Licentiate thesis 2013-001

Parallelization of Stochastic Estimation Algorithms on Multicore Computational Platforms

Olov Rosén

19 April 2013

Abstract:
The main part of this licentiate thesis concerns parallelization of recursive estimation methods, both linear and nonlinear. Recursive estimation deals with the problem of extracting information about parameters or states of a dynamical system, given noisy measurements of the system output and plays a central role in many applications of signal processing, system identification, and automatic control. Solving the recursive Bayesian estimation problem is known to be computationally expensive, which often makes the methods infeasible in real-time applications and for problems of large dimension. As the computational power of the hardware is today increased by adding more processors on a single chip rather than increasing the clock frequency and shrinking the logic circuits, parallelization is the most powerful way of improving the execution time of an algorithm. It has been found in this thesis that several of the optimal filtering methods are suitable for parallel implementation, in certain ranges of problem sizes. It has been concluded from the experiments that substantial improvements can be achieved by performing "tailor"-made parallelization, compared to straightforward implementations based on multi-threaded libraries. For many of the suggested parallelizations, a linear speedup in the number of cores has been achieved that have provided up to 8 times speedup on a double quad-core computer. As the evolution of the parallel computer architectures is unfolding rapidly, many more processors on the same chip will become available. The developed methods do not, of course, scale infinitely, but definitely can exploit and harness some of the computational power of the next generation of parallel platforms, allowing for optimal state estimation in real-time applications.

Available as PDF (2.44 MB)

Download BibTeX entry.