09.09.2014 Views

13th International Conference on Membrane Computing - MTA Sztaki

13th International Conference on Membrane Computing - MTA Sztaki

13th International Conference on Membrane Computing - MTA Sztaki

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Zs. Gazdag, G. Kol<strong>on</strong>its<br />

• In the last three steps of the system, the symbol yes goes out to the<br />

envir<strong>on</strong>ment, and the computati<strong>on</strong> halts.<br />

It is not difficult to see that Π(n) works correctly. Indeed, Π(n) sends in<br />

every computati<strong>on</strong> to the envir<strong>on</strong>ment either the symbol no or the symbol yes.<br />

The symbol no can be introduced <strong>on</strong>ly in Case 1 above, but in this case ϕ ′<br />

should c<strong>on</strong>tain every complete clause in C n , and it follows from Propositi<strong>on</strong> 1<br />

that in this case ϕ is not satisfiable. On the other hand, yes can be introduced<br />

<strong>on</strong>ly in Case 2, but in this case there is a complete clause in C n that does not<br />

occur in ϕ ′ , which, again by Propositi<strong>on</strong> 1 means that ϕ is satisfiable. Moreover,<br />

for every formula ϕ in CNF over X n , it is easy to see that Π(n) halts in n + 5<br />

steps. Thus we have the following theorem.<br />

Theorem 1. The SAT can be solved by a uniform family of a polarizati<strong>on</strong>less<br />

deterministic P systems with active membranes using separati<strong>on</strong> rules in weak<br />

linear time, where the size of an input formula is described by the number of<br />

variables occurring in the formula.<br />

4 C<strong>on</strong>clusi<strong>on</strong>s<br />

We proposed a new approach for solving SAT by P systems with active membranes.<br />

Although our family of P systems can not be c<strong>on</strong>structed in polynomial<br />

time, <strong>on</strong>ce a P system Π(n) for a given n ∈ N is c<strong>on</strong>structed, for every formula<br />

ϕ in CNF over the variables in X n , Π(n) can decide whether ϕ is satisfiable<br />

or not in linear steps in n. Our P systems use the standard rules of active<br />

membrane systems and, in additi<strong>on</strong>, separati<strong>on</strong> rules that can change membrane<br />

labels. Moreover, our P systems are polarizati<strong>on</strong>less. It is an interesting questi<strong>on</strong>,<br />

whether we could somehow get rid of membrane label changing in our P<br />

systems. However, this seems to be difficult without using other types of rules<br />

(for example, in Secti<strong>on</strong> 5 of [2], rules of type [[ ] i [ ] j ] k → [[ ] i ] k [[ ] j ] k , where<br />

i, j, and k are labels, were used to get a linear time soluti<strong>on</strong> of SAT without<br />

polarizati<strong>on</strong>s and membrane label changing).<br />

It should be also menti<strong>on</strong>ed that the rules in (a), (c)-(f) in the definiti<strong>on</strong><br />

of Π(n) have c<strong>on</strong>stant size. Moreover, it is not difficult to see that during the<br />

evoluti<strong>on</strong> of Π(n), membranes with label i (3 ≤ i ≤ n + 3) have no more objects<br />

than the number m of the clauses in the input formula. Thus the separati<strong>on</strong><br />

rules in (b) always should act <strong>on</strong> membranes with no more than m objects.<br />

Our P system Π(n) has exp<strong>on</strong>ential size in n, thus it is a reas<strong>on</strong>able questi<strong>on</strong><br />

whether a c<strong>on</strong>stant time soluti<strong>on</strong> of SAT exists based <strong>on</strong> our P systems. Since our<br />

soluti<strong>on</strong> is uniform, i.e., the size of Π(n) depends <strong>on</strong>ly <strong>on</strong> n, and the encoding<br />

of the input formula is computable in linear time, it seems that such a c<strong>on</strong>stant<br />

time soluti<strong>on</strong> can not be easily given. On the other hand, <strong>on</strong>e can see that slightly<br />

modifying Π(n), a P system Π ′ (n) could be easily given such that, for a formula<br />

ϕ over X n , Π ′ (n) can create the complete clauses of ϕ ′ <strong>on</strong>ly in <strong>on</strong>e step. However,<br />

it is not clear how could we ensure Π ′ (n) to send out to the envir<strong>on</strong>ment the<br />

correct symbol yes or no using <strong>on</strong>ly c<strong>on</strong>stant number of steps.<br />

218

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!