Network synchronization landscape reveals

compensatory structures, quantization, and

the positive effect of negative interactions
Takashi Nishikawaa,b,1 and Adilson E. Mottera,c,1
a
Department of Physics and Astronomy, Northwestern University, Evanston, IL 60208; bDepartment of Mathematics, Clarkson University, Potsdam, NY
13699; and cNorthwestern Institute on Complex Systems, Northwestern University, Evanston, IL 60208

Edited by Giorgio Parisi, University of Rome, Rome, Italy, and approved April 7, 2010 (received for review November 6, 2009)

Synchronization, in which individual dynamical units keep in pace

with each other in a decentralized fashion, depends both on the
dynamical units and on the properties of the interaction network.
Yet, the role played by the network has resisted comprehensive
characterization within the prevailing paradigm that interactions
facilitating pairwise synchronization also facilitate collective synchronization. Here we challenge this paradigm and show that networks with best complete synchronization, least coupling cost, and
maximum dynamical robustness, have arbitrary complexity but
quantized total interaction strength, which constrains the allowed
number of connections. It stems from this characterization that
negative interactions as well as link removals can be used to
systematically improve and optimize synchronization properties
in both directed and undirected networks. These results extend
the recently discovered compensatory perturbations in metabolic
networks to the realm of oscillator networks and demonstrate
why less can be more in network synchronization.
complex networks nonlinear dynamics stability analysis network
interactions collective behavior

locking animals (1, 2), self-coordinated moving sensors (3),

bursting neurons (4), pace-matching chemical oscillators
(5), and frequency-locked power generators (6) are some of many
physical manifestations of spontaneous synchronization. Like
other forms of collective phenomena (7), synchronization depends critically on the properties of the interaction network
(8). Common wisdom suggests that synchronization is generally
easier to achieve with more interactions, that synchronization
properties change monotonically as the number of available interactions is varied, and that certain network structures facilitate
while others inhibit synchronization. These three expectations,
however, are all false because they ignore the possibility of compensatory structural effects. For example, removing a link from a
globally connected network makes it less likely to synchronize,
but targeted removal of additional links can enhance synchronization (915). Heterogeneous distribution of coupling strengths
or connectivity generally inhibits synchronization (16, 17), but
when combined they can compensate for each other (18, 19).
Bringing this argument one step further, while previous studies
have focused mainly on positive interactions (but see refs.
2022)presumably because negative interactions alone generally do not support synchronyit is actually easy to provide examples in which negative interactions help stabilize synchronous
states. This is illustrated in Fig. 1, where the network composed of
black and blue links is not optimal for synchronization but the
removal of the blue interactions or, alternatively, the addition
of interactions with negative strengths (red links) makes it optimal; the same is achieved by weighting the strengths of all three
input interactions of each purple node by a factor of 23. However, the counterintuitive properties that start to emerge from
such case studies currently lack a common in-depth explanation
that is both comprehensive and predictive in nature.

1034210347 PNAS June 8, 2010 vol. 107 no. 23

Fig. 1. Examples of compensatory network structures. Nodes represent

identical dynamical units and links diffusive couplings. The network consisting of the black and blue links, all having strength 1, is not an optimal
network for synchronization, namely, one that synchronizes over the widest
range of global coupling strength (to be formalized in Results). However, it
can be made optimal either by (i) removing the blue links, (ii) adding the red
links with negative strength 1, or (iii) scaling the strengths of all in-links of
the purple nodes by a factor of 23.

Here we show that these and other apparently disparate properties follow from the discovery we present below that networks
optimal for synchronization have a quantized number of links, in
multiples of a constant that depends only on the number of nodes
and on the connection strengths. We derive our results for the
local stability of synchronous states in networks of identical units
and provide evidence that our main results remain valid for
networks of nonidentical units. We choose to focus on optimal
networks because, as we show, this class is very rich, can be
dealt with analytically, and forms a multicusped synchronization
landscape, which underlies all synchronizable networks and from
which suboptimal networks can be studied using perturbation and
Author contributions: T.N. and A.E.M. designed research; T.N. and A.E.M. performed
research; T.N. analyzed data; and T.N. and A.E.M. wrote the paper.
The authors declare no conflict of interest.
This article is a PNAS Direct Submission.
1

To whom correspondence may be addressed. E-mail: tnishika@clarkson.edu or motter@

northwestern.edu.

