BOLYAI SOCIETY
MATHEMATICAL STUDIES, 2
Combinatorics,
Paul Erdős is Eighty (Volume 2)
Keszthely (Hungary), 1993, pp. 1–46.
Random Walks on Graphs: A Survey
L. LOVÁSZ
Dedicated to the marvelous random walk
of Paul Erdős
through universities, continents, and mathematics
Various aspects of the theory of random walks on graphs are surveyed. In
particular, estimates on the important parameters of access time, commute time,
cover time and mixing time are discussed. Connections with the eigenvalues
of graphs and with electrical networks, and the use of these connections in
the study of random walks is described. We also sketch recent algorithmic
applications of random walks, in particular to the problem of sampling.
0. Introduction
Given a graph and a starting point, we select a neighbor of it at random, and
move to this neighbor; then we select a neighbor of this point at random,
and move to it etc. The (random) sequence of points selected this way is a
random walk on the graph.
A random walk is a finite Markov chain that is time-reversible (see
below). In fact, there is not much difference between the theory of random
walks on graphs and the theory of finite Markov chains; every Markov chain
can be viewed as random walk on a directed graph, if we allow weighted
edges. Similarly, time-reversible Markov chains can be viewed as random
walks on undirected graphs, and symmetric Markov chains, as random walks
on regular symmetric graphs. In this paper we’ll formulate the results in
terms of random walks, and mostly restrict our attention to the undirected
case.
2
L. Lovász
Random walks arise in many models in mathematics and physics. In
fact, this is one of those notions that tend to pop up everywhere once you
begin to look for them. For example, consider the shuffling of a deck of
cards. Construct a graph whose nodes are all permutations of the deck, and
two of them are adjacent if they come by one shuffle move (depending on
how you shuffle). Then repeated shuffle moves correspond to a random walk
on this graph (see Diaconis [20]). The Brownian motion of a dust particle
is random walk in the room. Models in statistical mechanics can be viewed
as random walks on the set of states.
The classical theory of random walks deals with random walks on simple, but infinite graphs, like grids, and studies their qualitative behaviour:
does the random walk return to its starting point with probability one? does
it return infinitely often? For example, Pólya (1921) proved that if we do a
random walk on a d-dimensional grid, then (with probability 1) we return to
the starting point infinitely often if d = 2, but only a finite number of times
if d ≥ 3. See Doyle and Snell [25]; for more recent results on random walks
on infinite graphs, see also Thomassen [65].
More recently, random walks on more general, but finite graphs have
received much attention, and the aspects studied are more quantitative:
how long we have to walk before we return to the starting point? before we
see a given node? before we see all nodes? how fast does the distribution
of the walking point tend to its limit distribution?
As it turns out, the theory of random walks is very closely related
to a number of other branches of graph theory. Basic properties of a
random walk are determined by the spectrum of the graph, and also by
electrical resistance of the electric network naturally associated with graphs.
There are a number of other processes that can be defined on a graph,
mostly describing some sort of “diffusion” (chip-firing, load-balancing in
distributed networks etc.), whose basic parameters are closely tied with the
above-mentioned parameters of random walks. All these connections are
very fruitful and provide both tools for the study and opportunities for
applications of random walks. However, in this survey we shall restrict our
attention to the connections with eigenvalues and electrical networks.
Much of the recent interest in random walks is motivated by important
algorithmic applications. Random walks can be used to reach “obscure”
parts of large sets, and also to generate random elements in large and
complicated sets, such as the set of lattice points in a convex body or
the set of perfect matchings in a graph (which, in turn, can be used to
Random Walks on Graphs: A Survey
3
the asymptotic enumeration of these objects). We’ll survey some of these
applications along with a number of more structural results.
We mention three general references on random walks and finite Markov
chains: Doyle and Snell [25], Diaconis [20] and the forthcoming book of
Aldous [3].
Acknowledgement. My thanks are due to Peter Winkler, András Lukács
and Andrew Kotlov for the careful reading of the manuscript of this paper,
and for suggesting many improvements.
1. Basic notions and facts
Let G = (V, E) be a connected graph with n nodes and m edges. Consider
a random walk on G: we start at a node v0 ; if at the t-th step we are at
a node vt , we move neighbor of vt with probability 1/d(vt ). Clearly, the
sequence of random nodes (vt : t = 0, 1, . . .) is a Markov chain. The node
v0 may be fixed, but may itself be drawn from some initial distribution P0 .
We denote by Pt the distribution of vt :
Pt (i) = Prob(vt = i).
We denote by M = (pij )i,j∈V the matrix of transition probabilities of
this Markov chain. So
½
1/d(i), if ij ∈ E,
(1.1)
pij =
0,
otherwise.
Let AG be the adjacency matrix of G and let D denote the diagonal matrix
with (D)ii = 1/d(i), then M = DAG . If G is d-regular, then M = (1/d)AG .
The rule of the walk can be expressed by the simple equation
Pt+1 = M T Pt ,
(the distribution of the t-th point is viewed as a vector in RV ), and hence
Pt = (M T )t P0 .
It follows that the probability ptij that, starting at i, we reach j in t steps
is given by the ij-entry of the matrix M t .
4
L. Lovász
If G is regular, then this Markov chain is symmetric: the probability of
moving to u, given that we are at node v, is the same as the probability of
moving to node v, given that we are at node u. For a non-regular graph G,
this property is replaced by time-reversibility: a random walk considered
backwards is also a random walk. More exactly, this means that if we look
at all random walks (v0 , . . . , vt ), where v0 is from some initial distribution
P0 , then we get a probability distribution Pt on vt . We also get a probability
distribution Q on the sequences (v0 , . . . , vt ). If we reverse each sequence,
we get another probability distribution Q′ on such sequences. Now timereversibility means that this distribution Q′ is the same as the distribution
obtained by looking at random walks starting from the distribution Pt .
(We’ll formulate a more handy characterization of time-reversibility a little
later.)
The probability distributions P0 , P1 , . . . are of course different in general. We say that the distribution P0 is stationary (or steady-state) for the
graph G if P1 = P0 . In this case, of course, Pt = P0 for all t ≥ 0; we call
this walk the stationary walk.
A one-line calculation shows that for every graph G, the distribution
π(v) =
d(v)
2m
is stationary. In particular, the uniform distribution on V is stationary if the
graph is regular. It is not difficult to show that the stationary distribution
is unique (here one has to use that the graph is connected).
The most important property of the stationary distribution is that
if G is non-bipartite, then the distribution of vt tends to a stationary
distribution, as t → ∞ (we shall see a proof of this fact, using eigenvalues,
a little later). This is not true for bipartite graphs if n > 1, since then the
distribution Pt is concentrated on one color class or the other, depending
on the parity of t.
