Consider any network of n identical Kuramoto oscillators in which each oscillator is coupled bidirectionally with unit strength to at least μ(n1) other oscillators. Then, there is a critical value of μ above which the system is guaranteed to converge to the in-phase synchronous state for almost all initial conditions. The precise value of μ remains unknown. In 2018, Ling, Xu, and Bandeira proved that if each oscillator is coupled to at least 79.29% of all the others, global synchrony is ensured. In 2019, Lu and Steinerberger improved this bound to 78.89%. Here, we find clues that the critical connectivity may be exactly 75%. Our methods yield a slight improvement on the best known lower bound on the critical connectivity from 68.18% to 68.28%. We also consider the opposite end of the connectivity spectrum, where the networks are sparse rather than dense. In this regime, we ask how few edges one needs to add to a ring of n oscillators to turn it into a globally synchronizing network. We prove a partial result: all the twisted states in a ring of size n=2m can be destabilized by adding just O(nlog2n) edges. To finish the proof, one needs to rule out all other candidate attractors. We have done this for n8 but the problem remains open for larger n. Thus, even for systems as simple as Kuramoto oscillators, much remains to be learned about dense networks that do not globally synchronize and sparse ones that do.

The dynamics of a system of oscillators can be profoundly affected by its wiring diagram. For example, consider a network of n identical Kuramoto oscillators, all coupled to their neighbors with unit strength. The network is assumed to be connected but otherwise has arbitrary topology. Under what conditions can we be sure the oscillators will settle into a state of in-phase synchrony instead of another mode of organization? A few things are known: if the oscillators are densely coupled all-to-all, the system globally synchronizes (meaning it asymptotically approaches the in-phase state from almost all initial conditions). On the other hand, if the oscillators are arranged in a ring and sparsely coupled only to their nearest neighbors, the system can display other patterns besides in-phase synchrony; these patterns take the form of twisted states in which the oscillators’ phases differ by a constant amount from one to the next around the ring. To clarify how the topology of a network affects its propensity to synchronize, we analyze circulant networks of identical Kuramoto oscillators. For these highly symmetric networks, the linear stability of the twisted states can be solved completely. The results lead us to conjecture that global synchronization (sync) occurs in any network of identical Kuramoto oscillators in which each oscillator is connected to more than 75% of the others. At the sparse end of the connectivity spectrum, we provide evidence that a ring of n oscillators can be turned into a globally synchronizing network by adding as few as O(nlog2n) edges in the right places. These results could be potentially useful in the design of distributed systems of clocks, sensors, low-power radios, or other applications where many oscillators need to keep themselves in sync without any cues from a central timekeeper.

Coupled nonlinear oscillators often fall into sync spontaneously. Examples range from flashing fireflies and neural populations to arrays of Josephson junctions and nanoelectromechanical oscillators.1–7 On the theoretical side, many researchers have explored how the tendency to synchronize is affected by the network structure.8–22 This is the topic of the present paper.

We say that a system of oscillators globally synchronizes if it converges to a state with all the oscillators in phase, starting from all initial conditions except a set of measure zero. Intuitively, one expects that dense networks should favor global synchronization, whereas sparse networks might support waves and other patterns besides synchrony.

In 2012, Taylor11 explored these issues using the following system of identical Kuramoto oscillators,

dθjdt=ω+Kk=0n1Ajksin(θkθj),0jn1.
(1)

Here, θj(t) is the phase of oscillator j and ω is its natural frequency, which is assumed to be the same for all oscillators. The parameter K>0 represents the strength of the coupling. The connection topology is encoded by the adjacency matrix A with entries Ajk=Akj=1 if oscillator j is coupled to oscillator k, and Ajk=Akj=0 otherwise. Thus, all interactions are assumed to be symmetric and equally attractive. Without loss of generality, one can set ω=0 by going into a rotating frame with θjθj+ωt; then, one can further set K=1 by rescaling time. Hence, the governing equations boil down to

dθjdt=k=0n1Ajksin(θkθj),0jn1.
(2)

Since the adjacency matrix A is symmetric, (2) is a gradient system.8,9 Thus, all the attractors are equilibrium points, so to analyze the long-term global dynamics of (1), it suffices to analyze the local stability of equilibria of (2).

