Chinese Journal of Electronics
Vol.25, No.2, Mar. 2016
Parallel Solution for Maximum Independent Set
Problem by Programmable Tile Assembly
∗
HUANG Yufang
1,2
, XIAO Jianhua
3
, JIANG Keqin
4
and CHEN Zhihua
2
(1. School of Mathematics, Southwest Jiaotong University, Chengdu 610031, China)
(2. School of Automation, Huazhong University of Science and Te chnology, Wuhan 430074, China)
(3. Rese arch Center of Logistics, Nankai University, Tianjin 300071, China)
(4. School of Computer and Information, Anqing Normal University, Anqing 246133, China)
Abstract — Parallelism in the theoretical computation
mainly depends on the particular paradigm or compu-
tational environment considered, and its importance has
been confirmed with the emergence of each novel comput-
ing technique. Programmable tile assembly is a novel com-
puting technique to tackle computationally difficult prob-
lems, in which computing time grows exponentially corre-
sponding to problematic size. The Maximum independent
set (MIS) problem is a typical nondeterministic polyno-
mial problem, which is often associated with strategy ap-
plications. In this paper, a novel approach dealing with
the MIS problem is proposed based on the abstract tile
assembly model, which is believed to be better than the
conventional silicon-based computing in solving the same
problem. The method can get the solutions of the MIS
problem in θ(m + n) time complexity based on θ(mn)dis-
tinct tile types.
Key words — Molecular computing, Tile assembly
model, Maximum independent set (MIS), Self-assembly,
Parallel computing.
I. Introduction
Parallel computing was originally motivated by the
need to speed up computation, especially for those tasks
whose sequential running time is prohibitively long. Con-
ventional silicon-based computer may take months or even
years to process the calculations to solve specific classes
of problems, especially Nondeterministic polynomial (NP)
problems. Therefore, searching out other possibilities to
replace the current silicon-based computer to solve NP
problems catches the attention of researches from differ-
ent fields.
Biological systems are a rich source for parallel com-
puting devices design. Natural computing is a typical
computing model inspired by biological systems such as
neural computing, DNA computing and membrane com-
puting. In recent years, the computing devices inspired by
cells, tissues or neural networks are deeply investigated.
Most of such systems are proved to be computationally
complete (equivalent with Turing machine) as number
computing devices
[1,2]
, language generators
[3,4]
, and func-
tion computing devices
[5]
. They are also used to theoret-
ically solve presumably intractable problems in a feasible
time
[6,7]
. Among various approaches, tile assembly com-
puting seems to be the most feasible way to solve NP
problems. This work focuses on computing systems based
DNA tile assembly which is a prospective method to over-
come the ultimate limits of silicon-based technology
[8]
.
Self-assembly is a powerful process by which a collec-
tion of relatively simple components coalesce to form more
complex structures. The process is guided by only local
interactions between the components, which typically fol-
low a basic set of rules. Key features of self-assembly are
its probabilistic nature and local programmability which
can be leveraged to design better self-assembly. In order
to model the tile assembly, theoretical models have been
established, and one of the most popular among these
is the abstract Tile assembly model (aTAM) introduced
by Winfree
[9]
. aTAM was based on a cross between the
theoretical study of Wang tiles
[10]
and novel DNA com-
plexes being synthesized within Seeman’s laboratory
[11]
.
The aTAM provides a more high-level abstraction which
∗
Manuscript Received Feb. 13, 2014; Accepted Sept. 19, 2014. This work is supported by the National Natural Science Founda-
tion of China (No.61402382, No.61373066, No.61370105, No.61202204), the Fundamental Research Funds for the Central Universities
(No.2682014CX054), Postdoctoral Science Foundation of China (No.2012M521427), and Anhui Provincial Natural Science Foundation
(No.1408085MF131).
c
2016 Chinese Institute of Electronics. DOI:10.1049/cje.2016.03.002