In terms of the stationary distribution, it is easy to formulate the
property of time-reversibility: it is equivalent to saying that for every pair
i, j ∈ V , π(i)pij = π(j)pji . This means that in a stationary walk, we step as
often from i to j as from j to i. From (1.1), we have π(i)pij = 1/(2m) for
ij ∈ E, so we see that we move along every edge, in every given direction,
with the same frequency. If we are sitting on an edge and the random
walk just passed through it, then the expected number of steps before it
passes through it in the same direction again is 2m. There is a similar
fact for nodes: if we are sitting at a node i and the random walk just
Random Walks on Graphs: A Survey
5
visited this node, then the expected number of steps before it returns is
1/π(i) = 2m/d(i). If G is regular, then this “return time” is just n, the
number of nodes.
2. Main parameters
We now formally introduce the measures of a random walk that play the
most important role in the quantitative theory of random walks, already
mentioned in the introduction.
(a) The access time or hitting time Hij is the expected number of steps
before node j is visited, starting from node i. The sum
κ(i, j) = H(i, j) + H(j, i)
is called the commute time: this is the expected number of steps in a
random walk starting at i, before node j is visited and then node i is
reached again. There is also a way to express access times in terms of
commute times, due to Tetali [63]:
Ã
!
X
1
κ(i, j) +
π(u)[κ(u, j) − κ(u, i)] .
(2.1)
H(i, j) =
2
u
This formula can be proved using either eigenvalues or the electrical
resistance formulas (sections 3 and 4).
(b) The cover time (starting from a given distribution) is the expected
number of steps to reach every node. If no starting node (starting
distribution) is specified, we mean the worst case, i.e., the node from
which the cover time is maximum.
(c) The mixing rate is a measure of how fast the random walk converges to
its limiting distribution. This can be defined as follows. If the graph is
(t)
non-bipartite, then pij → dj /(2m) as t → ∞, and the mixing rate is
¯
¯
¯ (t)
dj ¯¯1/t
¯
µ = lim sup max ¯pij −
.
i,j
2m ¯
t→∞
(For a bipartite graph with bipartition {V1 , V2 }, the distribution of
vt oscillates between “almost proportional to the degrees on V1′′ and
6
L. Lovász
“almost proportional to the degrees on V2′′ . The results for bipartite
graphs are similar, just a bit more complicated to state, so we ignore
this case.)
One could define the notion of “mixing time” as the number of steps
before the distribution of vt will be close to uniform (how long should
we shuffle a deck of cards?). This number will be about (log n)/(1 − µ).
However, the exact value depends on how (in which distance) the phrase
“close” is interpreted, and so we do not introduce this formally here. In
section 5 we will discuss a more sophisticated, but “canonical” definition of
mixing time.
The surprising fact, allowing the algorithmic applications mentioned
in the introduction, is that this “mixing time” may be much less than
the number of nodes; for an expander graph, for example, this takes only
O(log n) steps!
Example 1. To warm up, let us determine the access time for two points
of a path on nodes 0, . . . , n − 1.
First, observe that the access time H(k − 1, k) is one less than the
expected return time of a random walk on a path with k + 1 nodes, starting
at the last node. As remarked, this return time is 2k, so H(k−1, k) = 2k−1.
Next, consider the access times H(i, k) where 0 ≤ i < k ≤ n. In order
to reach k, we have to reach node k−1; this takes, on the average, H(i, k−1)
steps. From here, we have to get to k, which takes, on the average, 2k − 1
steps (the nodes beyond the k-th play no role). This yields the recurrence
H(i, k) = H(i, k − 1) + 2k − 1,
whence H(i, k) = (2i + 1) + (2i + 3) + . . . + (2k − 1) = k 2 − i2 . In particular,
H(0, k) = k 2 (this formula is closely√related to the well-known fact that
Brownian motion takes you distance t in t time).
Assuming that we start from 0, the cover time of the path on n nodes
will also be (n − 1)2 , since it suffices to reach the other endnode. The reader
might find it entertaining to figure out the cover time of the path when
starting from an internal node.
¿From this it is easy to derive that the access time between two nodes
at distance k of a circuit of length n is k(n − k). To determine the cover
time f (n) of the circuit, note that it is the same as the time needed on a
very long path, starting from the midpoint, to reach n nodes. Now we have
to reach first n − 1 nodes, which takes f (n − 1) steps on the average. At this
7
Random Walks on Graphs: A Survey
point, we have a subpath with n−1 nodes covered, and we are sitting at one
of its endpoints. To reach a new node means to reach one of the endnodes
of a path with n + 1 nodes from a neighbor of an endnode. Clearly, this is
the same as the access time between two consecutive nodes of a circuit of
length n. This leads to the recurrence
f (n) = f (n − 1) + (n − 1),
and through this, to the formula f (n) = n(n − 1)/2.
Example 2. As another example, let us determine the access times and
cover times for a complete graph on nodes {0, . . . , n − 1}. Here of course we
may assume that we start from 0, and to find the access times, it suffices
to determine H(0, 1). The probability that we first reach node 1 in the t-th
´t−1
³
1
n−2
step is clearly n−1
n−1 , and so the expected time this happens is
H(0, 1) =
¶
∞ µ
X
n − 2 t−1 1
t
= n − 1.
n−1
n−1
t=1
The cover time for the complete graph is a little more interesting, and
is closely related to the so-called Coupon Collector Problem (if you want to
collect each of n different coupons, and you get every day a random coupon
in the mail, how long do you have to wait?). Let τi denote the first time
when i vertices have been visited. So τ1 = 0 < τ2 = 1 < τ3 < . . . < τn .
Now τi+1 − τi is the number of steps while we wait for a new vertex to occur
— an event with probability (n − i)/(n − 1), independently of the previous
steps. Hence
n−1
E(τi−1 − τi ) =
,
n−i
and so the cover time is
E(τn ) =
n−1
X
i=1
E(τi+1 − τi ) =
n−1
X
i=1
n−1
≈ n log n.
n−i
A graph with particularly bad random walk properties is obtained by
taking a clique of size n/2 and attach to it an endpoint of a path of length
n/2. Let i be any node of the clique and j, the “free” endpoint of the path.
Then
H(i, j) = Ω(n3 ).
8
L. Lovász
In fact, starting from i, it takes, on the average, n/2 − 1 moves to reach
the attachment node u; then with probability 1 − 2/n, we move to another
node of the clique, and we have to come back about n/2 times before we can
expect to move into the path. But one can argue that on a path of length
n/2, if we start a random walk from one end, we can expect to return to the
starting node n/2 times. Each time, we can expect to spend Ω(n2 ) steps to
get back on the path.
Bounds on the main parameters
We start with some elementary arguments (as we shall see later, eigenvalues
provide more powerful formulas). Recall that if we have just traversed
an edge, then the expected number of steps before it is traversed in this
direction again is 2m. In other words, if we start from node i, and j is
an adjacent node, then the expected time before the edge ji is traversed
in this direction is 2m. Hence the commute time for two adjacent nodes is
bounded by 2m. It follows that the commute time between two nodes at
distance r is at most 2mr < n3 . A similar bound follows for the cover time,
by considering a spanning tree. It is an important consequence of this fact
that these times are polynomially bounded. (It should be remarked that
this does not remain true on directed graphs.)