This article contains supporting information online at www.pnas.org/lookup/suppl/

doi:10.1073/pnas.0912444107/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.0912444107

Results
Optimal Networks for Synchronization. We represent the structure
of a network with n nodes using its adjacency matrix A
Aij 1i;jn , where Aij is the strength of the link from node j to
node i. We consider the network dynamics governed by

x_ i Fxi

Aij Hxj Hxi
;

[1]

j1

where xi is the state vector of the ith dynamical unit, F represents

the dynamics of an isolated unit, Hxj is the signal that the jth
unit sends to other units (28), and d is the global coupling
strength 0 normalized by the average coupling strength per
node d n1 i ji Aij . As a result of this normalization, the
dynamics for a given is invariant under scaling of A by a
constant, which does not change the network structure. This system has served as a workforce model to study synchronization
because it allows analysis of the network influence without detailed specification of the properties of the dynamical units.
For example, using a stability function that is independent
of the network structure (7, 28, 29, 30), the condition for a
synchronous state x1 t xn t st to be linearly stable
is i < 0 for i 2; ; n, where 2 ; ; n are the nonidentically zero eigenvalues of the Laplacian matrix L defined by
Lij ij k Aik Aij (see Materials and Methods). Thus, the smaller the normalized spread of the eigenvalues in the complex
plane, which we quantify using
2

n
1
ji j2 ;
d2 n 1

where

i2

1 n
;
n 1 i

[2]

i2

the more synchronizable the network will generally be. Another

measure of synchronizability is the coupling cost K
s i ji Aij s m at the synchronization threshold s ,
whose minimization is equivalent to the condition
2 3 n > 0;

[3]

which is also equivalent to the condition 0. This condition

constrains all the eigenvalues to be real, even though the
networks can be directed and directed networks have complex
eigenvalues in general. Condition 3 is also equivalent to the maximization of the range of that allows for stable synchronization
as well as to the maximization of dynamical robustness, in that it
can achieve the fastest exponential convergence to synchronization (see Materials and Methods). [This may relate, for example,
to the finding that heterogeneous distribution of links, which tend
to make the distribution of i s heterogeneous, leads to longer
transients to synchronization in ecological network models with
strong couplings (31).] We emphasize that the equivalence of
Nishikawa and Motter

these conditions holds for a wide variety of possible stability functions, including those for which the region defined by < 0
may be finite or semiinfinite, or have multiple disconnected components. Having the maximum range, minimum coupling cost,
and maximum dynamical robustness, the networks satisfying
Eq. 3 are called the optimal networks for synchronization. Similar
results can be derived for networks of nonidentical units, in which
the functions F and H are possibly different for different nodes,
and this more general case will be discussed below.
Quantized Number of Links. A surprising consequence of having
these optimal synchronization properties is the quantization of
the number of links in the networks. We find, for example, that
for binary interaction networks (i.e., Aij 0; 1) satisfying condition 3, the number of links m i ji Aij is quantized to multiples of n 1. That is,

m qk kn 1;

where k 1; 2; ; n:

[4]

This follows from the identity m i Lii ni2 i n 1,

combined with the fact that condition 3 constrains the real eigenvalue further to be an integer for networks with integer-valued
interactions (see SI Text, Section 1). Consequently, any network
with m strictly between these quantized values must have > 0
and hence cannot be optimal. What is then the minimum for all
such networks with a given m? We denote this minimum by
min m. Based on our analysis of all networks with n 6,
we conjecture that the condition to achieve this minimum is
that the Laplacian eigenvalues (counting multiplicity) have the
form
qk1 m

mqk

z}|{ z}|{
0; k; ; k; k 1; ; k 1;

[5]

where k is the integer satisfying qk m qk1 . Note that,

analogously to Eq. 4 for m qk , this condition asserts that the
eigenvalues are not only real but also integer for any m. This leads
to our prediction that
min m

1
n 1d

q
m qk qk1 m;

[6]

which is expected to be valid for binary interaction networks with

arbitrary number of nodes and links. Indeed, Fig. 2A shows that
for 10 n 12, a simulated annealing procedure identified networks (blue dots) with within 103 of (but not smaller than)
min m predicted by Eq. 6 (black curve). Starting from the (optimal) fully connected network [with m nn 1] at the right
end of the curve min m, any initial link deletion necessarily
makes synchronization more difficult. Further deletions, however, can make it easier, bringing the networks back to optimal
periodically as a function of m at the cusp points and eventually
leading to (optimal) directed spanning trees with m n 1 (see
Movie S1). The optimization of synchronization at these cusp
points is analogous to the optimization of the dynamical range
of excitable networks at the critical points of phase transitions
(32). Similar cusps are also observed for the Laplacian
eigenvalues of a structurally perturbed optimal network as a function of the perturbation (see SI Text, Section 2). Note that,
although the cost K s m generally depends on m, it is actually
independent of m for optimal networks of given size n because
the synchronization threshold s s m compensates for any
change in m.
To reveal the intricate dependence of synchronization on the
number of nodes and links, we now consider min as a function
of both m and n (Fig. 2B) based on our prediction 6. Each point
on this synchronization landscape represents all networks with
min for a given m and n, and the number of such
PNAS

