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.

B. Nagy<br />

2.8 Neural-like <strong>Membrane</strong> Systems<br />

Neural-like membrane systems (or tissue P systems) can solve SAT in linear<br />

time by using an alphabet of chemical objects (or excitati<strong>on</strong>s/impulses) with<br />

cardinality 2 k+1 −1 [34], since the system first generates all the truth-assignments<br />

in the form of strings of length k using letters t i and f i with 1 ≤ i ≤ n.<br />

2.9 Quantum P-systems<br />

In this secti<strong>on</strong> we recall a mixed paradigm, the quantum UREM P-systems [19].<br />

Quantum computing is also counted as a new computing paradigm based <strong>on</strong> some<br />

‘unc<strong>on</strong>venti<strong>on</strong>al’ features of quantum mechanics. There is no space here to recall<br />

all details, there are various textbooks for this topic also (see, e.g., [11]). The<br />

main features of this paradigm are the following. A quantum bit (qubit) can have<br />

infinitely many values, technically any unit vector of a four dimensi<strong>on</strong>al space<br />

(complex coefficients for both of the possible values |0〉 and |1〉), the quantum<br />

superpositi<strong>on</strong> of the two possible states. The used unitary operati<strong>on</strong>s (rotati<strong>on</strong>s)<br />

can be written by 2 by 2 (complex valued) matrices. However by measurement<br />

<strong>on</strong>ly the ‘projecti<strong>on</strong>’ of the superpositi<strong>on</strong> is obtained, the system reaches <strong>on</strong>e of<br />

the states |0〉 and |1〉 with the probability based <strong>on</strong> their coefficients. Having<br />

a system with n qubits the dimensi<strong>on</strong> of its state (i.e., the stored informati<strong>on</strong>)<br />

grows exp<strong>on</strong>entially: the state can be described by a 2 n dimensi<strong>on</strong>al vector.<br />

The corresp<strong>on</strong>ding operators are described by matrices of size 2 n by 2 n (can be<br />

obtained by tensor product). By special quantum effect, called entanglement,<br />

a state in superpositi<strong>on</strong> of some qubits together may not be c<strong>on</strong>structible by<br />

the tensor products of the qubits. In this way exp<strong>on</strong>ential ‘space’ can be used.<br />

(Theoretically it is nice, technologically it is very hard task to produce systems<br />

that can use larger (e.g., 30) number of qubits in a system.)<br />

In UPREM P-systems there are unit rules and energy assigned to membranes.<br />

The rules in these systems are applied in a sequential way: at each computati<strong>on</strong><br />

step, <strong>on</strong>e rule is selected from the pool of currently active rules, and it is applied.<br />

The system further developed by mixing it with quantum computing. Quantum<br />

UPREM P-systems are proved to be universal without priority relati<strong>on</strong> am<strong>on</strong>g<br />

the rules [18]. In this way, a quantum computing technique: soluti<strong>on</strong> to SAT by<br />

the quantum register machine is simulated. The given semi-uniform algorithm<br />

uses the alphabet to describe the possible quantum states, and as the number<br />

of possible states of the system is exp<strong>on</strong>ential <strong>on</strong> the number of used qubits, the<br />

size of the alphabet is exp<strong>on</strong>ential <strong>on</strong> the input formula.<br />

2.10 Mutual Mobile <strong>Membrane</strong> Systems<br />

In mutual mobile membrane systems endocytosis and exocytosis work whenever<br />

the involved membranes ‘agree’ <strong>on</strong> the movement. Actually, in [4] <strong>on</strong>ly weak NPcomplete<br />

problems, i.e., the knapsack, the subset-sum and the 2-partiti<strong>on</strong> are<br />

solved and our main problem, the str<strong>on</strong>g NP-complete SAT does not. However,<br />

this branch of P-systems seems to be interesting and therefore we decided to<br />

330

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

Saved successfully!

Ooh no, something went wrong!