The following proposition summarizes some known results about cover
and commute times. An O(n3 ) upper bound on the access and cover times
was first obtained by Aleliunas, Karp, Lipton, Lovász and Rackoff [4]. The
upper bound on the access time in (a), which is best possible, is due to
Brightwell and Winkler [13].
It is conjectured that the graph with smallest cover time is the complete
graph (whose cover time is ≈ n log n, as we have seen, and this is of course
independent of the starting distribution). Aldous [1] proved that this is true
up to a constant factor if the starting point is drawn at random, from the
stationary distribution. The asymptotically best possible upper and lower
bounds on the cover time given in (b) are recent results of Feige [31,32].
For the case of regular graphs, a quadratic bound on the cover time was
first obtained by Kahn, Linial, Nisan and Saks (1989). The bound given in
(c) is due to Feige [33].
Theorem 2.1. (a) The access time between any two nodes of a graph on
n nodes is at most
(4/27)n3 − (1/9)n2 + (2/3)n − 1 if n ≡ 0 (mod 3),
Random Walks on Graphs: A Survey
9
(4/27)n3 − (1/9)n2 + (2/3)n − (29/27) if n ≡ 1 (mod 3),
(4/27)n3 − (1/9)n2 + (4/9)n − (13/27) if n ≡ 2 (mod 3).
(b) The cover time from any starting node in a graph with n nodes is at
least (1 − o(1))n log n and at most (4/27 + o(1))n3 .
(c) The cover time of a regular graph on n nodes is at most 2n2 .
It is a trivial consequence of these results that the commute time between any two nodes is also bounded by n3 , and for a regular graph, the
access time is at most 2n2 and the commute time is bounded by 4n2 .
No non-trivial lower bound on the commute time can be found in terms
of the number of nodes: the commute time between the two nodes in the
smaller color class of the complete bipartite graph K2,n is 8. It is true,
however, that κ(u, v) ≥ 2m/d(u) for all u and v (cf. Proposition 2.3 below,
and also Corollary 3.3). In particular, the commute time between two nodes
of a regular graph is always at least n.
The situation is even worse for the access time: this can remain bounded
even for regular graphs. Consider a regular graph (of any degree d ≥ 3) that
has a cutnode u; let G = G1 ∪ G2 , V (G1 ) ∩ V (G2 ) = {u}, and let v be a
node of G1 different from u. Then the access time from v to u is the same
as the access time from v to u in G1 , which is independent of the size of the
rest of the graph.
One class of graphs for which a lower bound of n/2 for any access time
can be proved is the class of graphs with transitive automorphism group;
cf. Corollary 2.6.
Symmetry and access time
The access time from i to j may be different from the access time from j
to i, even in a regular graph. There is in fact no way to bound one of these
numbers by the other. In the example at the end of the last paragraph,
walking from u to v we may, with probability at least 1/d, step to a node of
G2 . Then we have to walk until we return to u; the expected time before this
happens more than |V (G2 )|. So α(u, v) > |V (G2 )|, which can be arbitrarily
large independently of α(v, u).
Still, one expects that time-reversibility should give some sort of symmetry of these quantities. We formulate two facts along these lines. The
first is easy to verify by looking at the walks “backwards”.
10
L. Lovász
Proposition 2.2. If u and v have the same degree, then the probability
that a random walk starting at u visits v before returning to u is equal to
the probability that a random walk starting at v visits u before returning
to v.
(If the degrees of u and v are different, then the ratio of the given
probabilities is π(v)/π(u) = d(v)/d(u).)
The probabilities in Proposition 2.2 are related to the commute time
κ(u, v) in an interesting way:
Proposition 2.3. The probability that a random walk starting at u visits
v before returning to u is 1/(κ(u, v)π(u)).
Proof. Let q denote the probability in question. Let τ be the first time
when a random walk starting at u returns to u and σ, the first time when
it returns to u after visiting v. We know that E(τ ) = 2m/d(u) and by
definition, E(σ) = κ(u, v). Clearly τ ≤ σ and the probability of τ = σ is
exactly q. Moreover, if τ < σ then after the first τ steps, we have to walk
from u until we reach v and then return to u. Hence E(σ −τ ) = (1−q)E(σ),
and hence
E(τ )
2m
q=
=
.
E(σ)
d(u)κ(u, v)
A deeper symmetry property of access times was discovered by Coppersmith, Tetali and Winkler [19]. This can also be verified by elementary
means considering walks visiting three nodes u, v and w, and then reversing
them, but the details are not quite simple.
Theorem 2.4. For any three nodes u, v and w,
H(u, v) + H(v, w) + H(w, u) = H(u, w) + H(w, v) + H(v, u).
An important consequence of this symmetry property is the following.
Corollary 2.5. The nodes of any graph can be ordered so that if u precedes
v then H(u, v) ≤ H(v, u). Such an ordering can be obtained by fixing any
node t, and order the nodes according to the value of H(u, t) − H(t, u).
Proof. Assume that u precedes v in the ordering described. Then H(u, t)−
H(t, u) ≤ H(v, t) − H(t, v) and hence H(u, t) + H(t, v) ≤ H(v, t) + H(t, u).
By Theorem 2.4, this is equivalent to saying that H(u, v) ≤ H(v, u).
Random Walks on Graphs: A Survey
11
This ordering is not unique, because of the ties. But if we partition
the nodes by putting u and v in the same class if H(u, v) = H(v, u) (this
is an equivalence relation by Proposition 2.4), then there is a well-defined
ordering of the equivalence classes, independent of the reference node t. The
nodes in the lowest class are “difficult to reach but easy to get out of”, the
nodes in the highest class are “easy to reach but difficult to get out of”. It
is worth formulating a consequence of this construction:
Corollary 2.6. If a graph has a vertex-transitive automorphism group then
H(i, j) = H(j, i) for all nodes i and j.
Access time and cover time
The access times and commute times of a random walk have many nice
properties and are relatively easy to handle. The cover time is more elusive.
But there is a very tight connection between access times and cover times,
discovered by Matthews [56]. (See also Matthews [57]; this issue of the J.
Theor. Probability contains a number of other results on the cover time.)
Theorem 2.7. The cover time from any node of a graph with n nodes is
at most (1 + (1/2) + . . . + (1/n)) times the maximum access time between
any two nodes, and at least (1 + (1/2) + . . . + (1/n)) times the minimum
access time between two nodes.
Let us sketch a simple proof for the somewhat weaker upper bound of
2 log2 n times the maximum access time.
Lemma 2.8. Let b be the expected number of steps before a random walk
visits more than half of the nodes, and let h be the maximum access time
between any two nodes. Then b ≤ 2h.
¿From this lemma, the theorem is easy. The lemma says that in 2h
steps we have seen more than half of all nodes; by a similar argument, in
another 2h steps we have seen more than half of the rest etc.