Having made these simplifications, Taylor11 proved that system (2) globally synchronizes if each oscillator is coupled to at least 93.95% of the others. This result raised a natural question: What is the critical connectivity for systems of identical Kuramoto oscillators? To make this notion precise, define the connectivityμ of a network of size n as the minimum degree of the nodes in the network, divided by n1, the total number of other nodes. Then, define the critical connectivityμc as the smallest value of μ such that any network of n oscillators is globally synchronizing if μμc; otherwise, for any μ<μc, at least one network exists with some other attractors besides the in-phase state. In these terms, Taylor’s result is μc0.9395.

The striking thing about Taylor’s result is that the structure of the network could be arbitrary in all other respects (random, regular, or in between). Recently, Ling et al.21 refined Taylor’s argument to show that 79.29% connectivity ensures global synchronization, and Lu and Steinerberger22 lowered the bound still further to 78.89%. On the other hand, competing attractors called twisted states9,15 can coexist with the synchronous state for certain networks whose connectivity is less than 68.18%. Thus, 0.6818μc0.7889.

In this paper, we use symmetric structures called circulant networks to sharpen the understanding of μc. We construct a family of networks (Fig. 1, top row) whose connectivity approaches 75% and which lie on the razor’s edge of being able to avoid global synchrony: along with the synchronous state, they have a competing, nearly stable, twisted state. In each case, that twisted state has all negative eigenvalues, apart from four zero eigenvalues which render linear analysis insufficient. Simulations show that, sadly, this twisted state is weakly unstable. We had hoped it would be stable because that would have implied μc0.75. At least we can say that any future attempt to show μc<0.75 by linear analysis (in the style of Taylor11 and his successors21,22) must now contend with this sequence of networks and its zero eigenvalues. Its near-miss quality leads us to conjecture that μc=0.75. There is also a surprise at the opposite end of the connectivity spectrum (Fig. 1, bottom row). Merely connecting each oscillator to a logarithmically small number of neighbors suffices to destabilize all the twisted states of a ring, thereby converting it (we conjecture) into a globally synchronizing network.

FIG. 1.

Dense circulant networks and sparse ones of size n=8, 16, and 32. Panels (a)–(c) show networks with a twisted state whose associated Jacobian matrices have all eigenvalues 0. These networks are tantalizingly close to not being globally synchronizing. We wonder if they could be tweaked to be truly so. Panels (d)–(f) show sparse circulant networks that have no stable twisted states. These are promising candidates for the sparsest circulant networks that globally synchronize but we cannot say that for certain because we have not ruled out the possible existence of other attractors.

FIG. 1.

Dense circulant networks and sparse ones of size n=8, 16, and 32. Panels (a)–(c) show networks with a twisted state whose associated Jacobian matrices have all eigenvalues 0. These networks are tantalizingly close to not being globally synchronizing. We wonder if they could be tweaked to be truly so. Panels (d)–(f) show sparse circulant networks that have no stable twisted states. These are promising candidates for the sparsest circulant networks that globally synchronize but we cannot say that for certain because we have not ruled out the possible existence of other attractors.

Close modal

Circulant networks turn out to be a rich source of graphs for our purposes. A circulant network is a graph whose adjacency matrix A has constant diagonals such that each row is a circularly shifted version of the preceding row, i.e.,

A=[a0a1a2a1a1a0a1a2a1a0a2a1a1a2a1a0].

The matrix A is prescribed by selecting the values of a0,a1,,an/2 as either 0 or 1. We assume throughout that a0=0 so that no oscillator is coupled to itself.

The eigenvalues of any circulant matrix (with a0=0) are known analytically23 and are given by

λp(A)=s=1n1ase2πips/n,0pn1,
(3)

where as=ans for n/2<s<n. The reflectional symmetry as=ans implies that the eigenvalues in (3) are real and λp(A)=λnp(A) for 1pn1. Moreover, the eigenvectors of A are given by23 

Av_p=λp(A)v_p,0pn/2,Aw_p=λp(A)w_p,1pn/21,
(4)

where

v_p=[1cos(2πpn)cos(2πp(n1)n)AAAAAAA],w_p=[0sin(2πpn)sin(2πp(n1)n)AAAAAAA].

