CHIN. PHYS. LETT. Vol. 32, No. 3 (2015) 030501
Efficiency-Controllable Random Walks on a Class of Recursive Scale-Free Trees
with a Deep Trap
*
LI Ling(李玲)
1,2**
, GUAN Ji-Hong(关佶红)
1**
, ZHOU Shui-Geng(周水庚)
3,4
1
Department of Computer Science and Technology, Tongji University, Shanghai 201804
2
Department of Basic Courses, Zhejiang Shuren University, Hangzhou 310015
3
School of Computer Science, Fudan University, Shanghai 200433
4
Shanghai Key Lab of Intelligent Information Processing, Fudan University, Shanghai 200433
(Received 16 October 2014)
Controls, especially efficiency controls on dynamical processes, have become major challenges in many complex
systems. We study an important dynamical process, random walk, due to its wide range of applications for mod-
eling the transporting or searching process. For lack of control methods for random walks in various structures,
a control technique is presented for a class of weighted treelike scale-free networks with a deep trap at a hub
node. The weighted networks are obtained from original models by introducing a weight parameter. We compute
analytically the mean first passage time (MFPT) as an indicator for quantitatively measuring the efficiency of
the random walk process. The results show that the MFPT increases exponentially with the network size, and
the exponent varies with the weight parameter. The MFPT, therefore, can be controlled by the weight parameter
to behave superlinearly, linearly, or sublinearly with the system size. This work provides further useful insights
into controlling efficiency in scale-free complex networks.
PACS: 05.40.Fb, 05.60.Cd, 89.75.Hc, 89.75.Da DOI: 10.1088/0256-307X/32/3/030501
Structure and dynamic are two traditional issues
in complex networks studies, hence many previous
works focused on the problem of how the structure fea-
tures affect dynamical processes.
[1,2]
Recently, another
problem, the control of complex networks, has begun
to attract more attention in the scientific community
and has been viewed as the ultimate goal in network
studies.
[3]
Liu et al.
[3]
developed a general framework
to address structural controllability for an arbitrary
network in a quantitative fashion. Yuan et al.
[4]
of-
fered an exact minimum input theory for controllabil-
ity. Within the framework of controllability, several
topics about control have been discussed, including
centrality control,
[5]
energy requirement,
[6]
control of
edge dynamic,
[7]
and so on. Based on the control the-
ories, we concentrate our efforts on finding a practical
technique to control the dynamical processes of net-
works to achieve the target state.
A series of recent works focus on transport behav-
ior on complex networks, for instance, traffic flow,
[8,9]
path finding
[10]
and human mobility.
[11]
As a dy-
namic process,
[12]
random walk is one of the useful
tools in modeling the behavior of transport.
[13−15]
Various other dynamical processes, e.g., page
search and access,
[16−19]
exciton transport,
[20,21]
virus
spreading
[22]
and light-harvesting
[23−25]
can be de-
scribed by random walks. What we are interested in
is the discrete random walks with a trap fixed at a
hub node (also called the trapping process).
[26]
The most widely used quantity for measuring the
efficiency of random walk is the mean first passage
time (MFPT),
[11,16,27]
also called trapping time (TT).
The MFPT from a node to the target node (trap) is
the expected time required for a walker to hit the trap
for the first time. The average trapping time (ATT)
is defined as the average of TT over all the starting
nodes in the network.
Many endeavors have been devoted to uncov-
ering the effect of network structures on the ef-
ficiency of random walk dynamics measured by
ATT, such as lattices,
[28,29]
fractals,
[30−32]
tree-
like structures,
[31−36]
scale-free networks,
[37−41]
small-
world networks,
[29,41,42]
and modularity.
[41,43,44]
The
ATT in different types of systems exhibits different
behaviors, showing that structure has a great influ-
ence on the efficiency of transport in networks.
In addition to understanding the influence of struc-
ture on network efficiency, the problem of control-
ling efficiency has aroused interest. Some control
methods have been designed for control in some net-
works, such as local transition rates for super-molecule
systems,
[19,45]
and asymmetrical and weighted edges
for fractals.
[46]
However, this issue has not yet been
studied extensively and deeply. More effective con-
trol approaches need to be proposed and adopted in
more types of networks, especially networks with the
properties extensively existing in real networks, e.g.,
scale-free. In the following we study the control meth-
ods for scale-free networks.
Deterministic networks
[47−50]
are widely used in
studies due to their merits in analytical studies. We
focus on a class of deterministic models with scale-
free properties. Distinct from the fractal networks in
Ref. [46], our model is a non-fractal one. Motivated
by the method of setting edge weights for fractal or
hierarchical networks in Ref. [51], we generalize the
*
Supported by the National Natural Science Foundation of China under Grant Nos 61173118, 61373036 and 61272254.
**
Corresponding author. Email: LingLi_zjsru@163.com; jhguan@tongji.edu.cn
© 2015 Chinese Physical Society and IOP Publishing Ltd
030501-1