Proof. Assume, for simplicity, that n = 2k + 1 is odd. Let αv be the time
when node v is first visited. Then the time β when we reach more than half
of the nodes is the (k + 1)-st largest of the αv . Hence
X
αv ≥ (k + 1)β,
v
and so
b = E(β) ≤
n
1 X
E(αv ) ≤
h < 2h.
k+1 v
k+1
12
L. Lovász
Monotonicity
Let G′ be obtained from the graph G by adding a new edge ab. Since this
new graph is denser, one expects that a random walk on it turns back less
frequently, and therefore the access times, commute times, and cover times
decrease. As it turns out, this does not hold in general.
First, it is easy to see that some access times may increase dramatically
if an edge is added. Let G be a path on n nodes, with endpoints a and b.
Let s = a and let t be the unique neighbor of s. Then the access time from s
to t is 1. On the other hand, if we add the edge (a, b) then with probability
1/2, we have to make more than one step, so the access time from s to t
will be larger than one; in fact, it jumps up to n − 1, as we have seen.
One monotonicity property of access time that does hold is that if an
edge incident with t is added, then the access time from s to t is not larger
in G′ than in G.
The commute time, which is generally the best behaved, is not monotone either. For example, the commute time between two opposite nodes of
a 4-cycle is 8; if we add the diagonal connecting the other two nodes, the
commute time increases to 10. But the following “almost monotonicity”
property is true (we’ll return to its proof in section 4).
Theorem 2.9. If G′ arises from a graph G by adding a new edge, and G
has m edges, then the commute time between any two nodes in G′ is at
most 1 + 1/m times the commute time in G. In other words, the quantity
κ(s, t)/m does not decrease.
We discuss briefly another relation that one intuitively expects to hold:
that access time increases with distance. While such intuition is often
misleading, the following results show a case when this is true (Keilson
[42]).
Theorem 2.10. Let G be a graph and t ∈ V (G).
(a) If we choose s uniformly from the set of neighbors of t, then the expectation of H(s, t) is exactly (2m/d(t)) − 1.
(b) If we choose s from the stationary distribution over V , then the exµ
¶
d(t) 2
2m
1−
. So if we condition on
pectation of H(s, t) is at least
d(t)
2m
s 6= t, the expectation of H(s, t) is at least (2m/d(t)) − 1.
(c) If we choose t from the stationary distribution over V , then the expectation of H(s, t) is at least n − 2 + 1/n.
Random Walks on Graphs: A Survey
13
(a) is just a restatement of the formula for the return time. The proof
of (b) and (c) uses eigenvalue techniques. It is easy to derive either from
(b) or (c) that maxs,t H(s, t) ≥ n − 1. We remark that the expectation in
(c) is independent of s (see formula (3.3)).
Applications of the cover time and commute time bounds
Perhaps the first application of random walk techniques in computer science
was the following (Aleliunas, Karp, Lipton, Lovász and Rackoff [4]). Let
G = (V, E) be a connected d-regular graph, v0 ∈ V (G), and assume that
at each node, the ends of the edges incident with the node are labelled
1, 2, . . . , d. A traverse sequence (for this graph, starting point, and labelling)
is a sequence (h1 , h2 , . . . , ht ) ⊆ {1, . . . , d}t such that if we start a walk at v0
and at the ith step, we leave the current node through the edge labelled hi ,
then we visit every node. A universal traverse sequence (for parameters n
and d) is a sequence which is a traverse sequence for every d-regular graph
on n nodes, every labelling of it, and every starting point.
It is quite surprising that such sequences exist, and in fact need not be
too long:
Theorem 2.11. For every d ≥ 2 and n ≥ 3, there exists a universal traverse
sequence of length O(d2 n3 log n).
A consequence of this fact is that the reachability problem on undirected
graphs is solvable in non-uniform logspace. We do not discuss the details.
Proof. The “construction” is easy: we consider a random sequence. More
exactly, let t = 8dn3 log n, and let H = (h1 , . . . , ht ) be randomly chosen from
{1, . . . , d}t . For a fixed G, starting point, and labelling, the walk defined
by H is just a random walk; so the probability p that H is not a traverse
sequence is the same as the probability that a random walk of length t does
not visit all nodes.
By Theorem 2.1, the expected time needed to visit all nodes is at most
2n2 . Hence (by Markov’s Inequality) the probability that after 4n2 steps
we have not seen all nodes is less than 1/2. Since we may consider the next
4n2 steps as another random walk etc., the probability that we have not
2
seen all nodes after t steps is less than 2−t/(4n ) = n−2nd .
Now the total number of d-regular graphs G on n nodes, with the ends
of the edges labelled, is less than ndn (less than nd choices at each node),
and so the probability that H is not a traverse sequence for one of these
graphs, with some starting point, is less than nnnd n−2nd < 1. So at least
one sequence of length t is a universal traverse sequence.
14
L. Lovász
The results of Coppersmith, Tetali and Winkler [19] discussed above
served to solve the following problem: let us start two random walks on a
graph simultaneously; how long does it take before they collide? There are
variations depending on whether the two random walks step simultaneously,
alternatingly, or in one of several other possible ways. Here we only consider
the worst case, in which a “schedule daemon” determines which random
walk moves at any given time, whose aim is to prevent collision as long as
possible.
The motivation of this problem is a self-stabilizing token-management
scheme for a distributed computing network. The “token” is the authorization for the processor carrying it to perform some task, and at any time,
only one processor is supposed to carry it. Assume that by some disturbance, two processors carry the token. They pass it around randomly, until
the two tokens collide; from then on, the system is back to normal. How
long does this take?
Let M (u, v) denote the expected number of steps before two random
walks, starting from nodes u and v, collide. It is clear that M (u, v) ≥
H(u, v) (v may never wake up to move). Coppersmith, Tetali and Winkler
[19] prove the nice inequality
M (u, v) ≤ H(u, v) + H(v, w) − H(w, u)
for some vertex w. Thus it follows that the collision time is O(n3 ).
3. The eigenvalue connection
Recall that the probability ptij of the event that starting at i, the random
walk will be at node j after t steps, is an entry of the matrix M t . This
suggests that the powerful methods of the spectral theory of matrices can
be used.
The matrix M has largest eigenvalue 1, with corresponding left eigenvalue π and corresponding right eigenvalue 1, the all-1 vector on V . In fact,
M T π = π expresses the fact that π is the stationary distribution, while
M 1 = 1 says that exactly one step is made from each node.
Unfortunately, M is not symmetric unless G is regular; but it is easy
to bring it to a symmetric form. In fact, we kow that M = DA, where
A = AG is the adjacency matrix of G and D is the diagonal matrix in which
15
Random Walks on Graphs: A Survey
the i-th diagonal entry is 1/d(i). Consider the matrix N = D1/2 AD1/2 =
D−1/2 M D1/2 . This is symmetric, and hence can be written in a spectral
form:
n
X
N=
λk vk vkT ,
k=1
where λ1 ≥ λ2 ≥ . . . ≥ λn are the eigenvalues of N and v1 , . . . , vn are
the corresponding
eigenvectors of unit length. Simple substitution shows
p
that wi = d(i) defines an eigenvector of N with eigenvalue 1. Since this
eigenvector is positive, it follows from the Frobenius-Perron Theorem that
λ1 = 1 √
> λ2 ≥ . . . ≥ λn p
≥ −1 and that
p (possibly after flipping signs)
v1 = (1/ 2m)w, i.e., v1i = d(i)/2m = π(i). It also follows by standard
arguments that if G is non-bipartite then λn > −1.