Several equilibrium points of (2) can be derived from linear algebra when A is a circulant matrix. Let θ_=(θ0,,θn1), c_=cos(θ_), and s_=sin(θ_), where the superscript denotes the vector transpose. Then, since sin(xy)=sin(x)cos(y)cos(x)sin(y), a vector θ_ is an equilibrium point of (2) if and only if

Dc_As_=Ds_Ac_,Dc_=diag(c_),Ds_=diag(s_),

where diag(c_) is a diagonal matrix with c_ on the diagonal. Due to the structure of (2), if θ_ is an equilibrium point, then it remains an equilibrium point if all its arguments are shifted by the same angle Θ. To avoid these trivial rotations, we set θ0=0. Under this restriction, we find from (4) that

θ_(p)=(0,2πpn,,2πp(n1)n)
(5)

is an equilibrium point of (2) for any 0pn/2. When p=0, θ_(0)=(0,,0), and we call this the synchronous state, corresponding to all the oscillators being in phase. We call (5) a twisted state for 1pn/2 as such a state physically corresponds to arranging the oscillators so that their phases differ by a constant amount from one oscillator to the next, twisting uniformly through p full revolutions as we circulate once around the network. For any circulant network, the twisted states are always equilibrium points but there are usually other equilibria too.15 

To investigate the stability of the twisted states, we take a look at the Jacobian associated to (2). It is given by, for 0pn/2,

(Jp)jk={Ajkcos(θk(p)θj(p)),jk,s=0n1Ajscos(θs(p)θj(p)),j=k.

It can be verified that Jp is a symmetric circulant matrix for 0pn/2. This means that the eigenvalues of Jp at the equilibrium points in (5) can be analytically derived. For any 0pn/2, we have23 

λr(Jp)=s=1n1ascos(pts)[1+cos(rts)],
(6)

where ts=2πs/n. The stability of the twisted state (5) depends on the signs of λ0(Jp),,λn1(Jp). If one of these numbers is positive, then θ_(p) is unstable. The matrix Jp will always have at least one zero eigenvalue because of the trivial rotations by Θ mentioned earlier; if all other eigenvalues are negative, then θ_(p) is stable.

To illustrate the general theory, we begin by applying it to two familiar examples of circulant networks.

Consider a ring of oscillators with nearest-neighbor coupling so that a1=1 and ak=0 for 2kn/2. We find that

λr(Jp)=2cos(2πpn)[1+cos(2πrn)].

Hence, θ_(p) is unstable if and only if the twist p satisfies n/4<pn/2.

This is the network with the adjacency matrix given by a1==a=1, where 1<n/2. In other words, each oscillator on a ring is connected to its nearest neighbors on either side. Trigonometric identities yield

λr(Jp)=2s=1cos(2πpsn)[1+cos(2πrsn)]=12U2(xp+r)+12U2(xpr)U2(xp),

where U2 is the degree 2 Chebyshev polynomial of the second kind (see Table 18.3.1 in Olver et al.24) and xp=cos(pπ/n). From this formula, we can numerically verify that there is a sequence of networks with connectivity tending to 68.09% for which the 1-twist is a stable equilibrium. This matches the results derived in Wiley et al.9 Canale and Monzón15 formed a sequence of networks from the lexicographic product of a WSG network together with complete graphs to show that μc15/220.6818.

Having reviewed circulant networks, we now use them to look for dense networks that have other attractors besides perfect synchrony and which can, therefore, avoid having to globally synchronize. Every vertex in a circulant network G has the same degree, δG, so an equivalent aim is to maximize δG/(n1) over all networks that fail to globally synchronize. To do this, we conducted a numerical search over all circulant networks of size 5n50 having at least one stable twisted state (for which all nontrivial eigenvalues are negative). The blue dots of Fig. 2 show the maximum value of δG/(n1) for each n. They all lie below the best known lower bound15 of 15/220.6818.

FIG. 2.

Connectivity of the densest circulant networks with a twisted state that is either stable or weakly unstable for 5n50. The blue dots indicate networks with at least one stable twisted state. (All the nontrivial eigenvalues are negative.) The red squares show a sequence of dense circulant networks with a twisted state that has all negative eigenvalues apart from four zero eigenvalues. The dashed line is a rigorous bound for each n above which no stable states (of any kind) exist apart from synchrony; this conclusion follows from Theorem 3.1 of Ling et al.21 

FIG. 2.

