15h30 - 15h55
Optimization of the Global Structure of Integrated Circuits for Digital Signal Processing
We present a constructive metaheuristic that designs infinite response digital filters from a reduced set of component in order to obtain low complexity energy efficient integrated circuits. The method is validated through the comparison to the best current algorithm on several aspects of the problem (energy consumption, execution speed, surface occupied).
15h55 - 16h20
Constrained Triangular Enclosures
One of the basic problems in geometric optimization is the computation of the convex hull. This problem leads to many applications in computer vision, robotics, etc. What do we get if we constrain the hull to be a triangle? What if we fix one of the angles, the perimeter or the area of this triangle?
16h20 - 16h45
Solution of the Weber Problem on the Sphere with 3 Points
Given n (weighted) points on a sphere, the Weber problem consists in finding a point of the sphere that minimizes the weighted sum of (geodesic) distances to the n points. Contrarily to the Weber problem on the plane, this problem may admit local minima that are not global ones, as well as saddle points. We investigate the structure of this problem for n=3.
16h45 - 17h10
An Evolutionary Algorithm for a Versatile Use of a Multiprocessor System-on-Chip
The presentation introduces the combination of an evolutionary algorithm with a high-level emulator in order to determine the appropriate strategy to implement an application on a MultiProcessor System-on-Chip.Unlike standard methods that mostly focus on one the make-span, the proposed approach is versatile, as it also consider other aspects of the problem such as the energy consumption.