Now we have
M t = D1/2 N t D−1/2 =
n
X
λtk D1/2 vk vkT D−1/2 = Q +
k=1
n
X
λtk D1/2 vk vkT D−1/2
k=2
where Qij = π(j). In other words,
ptij = π(j) +
n
X
λtk vki vkj
k=2
s
d(j)
.
d(i)
(3.1)
If G is not bipartite then |λk | < 1 for k = 2, . . . , n, and hence
ptij → π(j)
(t → ∞)
as claimed above. We shall return to the rate of this convergence later.
Spectra and access times
We start a more in-depth study of connections between random walks and
spectra by deriving a spectral formula for access times. Let H ∈ RV ×V
denote the matrix in which Hij = H(i, j), the access time from i to j. Let
Γ(i) be the set of neighbors of node i. The key equation is that if i 6= j then
1 X
H(v, j)
H(i, j) = 1 +
d(i)
v∈Γ(i)
(since the first step takes us to a neighbor v of i, and then we have to reach
j from there). Expressing this equation in matrix notation, we get that
F = J + M H − H is a diagonal matrix. Moreover,
F T π = Jπ + H T (M − I)T π = Jπ = 1,
16
L. Lovász
whence
(F )ii =
1
2m
=
.
π(i)
d(i)
Thus F = 2mD, i.e.,
(I − M )H = J − 2mD.
(3.2)
We want to solve this “matrix equation” for H. Of course, this is not
possible since I − M is singular; in fact, with every X satisfying (3.2) (in
place of H), every matrix X + 1aT also satisfies it for any vector a. But
these are all, as elementary linear algebra shows, and so a can be determined
using the relations
H(i, i) = 0
(i ∈ V ).
So if we find any solution of (3.2), we can obtain H by subtracting the
diagonal entry from each column.
Let M ∗ denote the matrix 1π T , i.e., Mij∗ = π(j) (note that M ∗ is
the limit of M t as t → ∞). Substitution shows that the matrix X =
(I − M + M ∗ )−1 (J − 2mD) satisfies (3.2). Diagonalizing M as above, we
get the following formula:
Theorem 3.1.
H(s, t) = 2m
n
X
k=2
1
1 − λk
Ã
2
vkt
vks vkt
−p
d(t)
d(s)d(t)
!
.
As an immediate corollary we obtain a similar formula for the commute
time:
Corollary 3.2.
κ(s, t) = 2m
n
X
k=2
1
1 − λk
Ã
v
v
p kt − p ks
d(t)
d(s)
Using that
1
1
1
≤
≤
2
1 − λk
1 − λ2
along with the orthogonality of the matrix (vks ), we get
!2
.
17
Random Walks on Graphs: A Survey
Corollary 3.3.
µ
¶
¶
µ
1
1
1
1
2m
m
+
+
≤ κ(s, t) ≤
.
d(s) d(t)
1 − λ2 d(s) d(t)
If the graph is regular, the lower bound is n. If we have a expander graph,
which can be characterized as a regular graph for which 1/(1 − λ2 ) = O(1),
then it follows that the commute time between any pair of nodes is Θ(n).
In these formulas, the appearence of 1 − λk in the denominators suggest
that it will be necessary to find good bounds on the spectral gap: the
difference 1 − λ2 = λ1 − λ2 . This is an important parameter for many
other studies of graphs, and we shall return to its study in the next section.
To warm up to the many applications of Theorem 3.1, the reader is
encouraged to give a proof of the week symmetry property of access times
expressed in Theorem 2.4, and of the expression for access times in terms of
commute times (2.1). Another easy corollary is obtained if we average the
access time over all t. We have
s
Ã
!
n
X
XX
d(t)
1
2
vkt − vkt vks
π(t)H(s, t) =
1 − λk
d(s)
t
t k=2
s
Ã
!
n
X
X
X p
1
1
2
=
vkt
− vks
vkt d(t) .
1 − λk
d(s) t
t
k=2
Using that vk is of unit length and it is orthogonal to v1 for k ≥ 2, we get
the nice formula
n
X
X
1
.
(3.3)
π(t)H(s, t) =
1 − λk
t
k=2
Note that this value is independent of the starting node s.
As another application, we find the access time between two antipodal
nodes of the k-cube Qk . Let 0 = (0, . . . , 0) and 1 = (1, . . . , 1) represent two
antipodal nodes of the k-cube. As is well known, we get an eigenvector vb
of M (or A) for every 0-1 vector b ∈ {0, 1}k , defined by (vb )x = (−1)b·x .
The corresponding eigenvalue of M is 1 − (2/k)b · 1. Normalizing vb and
substituting in Theorem 3.1, we get that
k µ ¶
X
k 1
H(0, 1) = k
(1 − (−1)j ).
j 2j
j=1
18
L. Lovász
To find the asymptotic value of this expression, we substitute
Pk−1 ¡ p ¢
p=0 j−1 , and get
¡k ¢
j
=
µ
¶
k X
k−1
X
p
1
(1 − (−1)j )
H(0, 1) = k
2j j − 1
j=1 p=0
=k
¶
k µ
X
p+1
1
(1 − (−1)j )
2(p + 1)
j
k−1
X
p=0
=k
j=1
k−1
X
k−1
X 1 k
2p
= 2k−1
∼ 2k .
p+1
2j k − j
p=0
j=0
(It is easy to see that the exact value is always between 2k and 2k+1 .)
As a further application, let us prove that “more distant targets are
more difficult to reach” (Theorem 2.10.b). The argument is similar to the
proof of (3.3). We have
s
Ã
!
n
X
XX
d(s)
d(s)
1
2
vkt
− vkt vks
.
π(s)H(s, t) =
1
−
λ
d(t)
d(t)
k
s
s
k=2
Using again that vk is orthogonal to v1 for k ≥ 2, we have
n
X
π(s)H(s, t) =
s
2m X 1
v2 .
d(t)
1 − λk kt
k=2
By the inequality between arithmetic and harmonic means (considering the
2 as weights), we have
vkt
Pn
Now here
1
2
vkt
k=2
Pn 1−λk2
k=2 vkt
n
X
2
vkt
k=2
and
=
n
X
k=1
≥ Pn
Pn
2
k=2 vkt
k=2 (1
2
− λk )vkt
.
2
vkt
− π(t) = 1 − π(t)
n
n
n
X
X
X
2
2
2
λk vkt
= 1 − (N )t,t = 1.
(1 − λk )vkt = 1 −
(1 − λk )vkt =
k=2
k=1
k=1
19
Random Walks on Graphs: A Survey
Thus
X
s
1
(1 − π(t))2 ,
π(t)
π(s)H(s, t) ≥
which proves the assertion.