June 8, 2010

vol. 107

no. 23

10343

APPLIED PHYSICAL
SCIENCES

numerical methods. An immediate consequence of our quantization result is that the addition of links to an optimal network
generally results in a suboptimal network, providing systematic
examples of link removals that enhance synchronizability. Similar
quantization results hold true for optimal networks with negative
interactions, which we derive using a generalized complement
transformation that maps them into networks with only positive
interactions. We show that negative interactions can reveal antagonistic structural relations and counterbalance structural heterogeneities, with potential implications for inhibitory interactions in
neural (2325), power-grid (6, 26), and cell-regulatory networks
(27). The interactions between power-grid generators, for example, can be positive or negative depending on the inductive vs.
capacitive nature of the corresponding network elements [e.g.,
for the Western U.S. power grid they are split 97% vs. 3% (26)].

minimizes (red arrow in Fig. 2B). The sharp valleys are observed along the lines m qk and therefore correspond to
the cusp points in Fig. 2A. Because two adjacent points on
the landscape may include networks that do not have similar
structures, it is surprising that one can actually move from
one point to the next often by a simple structural modification
(see Movie S1). Indeed, moving in the direction of the m coordinate can be achieved by simply adding or removing a link that
induces the smallest increase or largest decrease in . Along the
line m qk , an optimal network can be grown and kept optimal, by adding a node and connecting any k existing nodes to the
new node (see SI Text, Section 3). The flexibility of choosing new
links in this construction strongly suggests that optimal networks
can have arbitrarily complex connectivity structures as the network size grows. Another interesting landscape appears when we
compute the cost K as a function of m and n (Fig. 2C) based on
condition 5. In this landscape, optimal networks lie along the
lines of discontinuity, resulting from the discontinuous change
in s that occurs when m changes from qk 1 to qk and defines
a sawtooth function along the m coordinate. Note that for optimal networks, the cost K is independent of m, as mentioned
above, but increases linearly with n and can be expressed as K
K 2 n 1, where K 2 is the coupling cost for a network of two
coupled units. A different but potentially related structure is
the roughness of the time horizon considered in the synchronization of parallel processing units in distributed computing
(33, 34).
While the presentation above focused on networks of identical
units, the main results also hold true for networks of nonidentical
units. Adopting the framework of ref. 35 and developing further
analysis for networks of one-dimensional maps, we show in SI Text
and Figs. S1 and S2 (Section 4) that complete synchronization is
possible even for nonidentical units. Moreover, we show that this
is possible only for networks satisfying Eq. 3 in addition to the
condition that the Laplacian matrix L is diagonalizable. Since
any such networks must also exhibit the same quantization expressed in Eq. 4, we also expect cusps similar to those shown
in Fig. 2. For each quantized number of links qk , we can show
that there is a network that can synchronize completely despite
the heterogeneity of the dynamical units. Therefore, for both
identical and nonidentical units, condition 5 can be used to systematically construct examples of suboptimal networks that can
be made optimal by either adding or removing links.

Fig. 2. Quantization in the synchronization and cost landscapes. (A) Estimated by simulated annealing for networks with n 1012 nodes, the minimum value min (blue dots) of (Eq. 2) shows cusp points at periodic values of
the total interaction strength, m qk (Eq. 4). In this representation as a function of m, the results for networks with positive (Aij 0; 1) and mixed positive/negative (Aij 0; 1) interactions are identical (blue dots) and coincide
with the theoretical prediction (black curves) (Eq. 6). The results for the case
Aij 1 (red dots) show a similar but different quantization behavior,
following Eq. 8. (B) Synchronization landscape min as a function of both
m and n, computed from Eq. 6. Through simulated annealing, a network with
initially high (top red dot) evolves to a network satisfying Eq. 5 and min
(bottom red dot). (C) The coupling cost K as a function of m and n in units of
K 2 (the cost for a network of two coupled units). The cost takes the lower
values along the lines of discontinuity (m qk ).

