Tag Archives: signal processing

Performance optimization of energy-efficient computer and communication systems


Project:


TR32023; 2011-2014

Leader:


dr Miroslav Lutovac

Description of research


Project consists of the development of new methods and algorithms, derivation of theoretical results, experimental verification, and development of systems for more efficient usage of computer and communication systems, with the focus on performance and to consume less energy. The topic belongs to Green technologies, and the focus is on consuming less energy or to reduce contamination and waste materials of computer and communication systems. Research is on efficient algorithms for signal and data processing, mathematical modeling, efficient spectral usage, optimization of computer systems, and usage of reconfigurable systems. The results will be used in communication systems, wireless communication, image and speech processing, and business decisions. Commercial software solutions will be produced, new or as upgrade of existing, or as part of equipments and systems produced by our participants.

Keywords


Signal processing, computer systems, mathematical modelling.

Project description


The expected results of this project are (1) the selection of modules and smaller units as parts of larger systems, (2) new methods, algorithms, and software solutions based on recent achievements of researchers employed by this project (modeling of analog passive systems implemented using digital systems, digital filters, nonlinear time-discrete systems). A part of team will work on theoretical-experimental research activities in the field of wireless communications. The main topic of research will be also on efficient reconfigurable filters, optimization algorithms, and implementation of efficient filters.
Another research topic is application of data selection and compression for building energy-efficient communication and information systems. This topic is a new area in signal processing with a wide possibility in building energy-efficient communication and information systems. The basic idea is on non-uniform sampling, with a limited number of data in such a way that it has all accurate relevant information. The key issue is to derive new efficient algorithms so that the minimal numbers of data contain the whole information without loss. Mathematical methods used in the research will cover liner data-base wavelets and L1 optimization.

Expected key results


Expected key results are in 5 sub-projects running 4 years in parallel.
Goal 1: Optimization of energy-efficient (EE) computer networks. Development of an Advanced Resource Utilization Metering System. Development of a Computer Network Simulation System. Research and development of design methods for EE Computer Network Architecture. Computer Network Protocol Stack Optimization.
Goal 2 and 4: Development of new methods, algorithms, and software solutions targeted on design and implementation of digital filters and non-linear systems, based on EE structures. Main topics are on reconfigurable hardware digital filters and the corresponding new efficient algorithms for the design of polynomial filters, reconfigurable band-pass digital filters, methods for optimization of multi-level decimation. Technical solutions in form of algorithms, methods and software products are planned. New date processing issue in medicine, image processing, sensor networks, understanding natural phenomena, business model based on this theory.
Goal 3: Development of mathematical method for EE spectra of mobile and wireless systems.
Goal 5: Results for managing various redundancies in sequent calculi proof search.

Participants from Singidunum University:


prof. dr Miroslav Lutovac