Algorithmic problems are abundant in many areas of science and technology, eg, in CAD networking, computer graphics and animation, cartography, and in transport systems. Efficient solutions to these algorithmic problems are of vital importance for competitive application systems in these and similar domains. The overall aim of ALCOM II is to provide the efficient algorithms and date structures needed for these solutions.
The research is concerned with the design, analysis, and implementation of efficient computer algorithms which are needed to build competitive information processing systems.
As well as the production of the software deliverables ALF, LEDA, LUO, DSS, and PARIX, over 15 Doctor of Philosophy dissertations and over 130 research reports have been completed. Workshops have been organized on Algorithms: Implementation, Libraries, and Use and on Graph Drawing and an electronic newsletter has been established.
In the area of sequential algorithms a new randomized algorithm was discovered for linear programming whose running time is subexpontential in the number of variables and constraints and does not depend on the size of the coefficients of the constraint matrix. Previously, only algorithms (eg the Simplex method) with exponential worst case complexity were known.
In the area of parallel and distributed algorithms a new load balancing techniques was developed. The new technique has the unique feature that it performs well in practice and can be analysed theoretically. Previously, this combination seemed elusive. The new technique has helped to win the Vice championship in Computer Chess.
In the area of analysis of algorithms and complexity a long standing open problem was solved and the exact number of comparators needed to merge 2 sorted sequences of length n was determined. In the area of implementation the functionality of the LEDA platform was increased for combinatorical computing. The platform is intensively used by several hundred sites worldwide.
APPROACH AND METHODS
Continuing the ALCOM I effort, the research in ALCOM II concentrates on the design of sequential, parallel and distributed algorithms, the analysis of algorithms, complexity theory, and on implementation. The consortium's coherent algorithmic approach is based on the following general methods: randomisation and probabilistic design, identification of algorithmic modules, dynamic data structures, and methodology of complexity analysis.
The dissemination of results and know-how is achieved through open schools, publications, an electronic bulletin board, and the disribution of prototype implementations. Prototype implementations were already distributed to several hundred sites.
ALCOM's research provides algorithms which have the potential of becoming fundamental building blocks for future systems in computer-aided design, computer graphics, symbolic manipulation, robotics, parallel and distributed computing, and VLSI design. The prototype implementations ease the transfer of the project's results to these application domains. The experience ot the first year shows that this potential is real.
Topic(s)Data not available
Call for proposalData not available
Funding SchemeData not available
78153 Le Chesnay
3508 TA Utrecht
CV4 7EZ Coventry