Abstract
This paper provides an overview of the field of parameterized parallel complexity by surveying previous work in addition to presenting a few new observations and exploring potential new directions. In particular, we present a general view of how known FPT techniques, such as bounded search trees, color coding, kernelization, and iterative compression, can be modified to produce fixed-parameter parallel algorithms.
Highlights
To how computational problems are classified according to their tractability or whether efficient algorithms exist for them, problems are classified according to whether efficient parallel algorithms exist to solve them
A problem is fixed-parameter parallel tractable (FPPT) if it can be solved in O( f (k ) ∗ logc1 (n))
Iterative compression algorithms are mainly applied to some minimization problems when it is possible to construct a feasible solution whose size is slightly larger than the target solution size k, which is the parameter in this case
Summary
To how computational problems are classified according to their tractability or whether efficient algorithms exist for them, problems are classified according to whether efficient parallel algorithms exist to solve them. A parallel problem is in NC if it can be solved in O(logc (n)) time using O(nc2 ) processors, where c1 and c2 are constants [4] To convert this notion into a parameterized definition that we can use (or relate to) later, and following the same assumptions of [1], we consider a parameterized problem to be in NC if it is in NC for any given value of the parameter. The more restrictive class of fixed-parameter parallel (FPP) problems was defined by Cesati and Ianni in [1] where a strictly poly-logarithmic factor in the running time is required, by replacing h(k ) with a constant It was observed in [6] that the function g(k ). In particular we overview the methods used for obtaining a parameterized parallel algorithm from a known sequential fixed-parameter algorithm
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have