Perhaps the most important applications of eigenvalue techniques concern the mixing rate, which we’ll discuss in a separate section.
Spectra and generating functions
One may obtain spectral formulas carrying even more information by introducing the probability generating function
F (x) =
∞
X
t=0
xt M t = (I − xM )−1 .
(the (i, j) entry Fij (x) of this matrix is the generating function for the
probabilities ptij ).
Using this function, we can express other probabilities via standard
t denote the
techniques of generating functions. As an example, let qij
probability that the random walk starting at i hits node j for the first
time in the t-th step. It is clear that
ptij =
t
X
s t−s
qij
pjj ,
s=0
We can translate this relation in terms of generating functions as follows.
Let
∞
X
t t
qij
x,
Gij (x) =
t=0
then
Fij (x) = Gij (x)Fjj (x).
So the matrix G(x) = (Gij (x)) arises from F (x) by scaling each column so
that the diagonal entry becomes 1.
We may use the spectral decomposition of M to get more explicit
formulas. We have
s
s
∞ n
n
d(j) X X
d(j) X
1
t
Fij (x) =
.
(xλk ) vki vkj =
vki vkj
d(i)
d(i)
1 − xλk
t=0 k=1
k=1
20
L. Lovász
Hence we also get the generating function
s
, n
n
X
1
1
d(j) X
2
vki vkj
.
vkj
Gij (x) =
d(i)
1 − xλk
1 − λk x
k=1
k=1
¿From this another proof of Theorem 3.1 follows easily, since
H(s, t) = G′st (1).
By calculating higher derivatives, we can derive similar (though increasingly
complicated) formulas for the higher moments of the time a node t is first
visited.
4. The electrical connection
Let G = (V, E) be a connected graph and S ⊆ V . A function φ : V → R is
called a “harmonic function with set of poles S” if
1 X
φ(u) = φ(v)
d(v)
u∈Γ(v)
holds for every v ∈
/ S (the set S is also called the boundary of the harmonic
function). Not surprisingly, harmonic functions play an important role in
the study of random walks: after all, the averaging in the definition can
be interpreted as expectation after one move. They also come up in the
theory of electrical networks, and also in statics. This provides a connection
between these fields, which can be exploited. In particular, various methods
and results from the theory of electricity and statics, often motivated by
physics, can be applied to provide results about random walks.
We start with describing three constructions of harmonic functions, one
in each field mentioned.
(a) Let φ(v) denote the probability that a random walk starting at node v
hits s before it hits t. Clearly, φ is a harmonic function with poles s
and t. We have φ(s) = 1 and φ(t) = 0.
More generally, if we have a set S ⊆ V and a function φ0 : S → R, then
we define φ(v) for v ∈ V \ S as the expectation of φ0 (s), where s is the
(random) node where a random walk starting at v first hits S. Then
Random Walks on Graphs: A Survey
21
φ(v) is a harmonic function with pole set S. Moreover, φ(s) = φ0 (s)
for all s ∈ S.
(b) Consider the graph G as an electrical network, where each edge represents a unit resistance. Assume that an electric current is flowing
through G, entering at s and leaving at t. Let φ(v) be the voltage of
node v. Then φ is a harmonic function with poles s and t.
(c) Consider the edges of the graph G as ideal springs with unit Hooke
constant (i.e., it takes h units of force to stretch them to length h).
Let us nail down nodes s and t to points 1 and 0 on the real line, and
let the graph find its equilibrium. The energy is a positive definite
quadratic form of the positions of the nodes, and so there is a unique
minimizing position, which is the equilibrium. Clearly all nodes will lie
on the segment between 0 and 1, and the positions of the nodes define
a harmonic function with poles s and t.
More generally, if we have a set S ⊆ V and we fix the positions of
the nodes in S (in any dimension), and let the remaining nodes find
their equilibrium, then any coordinate of the nodes defines a harmonic
function with pole set S.
Let us sum up some trivial properties of harmonic functions. Clearly,
φ(v) lies between the minimum and maximum of φ over S. Moreover, given
S ⊆ V and φ0 : S → R, there is a unique harmonic function on G with
pole set S extending φ0 . (The existence follows by either construction (a)
or (c); the uniqueness follows by considering the maximum of the difference
of two such functions.)
In particular it follows that every harmonic function with at most one
pole is constant. We denote by φst the (unique) harmonic function with
poles s and t such that φst (s) = 1 and φst (t) = 0.
Another consequence of the uniqueness property is that the harmonic
functions constructed in (a) and (c), and (for the case |S| = 2) in (b) are the
same. As an application of this idea, we show the following useful characterizations of commute times (see Nash-Williams [60], Chandra, Raghavan,
Ruzzo, Smolensky and Tiwari [16]).
Theorem 4.1. (i) Consider the graph G as an electrical network as in (b)
and let Rst denote the resistance between nodes s and t. Then the commute
time between nodes s and t is exactly 2mRst .
(ii) Consider the graph G as a spring structure in equilibrium, as in
example (c), with two nodes s and t nailed down at 1 and 0. Then the force
22
pulling the nails is
L. Lovász
2m
1
=
.
Rst
κ(s, t)
The energy of the system is
m
1
=
.
2Rst
κ(s, t)
Note that equation (2.1) can be used to express access times in terms
of resistances or spring forces (Tetali [63]).
Proof. By construction (b), φst (v) is the voltage of v if we put a current
through G from s to t, where the voltage of sP
is 0 and the voltage of t
is 1. The total current through the network is u∈Γ(t) φst (u), and so the
resistence is
−1
X
φst (u) .
Rst =
u∈Γ(s)
On the other hand, (a) says that φst (u) is the probability
that a random
1 P
walk starting at u visits s before t, and hence d(t) u∈Γ(t) φst (u) is the
probability that a random walk starting at t hits s before returning to t.
By Proposition 2.3, this probability is 2m/d(t)κ(s, t). This proves assertion
(i). The proof of (ii) is similar.
Using the “topological formulas” from the theory of electrical networks
for the resistance, we get the following characterization of commute time:
Corollary 4.2. Let G be a graph and s, t ∈ V . Let G′ denote the graph
obtained from G by identifying s and t, and let T (G) denote the number of
spanning trees of G. Then
κ(s, t) = 2m
T (G′ )
.
T (G)
The following fact is called Raleigh’s Principle in the theory of electrical
networks. We derive it as a consequence of Theorem 4.1.
Corollary 4.3. Adding any edge to a graph G does not increase any
resistance Rst . Consequently, no commute time κ(s, t) is increased by more
than a factor of (m + 1)/m.
In fact, it suffices to prove that deleting an edge from a graph G cannot
increase the energy of the equilibrium configuration in the spring structure
Random Walks on Graphs: A Survey
23
(c). Clearly, deleting an edge while keeping the positions of the nodes fixed
cannot increase the energy. If we let the new graph find its equilibrium then
the energy can only further decrease.
