Universität Koblenz-Landau FB 4 Informatik
Prof. Dr. Viorica Sofronie-Stokkermans∗1 24.01.2018
M.Ed. Dennis Peuter∗2
Exercises for Advances in Theoretical Computer Science Exercise Sheet 12
Due at 29.01.2018, 10:00 s.t.
Exercise 12.1
Give a function f : Σ∗ →Σ∗ which polynomially reducesL1 to L2, or explain why this is not possible:
I) Σ ={0,1,2};
L1 ={w∈ {0,1}∗|w is the representation of a prime number in base 2}; L2 ={w∈ {0,1,2}∗ |wis the representation of a prime number in base 3}. II) Σ ={0,1,2,3,4};
L1 ={w∈ {0,1}∗|w is the representation of a prime number in base 2}; L2 ={w∈ {0,1,2,3,4}∗ |wis the representation of a prime number in base 5}. III) Σ ={0,1};
L1 ={w∈ {0,1}∗|w is the representation of a prime number in base 2}; L2 ={w∈ {1}∗ |wis the representation of a prime number in base 1}. IV) Σ ={a, b,0,1};
L1 ={anbn|n≥0}; L2 ={1}.
Here, for IV), we require that for all w∈Σ∗\L1,f(w) = 0.
Exercise 12.2
We know that SAT is NP-complete. In the previous exercise we saw that satisabi- lity of formulae in DNF can be checked in polynomial time, so DNF-SAT = {F | F is a satisable formula of propositional logic in disjunctive normal form} is in P.
If we could construct a polynomial reduction of SAT to DNF-SAT (i.e. if we could prove that SAT ≺pol DNF-SAT) then we could show that P = NP.
Formulae in propositional logic can be transformed to DNF using distributivity:
A∧(B1∨ · · · ∨Bk)≡(A∧B1)∨ · · · ∨(A∧Bk).
Why does this not lead to a polynomial reduction?
∗1 B 225 sofronie@uni-koblenz.de https://userpages.uni-koblenz.de/~sofronie/
∗2 B 223 dpeuter@uni-koblenz.de https://userpages.uni-koblenz.de/~dpeuter/
If you want to submit solutions, please do so until 29.01.2018, 10:00 s.t. via the cardboard box in the shelf in room B 222 or via e-mail (with Homework ACTCS in the subject) to dpeuter@uni-koblenz.de.