded with Dr. Irene Macaluso. The startup promises to simplify the interaction between small business owners and online influencers for the purpose of starting an online marketing campaign. The work on this startup involved the whole business development cycle (i.e., problem definition, customer validation, building customer personas, value proposition, market analysis, competitor analysis, building a team and execution).

NearBy is a startup project that I co founded with Dr. Irene Macaluso. The startup was based on a patent that Dr. Macaluso and I filed together, and it takes advantage of distributed intelligence concepts to optimize the operation of a communication network for mobility purposes. The work on this startup involved the whole business development cycle (i.e., problem definition, customer validation, building customer personas, value proposition, market analysis, competitor analysis, building a team and execution). The project was successful and we got full funding to perform in-depth market research.

This project was funded by the US Air Force and its goal was to design and develop networks with degeneracy in mind. The self-healing and resilience of networks emerged as a result of the studies undertaken in this project. I was in charge of the scientific research on this project.

The goal was to implement a Software as Service (SaaS) based multi-functional and multi-vendor platform for administration and configuration of various radio access networks. The platform should aggregate various existing OSS tools and interfaces, thereby providing a unique interface for configuration and analysis of multi vendor networks. Because of the dynamically changing environment, all functional parts of the platform are organized as independent services. This implies the use of SaaS models, which should enable high scalability, flexibility and make the future customization and changes easier. I was in charge of the design, testing and development of the system.

The goal was to develop an application that includes a customized Porteus Linux distribution with a range of tools that enable smart (automatic) generation of configuration files, and configuration of Ericssons TCU / SIU (Transportation Connectivity Unit / Site Integration Unit) devices. The application simplifies the configuration for assembly teams within BH Mobile. I was in charge of the design, testing and development of the system.

As part of an expert team, I was in charge of evaluating the tender for procurement of digital microwave links for the public radio-television services of Bosnia and Herzegovina. The project required extensive telecommunication knowledge and required detailed documentation to be submitted for the purpose of evaluating the project plan.

I was in charge of designing the telecom network for the longest tunnel on the Koridor 5C highway in Bosnia and Herzegovina. The work on this project involved network design, project management, supervision of the project execution under difficult circumstances (tunnel building), communication with multiple stakeholders (including higher management, workers in the tunnel, safety officers).

As part of an expert team, I was in charge of defining the minimal requirements for DVB-T receivers in Bosnia and Herzegovina. The project required extensive telecommunication knowledge and required detailed documentation to be submitted for the purpose of evaluating the project plan.

This project was funded by Iskratel Kranj. I was in charge of the implementation of a VoIP application for Iskratel’s InnBox F68 network device. The work involved extensive c++ knowledge and cross-compilation from a 32bit Windows to a 64bit Linux based operating system.

I was part of the team that built the first prototype of the Ericsson Mobile Health System. My work included the definition, development and testing of various features ranging from health measurements (e.g. blood pressure, pulse, heart rate) to database design and data validation.