A5037736742- Green IT and Sustainability
- Smart Grid Energy Management
- Embedded Systems Design Techniques
- Simulation Techniques and Applications
- Advanced Data Storage Technologies
- Real-Time Systems Scheduling
- Energy Harvesting in Wireless Networks
- Parallel Computing and Optimization Techniques
- Biometric Identification and Security
- Brain Tumor Detection and Classification
- Medical Image Segmentation Techniques
- Caching and Content Delivery
- Chaos-based Image/Signal Encryption
- User Authentication and Security Systems
- Distributed and Parallel Computing Systems
St Petersburg University
2017-2023
Extracting various valuable medical information from head MRI and CT series is one of the most important challenging tasks in area image analysis. Due to lack automation for many these tasks, they require meticulous preprocessing experts. Nevertheless, some problems may have semi-automatic solutions, but are still dependent on person's competence. The main goal our research project create an instrument that maximizes processing degree. Our consists two parts: a set algorithms tools its...
Simultaneous perturbation stochastic approximation was previously shown to provide good results for optimizing resource consumption under dynamically changing workload. We employ this approach develop dynamic voltage-frequency scaling (DVFS) governor Android OS. design experimental methodology and show our algorithm performance compared common DVFS governors. A link the source code is also provided.
It was previously shown that simultaneous perturbation stochastic approximation can be used for developing general purpose dynamic voltage-frequency scaling (DVFS) governor Android OS. We lay down a theoretical foundation this approach and show in several experiments our algorithm performance is on par with the commonly DVFS governors. Links to complete results source code are also provided.
Simultaneous perturbation stochastic approximation demonstrate good results for control theory problems where system states dynamically changes. In this paper a dynamic voltage frequency scaling governor creation based on approach smartphone CPU under Android OS is described and ways to improve it are considered. The new shows comparable default governors
The power consumption of mobile devices is a hot topic these days, and it important to address when developing applications. One the most popular ways measure accessing internal sensors using Android Debug Bridge (ADB). We discovered that measurement frequency may skew readings. Based on this approach we propose our own algorithm for calculating smartphone energy constants — in milliamperes at nominal voltage different peripherals states. Our takes bias into account, its results are compared...
Extending smartphone working time is an ongoing endeavour becoming more and important with each passing year. It could be achieved by advanced hardware or introducing energy-aware practices to software, the latter a accessible approach. As CPU one of most power-hungry devices, Dynamic Voltage Frequency Scaling (DVFS) technique adjust frequency current computational needs, different algorithms were already developed, both energy-agnostic kinds. Following our previous work on subject, we...
Biometrie system users' classification based on their recognition quality is an important issue when developing and exploiting such systems. Existing approaches a biometric menagerie concept are described limitations shown. A new using errors proposed its feasibility demonstrated.