Combining Corollaries 4.2 and 4.3, a little algebraic manipulation gives
the following inequality for the numbers of spanning trees in a graph G and
in its subgraphs G − e, G − f , and G − e − f , where e and f are two edges
of G:
T (G − e)T (G − f ) ≥ T (G)T (G − e − f ).
(4.1)
5. Mixing rate
In several recent applications of random walks, the most important parameter is the mixing rate. Using eigenvalues, it is an easy task to determine
the mixing rate in polynomial time (see below), but this result does not tell
the whole story, since, as we shall see, the underlying graph in the cases
of interest is exponentially large, and the computation of the eigenvalues
by the tools of linear algebra is hopeless. Therefore, combinatorial techniques that lead to approximations only but are more manageable are often
preferable. Two main techniques that have been used are coupling and conductance. In this section we discuss these techniques; in the next, we give
several applications in algorithm design.
Mixing rate and coupling
We shall illustrate the methods for bounding the mixing rate on a special
class of graphs. (For reasons of comparison, we will also apply the other
methods to the same graph.) These graphs are the cartesian sum Cnk of
k circuits of length n, where n is odd. The node set of this graph is
{0, . . . , n − 1}k , and two nodes (x1 , . . . , xk ) and (y1 , . . . yk ) are adjacent
iff there exists an i, 1 ≤ i ≤ k, such that xj = yj for j 6= i and xi ≡
yi ± 1 (mod n).
Let us start a random walk (v0 , v1 , . . .) on Cnk from an arbitrary initial
distribution P0 . To estimate how long we have to walk to get close to
the stationary distribution (which is uniform in this case), let us start
another random walk (w0 , w1 , . . .), in which w0 is drawn from the uniform
distribution. Of course, wt is then uniformly distributed for all t.
The two walks are not independent; we “couple” them as follows. The
vertices of Cnk are vectors of length k, and a step in the random walk consists
24
L. Lovász
of changing a randomly chosen coordinate by one. We first generate the
step in the first walk, by selecting a random coordinate j, 1 ≤ j ≤ k, and
a random ε ∈ {−1, 1}. The point vt+1 is obtained by adding ε to the jth coordinate of vt . Now the trick is that if vt and wt agree in the j-th
coordinate, we generate wt+1 by adding ε to the j-th coordinate of wt ; else,
we subtract ε from the j-th coordinate of wt . (All operations are modulo
n.)
The important fact is that viewing (w0 , w1 , . . .) in itself, it is an entirely
legitimate random walk. On the other hand, the “coupling” rule above
implies that once a coordinate of vt becomes equal to the corresponding
coordinate of wt , it remains so forever. Sooner or later all coordinates
become equal, then vt will have the same distribution as wt , i.e., uniform.
To make this argument precise, let us look at the steps when the j-th
coordinate is selected. The expected number of such steps before the two
walks will have equal j-th coordinate is the average access time between two
nodes of the circuit on length n, which is (n2 −1)/6. So the expected number
of steps before all coordinates become equal is k(n2 − 1)/6. By Markov’s
inequality, the probability that after kn2 steps vt and wt are still different
is less than 1/6, and so the probability that after ckn2 steps these points
are still different is less than 6−c . Hence for any T that is large enough,
|P (vT ∈ S) −
|S|
2
| = |P (vT ∈ S) − P (wT ∈ S)| ≤ P (wT 6= vT ) < 6−T /(kn ) .
nk
2)
We obtain that the mixing rate is at most 6−1/(kn
<1−
1
.
kn2
This method is elegant but it seems that for most applications of interest, there is no simple way to find a coupling rule, and so it applies only in
lucky circumstances.
Mixing rate and the eigenvalue gap
An algebraic formula for the mixing rate is easily obtained.
min{|λ2 |, |λn |}, then from (3.1) it is easy to derive:
Theorem 5.1. For a random walk starting at node i,
s
d(j) t
|Pt (j) − π(j)| ≤
λ.
d(i)
More generally,
|Pt (S) − π(S)| ≤
s
π(S) t
λ.
π(i)
Let λ =
25
Random Walks on Graphs: A Survey
So the mixing rate is at most λ; it is not difficult to argue that equality
must hold here. Thus we obtain:
Corollary 5.2. The mixing rate of a random walk on a non-bipartite graph
G is λ = max{|λ2 |, |λn |}.
In most applications, we don’t have to worry about λn ; for example,
we can add d(i) loops at each point i, which only slows down the walk by
a factor of 2, but results in a graph with positive semidefinite adjacency
matrix. The crucial parameter is λ2 , or rather, the “spectral gap” 1 − λ2 .
Note that log(1/λ) ≈ (1 − λ)−1 .
Theorem 5.1 concerns the convergence to the stationary distribution
in terms of the total variation distance, which seems to be the most important for applications. Other measures have other, sometimes technical,
adventages. For example, using the χ2 measure has the adventage that the
distance is improving after each step (Fill [34]):
X (Pt+1 (j) − π(j))2
j
π(j)
≤λ
X (Pt (j) − π(j))2
j
π(j)
.
As an application of Theorem 5.1, let us determine the mixing rate of
a random walk on an n-dimensional cube. This graph is bipartite, so we
add loops; let’s add n loops at each node. The eigenvalues of the resulting
graph are 0, 2, 4, . . . , 2n, and so the eigenvalues of the transition matrix are
0, 1/n, 2/n, . . . , (n − 1)/n, 1. Hence the mixing rate is (n − 1)/n.
Next, let us do the graph Cnk , where n is odd. The eigenvalues of Cn
are 2 cos(2rπ/n), 0 ≤ r < n. Hence the eigenvalues of the adjacency matrix
Cnk are all numbers
2 cos(2r1 π/n) + 2 cos(2r2 π/n) + . . . + 2 cos(2rk π/n)
(see e.g. Lovász [48], exercise 11.7). In particular, the largest eigenvalue is
(of course) 2k, the second largest is 2(k − 1) + 2 cos(2π/n), and the smallest
is 2k cos((n − 1)π/n). From this it follows that the mixing rate is
1
1−
k
µ
2π
1 − cos
n
¶
≈1−
2π 2
.
kn2
26
L. Lovász
The eigenvalue gap and conductance
Let G be a graph and S ⊂ V , S 6= ∅. Let ∇(S) denote the set of edges
connecting S to V \ S. We define the conductance of the set S ⊂ V , S 6= ∅
by
|∇(S)|
Φ(S) =
2mπ(S)π(V \ S)
and the conductance of the graph by
Φ = min Φ(S),
S
where the minimum is taken over all non-empty proper subsets S ⊂ V . If
the graph is d-regular, then the conductance of S is
Φ(S) =
n|∇(S)|
.
d|S| · |V \ S|
To digest this quantity a little, note that |∇(S)|/2m is the frequency
with which a stationary random walk switches from S to V \ S; while
π(S)π(V \S) is the frequency with which a sequence of independent random
elements of V , drawn from the stationary distribution π, switches from S
to V \ S. So Φ can be viewed as a certain measure of how independent
consecutive nodes of the random walk are.