networks is expected to grow combinatorially with n (see SI Text,

Section 3). All other networks with the same m and n are represented by points directly above that point. The evolution
of one such network by rewiring links under pressure to optimize
synchronizability can be visualized as a vertical descent from a
point above the landscape toward a point on the landscape that
10344

www.pnas.org/cgi/doi/10.1073/pnas.0912444107

Stabilizing Effect of Negative Interactions. Interestingly, the exact

same quantization effect described above for binary interaction
networks is also observed when we allow for negative interactions
and interpret m i ji Aij as the net number of links. To see
this, we use a generalization of the complement network, which
we define for a given constant to be the network with adjacency
matrix Ac given by

Acij Aij 1 ij :

[7]

[This includes the special case 1, which for undirected

unweighted networks corresponds to the previously studied case
of complement graphs (36).] The transformation from a network
to its complement maps m to nn 1 m, i to n i , to
m
nn1m, and thus an optimal network to another optimal netm
work when > nn1
(see Materials and Methods). This also establishes a mapping between networks capable of complete
synchronization of nonidentical units, since the Laplacian matrix
for such a network remains diagonalizable under the transformation (see SI Text, Section 4). The condition on avoids (nonsynchronizable) networks having eigenvalues with negative real part
as an artifact of the transformation. We argue that this generalized complement transformation is a powerful tool in analyzing
Nishikawa and Motter

 min m

q
1
m qn2k1 qn2k m
n 1d

[8]

for qn2k1 m qn2k (red dots, Fig. 2A). The synchronization

landscape can be drastically different in some extreme cases. On
the one hand, within the widely studied class of undirected unweighted networks (corresponding to a stronger set of constraints,
Aij Aji , Aij 0; 1), no network with m < nn 1 satisfies
the optimality condition 3 (10, 37), which indicates that min m >
0 for all m < nn 1. On the other hand, within the class of
weighted networks corresponding to having no constraint on
the interaction strengths, an optimal network can be constructed
for any number of links n 1 (e.g., the hierarchical networks in
ref. 10), which implies that min is identically zero in this particular case.
To demonstrate that the synchronization enhancement by
negative interactions goes much beyond the realm of optimal networks, we propose a simple algorithm for assigning strength 1 to
directional links in an arbitrary network with all link strengths
initially equal to 1. Our strategy is based on the observation that
the in-degree distribution is a main factor determining synchronizability (1719, 38, 39), where the in-degree (or the total input
strength) of node i is defined to be ji Aij . Because heterogeneity
in the in-degree distribution tends to inhibit synchronization
(1719, 38, 40), here we use negative interactions to compensate
for positive interactions and to homogenize the in-degree distribution. For a given network, we first choose randomly a node with
the smallest in-degree and change the strength of each out-link of
that node to 1, unless it makes the in-degree of the target node
smaller than the mean in-degree of the original network. We keep
strength 1 for the other out-links, as well as all the in-links.
Having treated this node, we repeat this process for the subnetwork of untreated nodes, considering links and hence degrees
Nishikawa and Motter

only within that subnetwork. We continue this process until all

nodes are treated. Applying this algorithm to the network in
Fig. 3A, we see that the high in-degree nodes of the initial network (those in the shaded area) receive all of the compensatory
negative interactions (red arrows), reducing by nearly 35%
(Fig. 3B). The effectiveness of the method was further validated
(Fig. 3C) using random scale-free networks (41) generated by the
standard configuration model (42), where we see more dramatic
effect for more heterogeneous networks, reducing by as much
as 85%. The synchronization enhancement was indeed accompanied by the homogenization of the in-degree distribution (see
SI Text and Fig. S3, Section 5). The use of negative directional
interactions in our algorithm suggests that the enhancement is
partly due to link directionality (see ref. 43 for an enhancement
method purely based on link directionality), which generally plays
an important role in synchronization (see, for example, refs. 21
and 44). However, negative strength of interactions alone can
also produce similar enhancement in random scale-free networks
when they are assigned to bidirectional links between hubs (see
SI Text and Fig. S4, Section 6).
APPLIED PHYSICAL
SCIENCES

