Publications
Journal Articles
A Cyber-Physical-System Approach to Data Center Modeling and Control for Energy Efficiency
A cyber-physical systems approach to data center energy efficiency, coupling computational load management with thermal and cooling control through an integrated modelling and optimisation framework.
Book Chapters
Models and Control Strategies for Data Centers in the Smart Grid
Book chapter covering models and control strategies for operating data centers as flexible loads in the smart grid, exploiting demand-response capabilities to reduce costs and support grid stability.
Conference Papers
Adaptive Unicast-Multicast Strategies for Over-the-Air Updates in Multi-Receiver Wi-Fi Networks
Adaptive unicast/multicast switching strategies for reliable OTA data delivery to multiple Wi-Fi receivers under uncertain channel conditions, reducing transmissions by up to 75% over static strategies.
Online Map Validation for Autonomous Driving
Two complementary approaches for online HD map validation: a model-based framework and a deep similarity learning technique, enabling real-time detection of map inconsistencies from live sensor data.
Model Predictive Control of Data Centers in the Smart Grid Scenario
A cyber-physical approach to data center control as smart grid nodes, explicitly modelling computational and thermal dynamics to maximise financial returns from demand-response participation.
Dynamic power allocation in server farms: A Real Time Optimization approach
A real-time optimisation approach for dynamic power allocation across server farms, balancing computational performance against energy consumption under time-varying workloads and power caps.
A Hierarchical Approach to Energy Management in Data Centers
A hierarchical control architecture for data center energy management, decomposing the problem into fast server-level and slower thermal-level layers to enable scalable coordinated optimisation.
A cyber-physical systems approach to energy management in data centers
A cyber-physical systems framework for data center energy management that explicitly models the coupling between computation and thermal dynamics, enabling coordinated server and cooling control.
Workshop Papers
Siamese Networks for Online Map Validation in Autonomous Driving
A deep learning classifier based on Siamese Network architecture for validating HD map data against live sensor readings. Reaches an F1 score of 89.1%.
Correlation-based approach to online map validation
A model-based method for HD map validation using spatial and temporal correlation of smart sensors, fusing independent regional assessments to identify map areas inconsistent with live sensor data.
Benchmark problem: an air brake model for trains
A publicly available simplified hybrid model of a freight train air brake system, proposed as a benchmark for verification tools applied to cyber-physical systems with safety constraints.
A Unified Thermal-Computational Approach to Data Center Energy Management
A unified modelling approach that couples thermal and computational dynamics of data centers, enabling coordinated optimisation of cooling and server workloads for improved energy efficiency.
Reducing Data Center Energy Consumption via Coordinated Cooling and Load Management
Joint optimisation of cooling systems and workload placement in data centers, exploiting their thermal coupling to reduce total energy consumption compared to independently managed subsystems.
Thesis
Models and Control Strategies for Data Center Energy Efficiency
PhD thesis proposing a cyber-physical systems approach to data center energy management, integrating thermal and computational models under a unified MPC framework to minimise energy consumption while meeting SLA constraints.