Connectivity of the densest circulant networks with a twisted state that is either stable or weakly unstable for 5n50. The blue dots indicate networks with at least one stable twisted state. (All the nontrivial eigenvalues are negative.) The red squares show a sequence of dense circulant networks with a twisted state that has all negative eigenvalues apart from four zero eigenvalues. The dashed line is a rigorous bound for each n above which no stable states (of any kind) exist apart from synchrony; this conclusion follows from Theorem 3.1 of Ling et al.21 

Close modal

But, notice the red squares that are far above the blue dots. These represent networks perched on the razor’s edge of being able to avoid global synchrony. This sequence of remarkable networks squeezes up to 75% connectivity as n. Half of these networks have size n=8m and the other half have size n=8m+4. (The stability properties for these two cases are similar so from now on we restrict our attention to n=8m.)

For n=8m, consider the circulant network G such that aj=1 if and only if mod(j,4)2 for 1j4m. The degree of each vertex is δG=6m1. The twisted state θ_(2m) in (5) has no positive eigenvalues for this network. To see this, use (6) and the identity 1+cos(2x)=2sin2(x) to obtain

λr(J2m)=2s=12m1cos(2πs)sin2(πsr2m)2s=02m1cos(2πs+π/2)sin2(π(4s+1)r8m)2s=02m1cos(2πs+3π/2)sin2(π(4s+3)r8m).

Noting that cos(2πs+π/2)=cos(2πs+3π/2)=0 and cos(2πs)=1 for any integer s, we conclude that

λr(J2m)=2s=12m1sin2(πsr2m)0,0rn1.

This means that J2m has all negative eigenvalues apart from four that vanish (corresponding to r=0,2m,4m,6m), as promised earlier. We have found a dense network of size 8m where the stability or instability of the 2m-twist cannot be determined by the eigenvalues of the Jacobian. Unfortunately, numerical experiments suggest that these twisted states are weakly (nonlinearly) unstable, and so the system does not manage to avoid global synchrony; otherwise, this sequence would have dramatically improved the lower bound on μc from 0.6818 to 0.75.

Figure 3 shows the consequences of this weak instability. For initial conditions starting within an ε-neighborhood of the 2m-twisted state, it takes an algebraically long time of O(ε1) to escape that neighborhood before settling into the synchronous state. This super-slow escape time hints that these networks are poised at criticality. Perhaps, just a small tweak—in the network topology, the sine coupling function, or the edge weights—would be enough to nudge the system away from being globally synchronizing. For example, one may be able to avoid global synchrony by adding or removing just a handful of edges. In any case, this sequence of networks on the razor’s edge leads us to conjecture that μc=0.75.

FIG. 3.

Slow escape from a weakly unstable twisted state for n=8. We integrated (2) starting from random initial conditions within a ball of size ε about the twisted state (5) for p=2. For each ε, we repeated the simulation 100 times, to gather statistics, and computed the time taken to reach the in-phase state to within machine precision.

FIG. 3.

Slow escape from a weakly unstable twisted state for n=8. We integrated (2) starting from random initial conditions within a ball of size ε about the twisted state (5) for p=2. For each ε, we repeated the simulation 100 times, to gather statistics, and computed the time taken to reach the in-phase state to within machine precision.

Close modal

To sidestep the subtleties associated with the presence of zero eigenvalues, let us now restrict attention to circulant graphs with at least one certifiably stable twisted state; here, stable means that all its nontrivial eigenvalues lie strictly in the left half plane. For circulant graphs of size 5n50, the densest such graph is found at n=44 with δG/(n1)0.6744. This graph does not globally synchronize, of course, since it has a competing attractor (the stable twisted state), and it allows us to construct a sequence of even denser graphs with the same property, as follows. Given any non-globally synchronizing graph G and the complete graph Kτ on τ nodes (for any integer τ), Canale and Monzón15 proved that the lexicographic product G[Kτ] is also non-globally synchronizing. Intuitively, the lexicographic product of G and Kτ is formed by replacing each node in G by a clique of size τ; the nodes in different cliques are then connected in the same way as the parent nodes they replaced. We call this process twinning. To be more precise, the vertex set of the twinned graph G[Kτ] is defined as the Cartesian product G×Kτ, and any two vertices (u,v) and (x,y) are adjacent if and only if either u is adjacent with x in G or u=x. The twinned graph G=G[Kτ] is denser than G because δG=τ1+τδG and δG/(τn1)>δG/(n1) for τ>1. Thus, given any non-globally synchronizing graph G, one can use this idea to construct a sequence of non-globally synchronizing graphs with a limiting edge density of (δG+1)/n. Canale and Monzón15 used this to show that the critical connectivity μc is bounded below by 15/220.6818.