networks with negative interactions because it reduces problems

involving negative interactions to those involving only positive
interactions when we choose max Aij .
As an example of quantization with negative interactions, we
consider the class of networks for which Aij 0; 1 and assume
that 2 ; ; n have positive real parts to ensure that the network
can synchronize. Mapping these networks under the complement
transformation with 1, we obtain positive interaction networks with Aij 0; 1; 2. Conjecture 5 applied to the resulting networks then leads to a prediction identical to Eq. 6, which we
validated by simulated annealing for networks of size up to 12
(blue dots, Fig. 2A). Thus, all the results discussed above based
on Eq. 6, including Eq. 4 and the shape of the synchronization
landscape, are predicted to be valid even when negative interactions are allowed. Whether the networks with min m
actually do have negative interactions is not a priori clear because
the synchronization landscape can be built entirely by using the
subset of networks with only positive interactions. Our simulated
annealing shows, however, that many optimal networks (i.e., with
0) have negative interactions, as illustrated by an example in
Fig. 1. In addition, complete synchronization of nonidentical
units is possible in the presence of negative interactions (see
SI Text, Section 4). These networks provide clear and systematic
examples of negative interactions that improve synchronization,
as removing negative interactions from any such network would
in general push m off the quantized values (m qk ) and make the
network suboptimal.
The quantization of m and the shape of the synchronization
landscape, though they were identical for the two examples
above, do depend critically on the constraints imposed on the
interactions. Consider, for example, the fully connected networks
with interaction strengths 1. It can be shown that the implied
constraint Aij 0 leads to a different quantization compared
to Eq. 6, which involves multiples of 2n 1 rather than n 1,

Fig. 3. Improving synchronization with negative interactions. (A) Strength

1 is preferentially assigned to directional links (red arrows) connecting to
high in-degree nodes (those in the shaded area) in an initially undirected network. The black links have strength 1. (B) The initial spread of the Laplacian
eigenvalues is significantly reduced as a result (from blue to red dots). The
error bars indicate the mean and 1 standard deviation of the corresponding
distributions. (C) The mean and the standard deviation (red error bars) of
the percentage reduction in over 20 realizations (blue dots) of random
undirected scale-free networks with 1,000 nodes, minimum degree 5, and
the scaling exponent . The networks were generated by connecting nodes
randomly according to a power-law degree distribution P .
PNAS

June 8, 2010

vol. 107

no. 23

10345

Conclusions
Even though negative interactions and link removals by themselves tend to destabilize synchronous states, we have shown that
they can compensate for other instabilities, such as those resulting from a forbidden number of interactions or from a heterogeneous in-degree distribution. This establishes an unexpected
relation between network synchronization and recent work on
metabolic networks, where locally deleterious perturbations have
been found to generate similar compensatory effects that are
globally beneficial in the presence of other deleterious perturbations (45, 46). For example, the removal of a metabolic reaction
or, equivalently, of its enzyme-coding gene(s)can often be
partially or totally compensated by the targeted removal of a second metabolic reaction. That is, the inactivation of specific metabolic reactions can improve the performance of defective or
suboptimally operating metabolic networks, and in some cases
it can even rescue cells that would be nonviable otherwise
(45). Other research has shown that similar reaction inactivation
occurs spontaneously for cells evolved to optimize specific metabolic objectives (47). These apparently disparate examples share
the common property that the collective dynamics is controlled,
and in fact enhanced, by constraining rather than augmenting the
underlying network. Such network-based control, including the
conditionally positive impact of otherwise detrimental interactions and constraints, is most likely not unique to these contexts.
In neuronal networks, for example, inhibitory interactions
have been predicted to facilitate synchronous bursting (2325,
4850). Even more compelling, it has been established for both
animal (51) and human (52, 53) cerebral cortices that the density
of synapses first increases and then decreases during early
brain development, suggesting a positive role played by link
removal also in neuronal networks. More generally, we expect
that the eigenvalue spectrum analysis that led to our conclusions
will help investigate analogous behavior in yet other classes of
network processes governed by spectral properties, such as
epidemic spreading (54) and cascading failures (55, 56).
Taken together, the highly structured characteristics revealed
by the synchronization landscape explain why the characterization of the network properties that govern synchronization has
been so elusive. Numerous previous studies performed under
comparable conditions have sometimes led to apparently conflicting conclusions about the role played by specific network
structural properties such as randomness, betweenness centrality,
and degree distribution (30). Our results indicate that at least part
of these disagreements can be attributed to the sensitive dependence of the synchronization properties on the specific combination of nodes and links, as clearly illustrated by the nonmonotonic, periodic structure of cusps exhibited by the synchronization
landscape. We suggest that insights provided by these findings will
illuminate the design principles and evolution mechanisms of
both natural and engineered networks in which synchronization
is functionally important (57).
Synchronization Stability Analysis. The stability analysis can be carried out
using a master stability approach based on linearizing Eq. 1 around synchronous states (28). We apply a coordinate transformation to reduce the
Laplacian matrix L to its Jordan canonical form and decompose the variational equations into components along the corresponding eigenspaces
(9, 10). This leads to a master stability equation,

