1. SATORI (2019-?)

I am a team member of the SATORI (smart networking in the era of AI) project, a joint research collaboration between Trinity College Dublin (Ireland) and Tsinghua University (China). For more details, visit here.

[1] Gomes, A., Kibiłda, J., Farhang, A., Farrell, R., and DaSilva, L. A (2020). Network Sharing for Reliable Networks: A Data-Driven Study. In IEEE International Conference on Communication (ICC). PDF.

2. SOMAC (2017-2018)

SOMAC (Self-Organizing MAC sub-layer) is an extension of FS-MAC+ and also the outcome of my Master’s thesis. It copes with the non-stationary nature of wireless networks and does not rely on training sets beforehand (shortcomings of FS-MAC+), being able to learn how to switch the MAC protocol on the fly. SOMAC is also implemented using GNU-Radio and tested on a software-defined radio testbed.

[1] Gomes, A., Macedo, D. F. and Vieira, L. F. M. (2019). Automatic MAC protocol selection in wireless networks based on reinforcement learning. In ELSEVIER Journal of Computer Communications. Paper URL. PDF. Source code.

3. FS-MAC+ (2017-2018)

FS-MAC+ is a Machine Learning-based MAC sub-layer for wireless networks. It automatically selects and switches the MAC protocol in use, improving the network performance over time. FS-MAC+ is implemented using GNU-Radio and tested on a software-defined radio testbed of Ettus USRP B200 devices.

[1] Gonçalves, A. V. G. S., Macedo, D. F. and Vieira, L. F. M. (2018). FS-MAC+: Troca flexível do protocolo de acesso ao meio em redes sem fio empregando aprendizado de máquina. In XXXVI Brazilian Symposium on Computer Networks and Distributed Systems (SBRC). Paper URL. PDF - PT. Source code.

4. MAC protocols for SDR (2017-2018)

I had developed this project during the course “Intelligent Networks” at the Federal University of Minas Gerais in 2017.1, consisting of two MAC protocols (CSMA/CA and TDMA) for use in software-defined radio platforms. The PHY layer comes from the project gr-ieee802-11. Source code.

5. Alternative algorithm for ANR in LTE networks (2016)

This project is the outcome of my Bachelor’s degree thesis. It proposes an algorithm that efficiently and automatically fulfils ANR (Automatic Neighbour Relation) tables in LTE networks. Document URL (unpublished).