Unfortunately, we have not been able to improve on the lower bound with just twinning. However, with the help of a computer search, we have found a more elaborate four-step “twinning plus adding” construction that slightly improves the lower bound to 0.6828: (i) take the n=22 WSG network with degree 14, where the 1-twist is observed to be stable, based on its eigenvalues. We will denote this graph by G22 and label its vertices v1,,v22 so that v1 is connected to v2,,v8. (ii) Now, form the graph G22[K85] by twinning. This is a non-globally synchronizing graph of size n=1870 with degree 1274. This graph is just like G22 except vertex vk is now a connected cluster Ck of size 85. (iii) Pair up vertices from C1 with vertices from C9 (so that every vertex is paired up precisely once). Repeat this process with C2 and C10, C3 and C11, and so on, up to C22 and C8. Each vertex gets two additional edges, and the graph is still circulant. We now have a circulant graph G of size n=1870 with degree 1276. By looking at the eigenvalues of the associated Jacobian, the graph is observed to remain non-globally synchronizing (with the same stable equilibrium state from G22[K85]). (iv) One can finally consider the sequence G[K1],G[K2], of non-globally synchronizing twinned graphs. The connectivity of this sequence of graphs tends to 1277/18700.6828. (A computer search shows that n=22 and τ=85 are the best parameters in the range of 5n50 for this construction.)

As a counterpoint to the search for dense networks that fail to globally synchronize, we have also looked for sparse networks that do globally synchronize. Figure 4 shows, for 5n100, the sparsest circulant networks that lack any stable twisted state as revealed by a brute-force numerical search. The trend in Fig. 4 is δG=O(log2n). Hence, the connectivity δG/(n1) of these networks tends to zero as n.

FIG. 4.

Connectivity of the sparsest circulant networks for which all the twisted states are unstable. The trend observed here led us to a particular sequence of sparse circulant networks, all of which appear to be globally synchronizing.

FIG. 4.

Connectivity of the sparsest circulant networks for which all the twisted states are unstable. The trend observed here led us to a particular sequence of sparse circulant networks, all of which appear to be globally synchronizing.

Close modal

To understand this logarithmic trend, consider the circulant network of size n=2m, where aj=1 for all powers of 2 less than or equal to 2m1 (Fig. 1, bottom row). The other aj’s are zero or determined by the symmetry condition as=ans for 1sn/2. This network is constructed by connecting a logarithmically small number of neighbors to each vertex of a ring. The final network has a total of (2log2n1)n/2 edges.

All the twisted states θ_(1),,θ_(n1) are unstable for this network. To prove this, we need to show that each Jacobian J1,,Jn/2 has at least one strictly positive eigenvalue. Consider the pth eigenvalue of Jp. By (6) and the fact that cos(x)[1+cos(x)]=[12cos(x)+cos(2x)]/2, we find that

λp(Jp)=m+k=0m2cos(2πp2k+12m)2k=0m2cos(2πp2k2m)cos(pπ).

This expression can be simplified to

λp(Jp)=mk=1m2cos(2πp2k2m)2cos(2πp2m)m(m2)2=0,
(7)

where we can only hope to have an equality in (7) if cos(2πp/2m)=1, i.e., if 2mp is an integer. Since 0p2m1, we conclude that cos(2πp/2m)<1 for 1p2m1. In other words, λp(Jp)>0 for 1p2m1. Thus, all the twisted states are unstable.

The network we have just constructed is very sparse. We conjecture that if a circulant network is even more sparse, namely, δGlog2n as n, then at least one of the twisted states must become stable. We go further and conjecture that our sparse circulant examples not only lack stable twisted states but also they lack competing attractors of any kind. Proving this conjecture for all n=2m seems difficult because it requires checking that synchrony is the only stable state of (2). The conjecture holds for n=2 and n=4 because then the networks are complete graphs, which are known to be globally synchronizing.5 For n=8, we used the computational algebraic geometry package Macaulay225 to confirm that there are 262 real isolated equilibrium points of (2) and all these equilibria are unstable except for the synchronous state. (There are an additional 1008 isolated complex-valued equilibria, and also 5 continuous families of unstable equilibria.) For larger n, the number of real equilibria grows exponentially. For n=16, there are at least 32768 real equilibria but there could be as many as 1073741824 complex-valued ones. The lesson is that a brute-force search quickly becomes computationally hopeless. A new idea is needed.