[9]

where i for the eigenspace corresponding to i . The stability function

is the maximum Lyapunov exponent for the solution yt 0 of Eq. 9.
The Aij are not required to be nonnegative and, since the derivation of is
based on a Jordan form, L is not required to be diagonalizable either (9, 10).
Geometrically, the stability condition is that all the (possibly complex)
numbers 2 ; ; n lie in the region f C < 0g of the complex plane.
10346

www.pnas.org/cgi/doi/10.1073/pnas.0912444107

Dynamical Robustness of Optimal Networks. For a given stability function ,

a network satisfying condition 3 has the maximum rate of exponential convergence to synchronization among all networks. This is so under the assumption that there is a physical limit M to the coupling strength of individual
links, Aij M, and that the stability function is continuous and monotonically increasing with respect to the distance from the real axis, which
appears to hold true for all known examples of in the literature (58).
From the limit on coupling strengths, it follows that the real part of i is
bounded by a constant Mnn 1, which, combined with the assumption
on , implies that the exponential rate of convergence is completely determined by the value of in the interval 0; Mnn 1
on the real axis. In
this interval has a global minimum at some  , and thus r   > 0
is the maximum possible rate of exponential convergence. If the network
satisfies condition 3, all perturbation eigenmodes converge at the maximum

rate r  when we choose . In contrast, if the network violates condition 3,
there must be at least one eigenvalue i such that i  [excluding the
exceptional situation where multiple values of i fall precisely on multiple
global minima of ], resulting in an eigenmode that converges at a slower
rate r i < r  . Therefore, although optimal networks may suffer from
initially slower convergence to synchronization that is polynomial in time (9,
10), the long-term convergence rate is dominated by r  and is faster than for
any other network. Indeed, the deviation from the synchronous state can be

written as Pter t for an optimal network and as Qtert for a suboptimal network, where Pt and Qt are polynomials. The ratio between
the deviations in the two cases is then

Pt r rt
e
0 as t
Qt

[10]

and, in particular, is less than 1 for sufficiently large t, implying that the
deviation will eventually become smaller for the optimal network.

Laplacian Spectrum of Generalized Complements. We show that if the eigenvalues of L are 0, 2 ; ; n (counting multiplicity), then the eigenvalues of the
Laplacian matrix Lc of the generalized complement (defined by Eq. 7) are 0,
n 2 ; ; n n . This result follows from the relation

Lc ; x 1n1

x
L; n x;
n x

[11]

where L; x detL xI is the characteristic polynomial of the matrix L and

I is the n n identity matrix. We derive this relation by following the strategy
of the proof of Lemma 2.3 in ref. 59 for undirected networks with nonnegative link strengths, to now consider directional and weighted links, possibly
with negative strengths. From the definition of the complement transformation, we can write L Lc nI J, where J is the n n matrix with every
entry equal to one. Using this and well-known properties of the
determinant, we have

Lc ; x detLc xI detnI J L xI
1n detLT J n xI

1n LT J; n x;

Materials and Methods

y_ DFs DHs
y;

Because is generally positive on the left side of the imaginary axis, we

consider only the networks whose Laplacian eigenvalues have a nonnegative
real part, which ensures that complete synchronization is possible.

[12]

where LT denotes the transpose of L. Eq. 11 then follows from the fact that
LT J; zn z LT ; zz whenever L has row sums equal to zero
(which is the case, because L is the Laplacian matrix of a network). To prove
this fact, we will use elementary row operations, which do not change the
determinants. First, we replace the first row of the matrix LT J zI by the
sum of all rows, making each entry in this row n z. Next, we subtract this
row multiplied by n z (which makes it a row with all entries equal to )
from the remaining rows canceling out the contribution from the term J in
these rows. We denote the resulting matrix by M1. Finally, we take the matrix
LT zI and replace the first row by the sum of all rows. This results in a matrix
M2 with exactly the same entries as M1 except for the first row. For this
matrix, this is a row with all entries equal to z rather than n z. Dividing
the first rows of M1 and M2 by n z and z, respectively, which will scale the
determinants accordingly, we see that

Nishikawa and Motter

the total link strength m is mapped to nn 1 m. In particular, this

implies that the complement of any optimal network is also optimal
m
.
if > nn1

As an immediate consequence of this result, the normalized standard

m
deviation changes to nn1m
under the complement transformation, while

ACKNOWLEDGMENTS. The authors thank Marian Anghel for discussions on the

power-grid network and Sara Solla for providing important references on
the role of inhibitory neurons. This work was supported by NSF under Grant
DMS-0709212.

