About

I am a researcher at the AIR Institute in Salamanca, working on machine learning for network traffic monitoring and analysis. I completed my PhD in Computer Engineering at the University of Salamanca in July 2026, advised by Juan M. Corchado and Javier Prieto, as a Marie Skłodowska-Curie fellow in the EU Horizon 2020 IoTalentum training network. The thesis received sobresaliente cum laude with the International Doctorate mention.

My work covers knowledge distillation, ensemble methods and self-supervised representation learning for traffic classification and anomaly detection, and consensus protocols for IoT devices that cannot afford proof of work. More recently I work on language-model agents over live network measurement and on neuro-symbolic architectures for intent-based networking, where a symbolic layer constrains what the model is permitted to assert.

  • Network traffic classification
  • Anomaly & intrusion detection
  • Knowledge distillation
  • Self-supervised learning
  • Intent-based networking
  • LLM agents for observability
  • Blockchain for IoT

News

  • Defended my PhD thesis at the University of Salamanca: sobresaliente cum laude, International Doctorate mention.
  • Joined the AIR Institute, Salamanca, as a researcher.
  • Paper on autoencoding-based self-supervised learning for traffic representation accepted in the IEEE Open Journal of the Communications Society.
  • Paper on self-supervised traffic representation published at IEEE ICMLCN 2025.

Publications

Journal articles

  • OJ-COMS2026Autoencoding-based self-supervised learning for enhanced representation of network traffic patternsM. Abbasi, A. Shahraki, M. Plaza-Hernández, J. Prieto, J. M. CorchadoIEEE Open Journal of the Communications SocietyAccepted
  • Access2025Class imbalance in network traffic classification: an adaptive weight ensemble-of-ensemble learning methodM. Abbasi, S. López Flórez, A. Shahraki, A. Taherkordi, J. Prieto, J. M. CorchadoIEEE Access, vol. 13, pp. 26171–26192DOI
  • TMLCN2024Unleashing the potential of knowledge distillation for IoT traffic classificationM. Abbasi, A. Shahraki, J. Prieto, A. González Arrieta, J. M. CorchadoIEEE Transactions on Machine Learning in Communications and NetworkingDOI
  • EAI IoT2024Proof-of-resource: a resource-efficient consensus mechanism for IoT devices in blockchain networksM. Abbasi, J. Prieto, M. Plaza-Hernández, J. M. CorchadoEAI Endorsed Transactions on Internet of Things, vol. 10DOI
  • IoT2023Industrial data monetization: a blockchain-based industrial IoT data trading systemM. Abbasi, J. Prieto, A. Shahraki, J. M. CorchadoInternet of Things, ElsevierDOICode
  • Access2022Security in the Internet of Things application layer: requirements, threats and solutionsM. Abbasi, M. Plaza-Hernández, J. Prieto, J. M. CorchadoIEEE Access, vol. 10, pp. 97197–97216DOI
  • ComNet2022A comparative study on online machine learning techniques for network traffic streams analysisA. Shahraki, M. Abbasi, A. Taherkordi, A. D. JurcutComputer Networks, Elsevier, vol. 207DOI
  • TCCN2021Active learning for network traffic classification: a technical studyA. Shahraki, M. Abbasi, A. Taherkordi, A. D. JurcutIEEE Transactions on Cognitive Communications and NetworkingDOI
  • ComCom2021Deep learning for network traffic monitoring and analysis (NTMA): a surveyM. Abbasi, A. Shahraki, A. TaherkordiComputer Communications, Elsevier, vol. 170, pp. 19–41DOI
  • EAAI2021Deep reinforcement learning for QoS provisioning at the MAC layer: a surveyM. Abbasi, A. Shahraki, M. J. Piran, A. TaherkordiEngineering Applications of Artificial Intelligence, vol. 102DOI
  • EAAI2020Boosting algorithms for network intrusion detection: a comparative evaluation of Real AdaBoost, Gentle AdaBoost and Modest AdaBoostA. Shahraki, M. Abbasi, Ø. HaugenEngineering Applications of Artificial IntelligenceDOI

Conference papers

  • ICMLCN2025Autoencoding-based self-supervised learning for enhanced representation of network traffic patternsM. Abbasi, J. Prieto, J. M. CorchadoIEEE International Conference on Machine Learning for Communication and Networking
  • SSCT2023A novel aging-based proof of stake consensus mechanismM. Abbasi, J. Prieto, M. Plaza-Hernández, J. M. CorchadoSustainable Smart Cities and Territories, LNNS, SpringerDOI
  • SMARTCOMP2022FLITC: a novel federated learning-based method for IoT traffic classificationM. Abbasi, A. Shahraki, A. TaherkordiIEEE International Conference on Smart ComputingDOI
  • FlexNets2021Internet traffic classification using an ensemble of deep convolutional neural networksA. Shahraki, M. Abbasi, A. Taherkordi, M. KaosarACM SIGCOMM Workshop on Flexible NetworksDOI
  • IICITA2019Internet of Things in agriculture: a surveyM. Abbasi, M. H. Yaghmaee, F. Rahnama3rd International Conference on Internet of Things and Applications (IoT), IEEEDOIPDF
Full list on Google Scholar →

Software

agentic-observability

An LLM-based network monitor that reports the limits of its own instrument. Before it calls a change real, it measures the path’s noise floor from its own history and asks whether a change of that size could have been observed at all, so “no change” is never confused with “could not see”. Every claim the model makes is verified in code against the same statistics the monitor alerts on.

Instrument floor
0.1 ms steps → <5.0% not expressible
Statistical floor
75% at 80% power
Can it detect +10%?
No. “No change” would mean invisible
All repositories on GitHub →

Experience & education

  • 2026 – nowResearcherAIR Institute, Salamanca, Spain
  • 2022 – 2026PhD in Computer EngineeringUniversity of Salamanca. Sobresaliente cum laude, International Doctorate mention
  • 2022 – 2025Marie Skłodowska-Curie FellowIoTalentum, University of Salamanca. EU H2020, grant 953442
  • 2024 – 2025Secondment: energy market modellingEnergia Simples, Portugal
  • 2024Secondment: blockchain for peer-to-peer energy tradingInstituto Superior de Engenharia do Porto, Portugal
  • 2022 – 2023Secondment: proof-of-work trade-offsDigitel TS, Valladolid, Spain

Service

Reviewer for IEEE Communications Letters and Engineering Applications of Artificial Intelligence.