Much remains to be learned about the conditions that ensure global synchrony, as well as the conditions that allow coupled oscillators to avoid it. Even in the simplified setting of Kuramoto oscillators, many fascinating problems remain about the impact of network topology on global synchronization. We have highlighted two such problems in the conjectures mentioned above. Similar questions could be investigated for other kinds of oscillators, such as Kuramoto–Sakaguchi oscillators, van der Pol oscillators, Stuart–Landau oscillators, and so on.

The study of global synchronization can also have practical implications. Existing theorems for a different class of oscillators (pulse-coupled oscillators4) have been applied to sensor networks and distributed clock synchronization26–29 and to the design and implementation of ultra-low-power radio systems30 with potential uses that include intrusion detection, keeping track of workers in coal mines, personal health care, and automated drug-delivery.27 In the same vein, the results presented here for Kuramoto oscillators may open up additional ways to build useful networks that reliably keep themselves in sync.

We thank Heather Harrington for discussions of the algebraic geometry computations; Mateo Diaz for informing us of the work of Ling et al.;21 Eduardo Canale for pointing out the weak instability of the critical twisted state in our dense circulant networks; and Lee DeVille, Gokul Nair, and Heather Wilber for helpful discussions. Research was supported by the National Science Foundation (NSF) as follows: NSF Grant No. DMS-1818757 to A.T., NSF Grant No. DMS-1502294 to M.S., and NSF Grant Nos. DMS-1513179 and CCF-1522054 to S.H.S.

The data that support the findings of this study are almost all available within the article. Any data that are not available can be found from the corresponding author upon reasonable request.

