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

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

A. Alhazov, R. Freund, H. Heikenwälder, M. Oswald, Yu. Rogozhin, S. Verlan<br />

Simulati<strong>on</strong> of ADD-instructi<strong>on</strong> j : (ADD (a) , k, l)<br />

P 1 : p j → p ′ j<br />

P 2 : ˜p j → λ<br />

P 3 : ˜p j → #, p j → #<br />

P 4 : p ′ j → ¯p j<br />

P 5 : ¯p j → ¯p ′ j<br />

P 6 : p ′ j → #, ¯p j → #<br />

P 7 : ¯p ′ j → ˆp j<br />

P 8 : ˆp j → ˆp ′ j<br />

P 9 : ¯p ′ j → #, ˆp j → #<br />

P 10 : ˆp ′ j → ˆp′′ j<br />

P 11 : ˆp ′′<br />

j → o ap k ˜p k , ˆp ′′<br />

j → o ap l ˜p l<br />

P 12 : ˆp ′ j → #, ˆp′′ j → #<br />

The trap rules introduced in the rule sets P 3 , P 6 , P 9 , and P 12 guarantee that<br />

the rules in the rule sets P 1 , P 2 , P 4 , P 5 , P 7 , P 8 , P 10 , and P 11 have to be applied<br />

in a correct way to avoid the introducti<strong>on</strong> of the trap symbol #.<br />

Without loss of generality, we may assume that the last instructi<strong>on</strong> applied in<br />

the register machine M is a SUB-instructi<strong>on</strong> (labeled by j) being applied to the<br />

empty register 1; instead of taking the rule p ′ j → p f ˜p f we take the rule p ′ j → λ<br />

into P 10 . If until then the acti<strong>on</strong>s of the register machine have been simulated<br />

correctly in Π C , <strong>on</strong>ly the terminal results c<strong>on</strong>sisting of specific numbers of copies<br />

of the symbols o i , 3 ≤ i ≤ m, remain in the membrane regi<strong>on</strong>. The P system<br />

therefore finally stops before entering a new cycle P 1 · · · P 12 ; hence, in sum we<br />

have shown that the language generated by the register machine is also generated<br />

by the time-varying P system Π C with delay 2, i.e.,<br />

P sRE ⊆ L (ncoo-T V ut OP 1 (12) , 2) .<br />

As weak c<strong>on</strong>trol is the less restrictive c<strong>on</strong>trol variant, we immediately infer<br />

P sRE ⊆ L (ncoo-wT V ut OP 1 (12) , 2) .<br />

too.<br />

□<br />

As a challenge for future research it remains to search for a proof which eventually<br />

allows to obtain computati<strong>on</strong>al completeness with delay <strong>on</strong>e <strong>on</strong>ly. Another<br />

parameter to be improved is the period of the c<strong>on</strong>trol language. Eventually there<br />

might also be a trade-off between these two parameters: It is easy to see that<br />

using p j <strong>on</strong>ly instead of the pair p j ˜p j we could save the sec<strong>on</strong>d rule set at the<br />

beginning of a simulati<strong>on</strong>; then, in additi<strong>on</strong>, we might even omit P 3 and P 12 ,<br />

but with these rather obvious changes we would increase the delay to 3.<br />

The c<strong>on</strong>structi<strong>on</strong> given in the preceding proof shows that any acti<strong>on</strong> in the<br />

time-varying P system can be seen as simple multiset rewriting, hence, we obviously<br />

get the following result:<br />

112

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

Saved successfully!

Ooh no, something went wrong!