1. Vicsek T, Czirk A, Ben-Jacob E, Shochet O (1995) Novel type of phase transition in a

system of self-driven particles. Phys Rev Lett 75:12261229.
2. Cucker F, Smale S (2007) Emergent behavior in flocks. IEEE T Automat Contr
52:852862.
3. Yang P, Freeman R, Lynch K (2008) Multi-agent coordination by decentralized
estimation and control. IEEE T Automat Contr 53:24802496.
4. Izhikevich E (2001) Synchronization of elliptic bursters. SIAM Rev 43:315344.
5. Taylor AF, Tinsley MR, Wang F, Huang Z, Showalter K (2009) Dynamical quorum sensing
and synchronization in large populations of chemical oscillators. Science 323:614617.
6. Grainger J, Stevenson W, Jr (1994) Power System Analysis (McGraw-Hill Co, Singapore).
7. Barrat A, Barthelemy M, Vespignani A (2008) Dynamical Processes on Complex Networks (Cambridge Univ Press, Cambridge, UK).
8. Strogatz SH (2001) Exploring complex networks. Nature 410:268276.
9. Nishikawa T, Motter AE (2006) Synchronization is optimal in non-diagonalizable
networks. Phys Rev E 73:065106(R).
10. Nishikawa T, Motter AE (2006) Maximum performance at minimum cost in network
synchronization. Physica D 224:7789.
11. Yin CY, Wang WX, Chen G, Wang BH (2006) Decoupling process for better
synchronizability on scale-free networks. Phys Rev E 74:047102.
12. Duan Z, Chen G, Huang L (2007) Complex network synchronizability: Analysis and
control. Phys Rev E 76:056103.
13. Gu Y, Sun J (2009) Altering synchronizability by adding and deleting edges for
scale-free networks. Physica A 388:32613267.
14. Hagberg A, Schult DA (2008) Rewiring networks for synchronization. Chaos 18:037105.
15. Zhi-Sheng D, Wen-Xu W, Chao L, Guan-Rong C (2009) Are networks with more edges
easier to synchronize, or not?. Chinese Phys B 18:31223130.
16. Nishikawa T, Motter AE, Lai YC, Hoppensteadt FC (2003) Heterogeneity in oscillator
networks: Are smaller worlds easier to synchronize?. Phys Rev Lett 91:014101.
17. Zhou C, Motter AE, Kurths J (2006) Universality in the synchronization of weighted
random networks. Phys Rev Lett 96:034101.
18. Motter AE, Zhou C, Kurths J (2005) Enhancing complex-network synchronization.
Europhys Lett 69:334340.
19. Motter AE, Zhou C, Kurths J (2005) Network synchronization, diffusion, and the
paradox of heterogeneity. Phys Rev E 71:016116.
20. Restrepo JG, Ott E, Hunt BR (2006) Characterizing the dynamical importance of
network nodes and links. Phys Rev Lett 97:094102.
21. Restrepo JG, Ott E, Hunt BR (2006) Synchronization in large directed networks of
coupled phase oscillators. Chaos 16:015107.
22. Bauer F, Atay FM, Jost J (2010) Synchronized chaos in networks of simple units.
Europhys Lett 89:20002.
23. Bragin A, et al. (1995) Gamma (40100 hz) oscillation in the hippocampus of the
behaving rat. J Neurosci 15:4760.
24. Whittington M, Traub R, Jefferys J (1995) Synchronized oscillations in interneuron networks driven by metabotropic glutamate receptor activation. Nature 373:612615.
25. Ylinen A, et al. (1995) Sharp wave-associated high-frequency oscillation (200 hz) in the
intact hippocampus: network and intracellular mechanisms. J Neurosci 15:3046.
26. Myers SA, Anghel M, Motter AE (2009) Spontaneous synchrony in power grids (Los
Alamos Natl Laboratory) Tech Rep LA-UR-09-02778.
27. McAdams HH, Shapiro L (2003) A bacterial cell-cycle regulatory network operating in
time and space. Science 301:18741877.
28. Pecora LM, Carroll TL (1998) Master stability functions for synchronized coupled
systems. Phys Rev Lett 80:21092112.
29. Dorogovtsev SN, Goltsev AV, Mendes JFF (2008) Critical phenomena in complex
networks. Rev Mod Phys 80:12751335.
30. Arenas A, Daz-Guilera A, Kurths J, Moreno Y, Zhou C (2008) Synchronization in
complex networks. Phys Rep 469:93153.