1.
A. T.
Winfree
, “
Biological rhythms and the behavior of populations of coupled oscillators
,”
J. Theor. Bio.
16
(
1
),
15
42
(
1967
).
2.
C. S.
Peskin
, “
Mathematical aspects of heart physiology
,”
Courant Institute Lecture Notes
(
Courant Institute of Mathematical Sciences
,
1975
), pp.
268
278
, available at https://www.math.nyu.edu/faculty/peskin/heartnotes/index.html.
3.
Y.
Kuramoto
,
Chemical Oscillations, Waves, and Turbulence
(
Springer
,
1984
).
4.
R. E.
Mirollo
and
S. H.
Strogatz
, “
Synchronization of pulse-coupled biological oscillators
,”
SIAM J. Appl. Math.
50
(
6
),
1645
1662
(
1990
).
5.
S.
Watanabe
and
S. H.
Strogatz
, “
Constants of motion for superconducting Josephson arrays
,”
Physica D
74
(
3–4
),
197
253
(
1994
).
6.
A.
Pikovsky
,
M.
Rosenblum
, and
J.
Kurths
,
Synchronization: A Universal Concept in Nonlinear Sciences
(
Cambridge University Press
,
2003
), Vol. 12.
7.
M. H.
Matheny
,
J.
Emenheiser
,
W.
Fon
,
A.
Chapman
,
A.
Salova
,
M.
Rohden
,
J.
Li
,
M. H.
de Badyn
,
M.
Pósfai
,
L.
Duenas-Osorio
et al., “
Exotic states in a simple network of nanoelectromechanical oscillators
,”
Science
363
(
6431
),
eaav7932
(
2019
).
8.
A.
Jadbabaie
,
N.
Motee
, and
M.
Barahona
, “On the stability of the Kuramoto model of coupled nonlinear oscillators,” in Proceedings of the 2004 American Control Conference (IEEE, 2004), Vol. 5, pp. 4296–4301.
9.
D. A.
Wiley
,
S. H.
Strogatz
, and
M.
Girvan
, “
The size of the sync basin
,”
Chaos
16
(
1
),
015103
(
2006
).
10.
E.
Mallada
and
A.
Tang
, “Synchronization of phase-coupled oscillators with arbitrary topology,” in Proceedings of the 2010 American Control Conference (IEEE, 2010), pp. 1777–1782.
11.
R.
Taylor
, “
There is no non-zero stable fixed point for dense networks in the homogeneous Kuramoto model
,”
J. Phys. A Math. Theor.
45
(
5
),
055102
(
2012
).
12.
F.
Dörfler
and
F.
Bullo
, “
Synchronization in complex networks of phase oscillators: A survey
,”
Automatica
50
(
6
),
1539
1564
(
2014
).
13.
L. M.
Pecora
,
F.
Sorrentino
,
A. M.
Hagerstrom
,
T. E.
Murphy
, and
R.
Roy
, “
Cluster synchronization and isolated desynchronization in complex networks with symmetries
,”
Nat. Comm.
5
,
4079
(
2014
).
14.
A.
Pikovsky
and
M.
Rosenblum
, “
Dynamics of globally coupled oscillators: Progress and perspectives
,”
Chaos
25
(
9
),
097616
(
2015
).
15.
E. A.
Canale
and
P.
Monzón
, “
Exotic equilibria of Harary graphs and a new minimum degree lower bound for synchronization
,”
Chaos
25
(
2
),
023106
(
2015
).
16.
D.
Mehta
,
N. S.
Daleo
,
F.
Dörfler
, and
J. D.
Hauenstein
, “
Algebraic geometrization of the Kuramoto model: Equilibria and stability analysis
,”
Chaos
25
(
5
),
053103
(
2015
).
17.
F. A.
Rodrigues
,
T. K.
DM. Peron
,
P.
Ji
, and
J.
Kurths
, “
The Kuramoto model in complex networks
,”
Phys. Rep.
610
,
1
98
(
2016
).
18.
D. M.
Abrams
,
L. M.
Pecora
, and
A. E.
Motter
, “
Introduction to focus issue: Patterns of network synchronization
,”
Chaos
26
(
9
),
094601
(
2016
).
19.
L.
DeVille
and
B.
Ermentrout
, “
Phase-locked patterns of the Kuramoto model on 3-regular graphs
,”
Chaos
26
(
9
),
094820
(
2016
).
20.
Y.
Sokolov
and
G. B.
Ermentrout
, “When is sync globally stable in sparse networks of identical Kuramoto oscillators?,” arXiv:1812.08153 (2018).
21.
S.
Ling
,
R.
Xu
, and
A. S.
Bandeira
, “On the landscape of synchronization networks: A perspective from nonconvex optimization,” arXiv:1809.11083 (2018).
22.
J.
Lu
and
S.
Steinerberger
, “Synchronization of Kuramoto oscillators in dense networks,” arXiv:1911.12336 (2019).
23.
G. H.
Golub
and
C. F.
Van Loan
,
Matrix Computations
(
JHU Press
,
2012
), Vol. 3.
24.
F. W. J.
Olver
,
D. W.
Lozier
,
R. F.
Boisvert
, and
C. W.
Clark
,
NIST Handbook of Mathematical Functions Hardback and CD-ROM
(
Cambridge University Press
,
2010
).
25.
D. R.
Grayson
and
M. E.
Stillman
, see http://www.math.uiuc.edu/Macaulay2/ for “Macaulay2, A Software System for Research in Algebraic Geometry.”
26.
O.
Simeone
,
U.
Spagnolini
,
Y.
Bar-Ness
, and
S. H.
Strogatz
, “
Distributed synchronization in wireless networks
,”
IEEE Sig. Proc. Mag.
25
(
5
),
81
97
(
2008
).
27.
J.
Yick
,
B.
Mukherjee
, and
D.
Ghosal
, “
Wireless sensor network survey
,”
Comput. Netw.
52
(
12
),
2292
2330
(
2008
).
28.
Y.-W.
Hong
and
A.
Scaglione
, “
A scalable synchronization protocol for large scale sensor networks and its applications
,”
IEEE J. Sel. Areas Comm.
23
(
5
),
1085
1099
(
2005
).
29.
G.
Werner-Allen
,
G.
Tewari
,
A.
Patel
,
M.
Welsh
, and
R.
Nagpal
, “Firefly-inspired sensor network synchronicity with realistic radio effects,” in Proceedings of the 3rd International Conference Embedded Network Sensor Systems (ACM, 2005), pp. 142–153.
30.
R. K.
Dokania
,
X. Y.
Wang
,
S. G.
Tallur
, and
A. B.
Apsel
, “
A low power impulse radio design for body-area-networks
,”
IEEE Trans. Circuits Syst. I Regul. Pap.
58
(
7
),
1458
1469
(
2011
).