31. Holland MD, Hastings A (2008) Strong effect of dispersal network structure on
ecological dynamics. Nature 456:792794.
32. Kinouchi O, Copelli M (2006) Optimal dynamical range of excitable networks at
criticality. Nat Phys 2:348351.
33. Guclu H, Korniss G, Novotny MA, Toroczkai Z, Rcz Z (2006) Synchronization landscapes in small-world-connected computer networks. Phys Rev E 73:066115.
34. Korniss G, Novotny MA, Guclu H, Toroczkai Z, Rikvold PA (2003) Suppressing roughness
of virtual times in parallel discrete-event simulations. Science 299:677679.
35. Sun J, Bollt EM, Nishikawa T (2009) Master stability functions for coupled nearly
identical dynamical systems. Europhys Lett 85:60011.
36. Diestel R (2005) Graph Theory, Graduate Texts in Mathematics (Springer-Verlag,
Berlin), 3rd Ed, Vol 173.
37. Donetti L, Hurtado PI, Muoz MA (2005) Entangled networks, synchronization, and
optimal network topology. Phys Rev Lett 95:188701.
38. Motter AE (2007) Bounding network spectra for network design. New J Phys 9:182.
39. Belykh I, de Lange E, Hasler M (2005) Synchronization of bursting neurons: What
matters in the network topology. Phys Rev Lett 94:188101.
40. Denker M, Timme M, Diesmann M, Wolf F, Geisel T (2004) Breaking synchrony by
heterogeneity in complex networks. Phys Rev Lett 92:074103.
41. Barabsi AL, Albert R (1999) Emergence of scaling in random networks. Science
286:509512.
42. Newman MEJ, Strogatz SH, Watts DJ (2001) Random graphs with arbitrary degree
distributions and their applications. Phys Rev E 64:026118.
43. Son SW, Kim BJ, Hong H, Jeong H (2009) Dynamics and directionality in complex
networks. Phys Rev Lett 103:228702.
44. Belykh I, Belykh V, Hasler M (2006) Synchronization in asymmetrically coupled
networks with node balance. Chaos 16:015102.
45. Motter AE, Gulbahce N, Almaas E, Barabasi AL (2008) Predicting synthetic rescues in
metabolic networks. Mol Syst Biol 4:168.
46. Motter AE (2010) Improved network performance via antagonism: From synthetic
rescues to multi-drug combinations. BioEssays 32:236245.
47. Nishikawa T, Gulbahce N, Motter AE (2008) Spontaneous reaction silencing in metabolic optimization. PLoS Comput Biol 4:e1000236.
48. Belykh I, Shilnikov A (2008) When weak inhibition synchronizes strongly desynchronizing networks of bursting neurons. Phys Rev Lett 101:078102.
49. Vreeswijk C, Abbott LF, Bard Ermentrout G (1994) When inhibition not excitation
synchronizes neural firing. J Comput Neurosci 1:313321.
50. Wang XJ, Buzsaki G (1996) Gamma oscillation by synaptic inhibition in a hippocampal
interneuronal network model. J Neurosci 16:64026413.
51. Bourgeois J, Rakic P (1993) Changes of synaptic density in the primary visual cortex of
the macaque monkey from fetal to adult stage. J Neurosci 13:28012820.
52. Huttenlocher P (1979) Synaptic density in human frontal cortexdevelopmental
changes and effects of aging. Brain Res 163:195205.
53. Huttenlocher PR, Dabholkar AS (1997) Regional differences in synaptogenesis in
human cerebral cortex. J Comp Neurol 387:167178.
54. Bogu M, Pastor-Satorras R (2002) Epidemic spreading in correlated complex
networks. Phys Rev E 66:047104.
55. Motter AE (2004) Cascade control and defense in complex networks. Phys Rev Lett
93:098701.
56. Simonsen I, Buzna L, Peters K, Bornholdt S, Helbing D (2008) Transient dynamics
increasing network vulnerability to cascading failures. Phys Rev Lett 100:218701.
57. Kiss IZ, Rusin CG, Kori H, Hudson JL (2007) Engineering complex dynamical structures:
Sequential patterns and desynchronization. Science 316:18861889.
58. Huang L, Chen Q, Lai YC, Pecora LM (2009) Generic behavior of master-stability
functions in coupled nonlinear dynamical systems. Phys Rev E 80:036204.
59. Mohar B, Poljak S (1990) Eigenvalues and the max-cut problem. Czech Math J
40:343352.

Nishikawa and Motter

PNAS

June 8, 2010

vol. 107

no. 23

10347

APPLIED PHYSICAL
SCIENCES

LT J; z detLT J zI detM 1 detM 2

n z
n z
n z
z
T
T
detL zI
L ; z

:
[13]

z
z