Sinclair and Jerrum [62] established a connection between the spectral
gap and the conductance of the graph. A similar result for the related, but
somewhat different parameter called expansion rate was proved by Alon
[5] and, independently, by Dodziuk and Kendall [24]; cf. also Diaconis and
Stroock [21]. All these results may be considered as discrete versions of
Cheeger’s inequality in differential geometry.
Theorem 5.3.
Φ2
≤ 1 − λ2 ≤ Φ.
8
We’ll sketch the proof of this fundamental inequality; but first, we state
(without proof) a simple lemma that is very useful in the study of the
spectral gap.
Lemma 5.4.
1 − λ2 = min
X
(xi − xj )2 :
ij∈E(G)
X
i
π(i)xi = 0,
X
i
1
π(i)x2i =
2m
27
Random Walks on Graphs: A Survey
(each edge ij is considered only once in the sum).
Proof. Proof of Theorem 5.3 To warm up, let us prove the upper bound
first. By Lemma 5.4, it suffices to exhibit a vector x ∈ RV such that
X
X
π(i)xi = 0,
i
π(i)x2i = 1/(2m),
(5.1)
i
and
X
ij∈E(G)
(xi − xj )2 = Φ.
(5.2)
Let S be a set with minimum conductance, and consider a vector of the
type
½
a, if i ∈ S,
xi =
b, if i ∈ V \ S.
Such a vector satisfies (5.1) if
a=
s
π(V \ S)
,
2mπ(S)
s
b=−
π(S)
,
2mπ(V \ S)
and then straightforward substitution shows that (5.2) is also satisfied.
To prove the lower bound, we again invoke Lemma 5.4: we prove that
for every vector x ∈ RV satisfying (5.1), we have
X
(xi − xj )2 ≥
ij∈E(G)
Φ2
.
8
(5.3)
Conductance enters the picture through the following inequality, which
is, in a sense, the “ℓ1 -version” of (5.3).
Lemma 5.5. Let G be a graph with conductance Φ. Let P
y ∈ RV and
assume that π({i : yi > 0}) ≤ 1/2, π({i : yi < 0}) ≤ 1/2 and i π(i)|yi | =
1. Then
X
|yi − yj | ≥ mΦ.
(i,j)∈E
Proof. Proof of the Lemma Label the points by 1, . . . , n so that
y1 ≤ y2 ≤ yt < 0 = yt+1 = . . . = ys < ys+1 ≤ . . . ≤ yn .
28
L. Lovász
Set Si = {1, . . . , i}. Substituting yj − yi = (yi+1 − yi ) + · · · + (yj − yj−1 ),
we get
X
|yi − yj | =
(i,j)∈E
n−1
X
i=1
|∇(Si )|(yi+1 − yi ) ≥ 2mΦ
n−1
X
i=1
(yi+1 − yi )π(Si )π(V \ Si ).
Using that π(Si ) ≤ 1/2 for i ≤ t, π(Si ) ≥ 1/2 for i ≥ s + 1, and that
yi+1 − yi = 0 for t < i < s, we obtain
X
(i,j)∈E
|yi − yj | ≥ mΦ
= mΦ
t
X
i=1
X
i
(yi+1 − yi )π(Si ) + mΦ
n−1
X
i=t+1
(yi+1 − yi )π(V \ Si )
π(i)|yi | = mΦ.
Now we return to the proof of the theorem. Let x be any vector
satisfying (5.1). We may assume that x1 ≥ x2 ≥ . . . ≥ xn . Let k (1 ≤ k ≤ n)
be the index for which π({1, . . . , k − 1}) ≤ 1/2 and π({k + 1, . . . , n}) < 1/2.
Setting zi = max{0, xi − xk } and choosing the sign of x appropriately, we
may assume that
X
i
π(i)zi2 ≥
X
1X
1
1X
π(i)xi + x2k
π(i)(xi − xk )2 =
π(i)x2i − xk
2
2
2
i
i
i
1
1
1
+ x2k ≥
.
=
2m 2
2m
Now Lemma 5.5 can be applied to the numbers yi = zi2 /
obtain that
X
X
π(i)zi2 .
|zi2 − zj2 | ≥ mΦ
2
i π(i)zi ,
P
and we
i
(i,j)∈E
On the other hand, using the Cauchy-Schwartz inequality,
X
(i,j)∈E
|zi2 − zj2 | ≤
X
(i,j)∈E
1/2
(zi − zj )2
X
(i,j)∈E
1/2
(zi + zj )2
Here the second factor can be estimated as follows:
X
X
X
(zi + zj )2 ≤ 2
(zi2 + zj2 ) = 4m
π(i)zi2 .
(i,j)∈E
(i,j)∈E
i
.
29
Random Walks on Graphs: A Survey
Combining these inequalities, we obtain
X
(i,j)∈E
(zi − zj )2 ≥
X
(i,j)∈E
≥ Φ2 m2
2
|zi2 − zj2 |
Ã
X
π(i)zi2
i
. X
!2
(zi + zj )2
(i,j)∈E
.
Φ2
Φ2 m X
π(i)zi2 ≥
.
=
4
8
4m
X
π(i)zi2
i
i
Since trivially
X
(i,j)∈E
(xi − xj )2 ≥
X
(i,j)∈E
(zi − zj )2 ,
the theorem follows.
Corollary 5.6. For any starting node i, any node j and any t ≥ 0,
s
µ
¶
2 t
¯ t
¯
¯P (j) − π(j)¯ ≤ d(j) 1 − Φ
.
d(i)
8
In another direction, Chung and Yau [17] considered a refined notion of
conductance, replacing π(S)π(V \ S) in the denominator by some power
of it, and showed how this relates to higher eigenvalues. Diaconis and
Saloff-Coste [23] used similar inequalities to get improved bounds on the
mixing time, in particular on the early part when the distribution is highly
concentrated. Theorem 5.3 is a discrete analogue of Cheeger’s inequality
from differential geometry, and these inequalities are discrete analogues of
the Harnack, Sobolev and Nash inequalities known from the theory of the
heat equation, and in fact, these results represent first steps in the exciting
area of studying “difference equations” on graphs as discrete analogues of
differential equations.
Conductance and multicommodity flows
Conductance itself is not an easy parameter to handle; it is NP-hard to determine it even for an explicitly given graph. But there are some methods
to obtain good estimates. The most important such method is the construction of multicommodity flows. Let us illustrate this by a result of Babai and
Szegedy [11].
30
L. Lovász
Theorem 5.7. Let G be a connected graph with a node-transitive automorphism group, with diameter D. Then the conductance of G is at least
1/(dD). If the graph is edge-transitive, its conductance is at least 1/D.
Proof. For each pair i, j of points, select a shortest path Pij connecting
them. Let P denote the family of these paths and all their images under
automorphisms
of G. The total number of paths in P (conting multiplicities)
µ ¶
n
is
g, where g is the number of automorphisms of G. Moreover, P
2
contains exactly g paths connecting any given pair of points.
We claim that every edge occurs in at most Dg(n − 1) paths of P. In
fact, if an edge e occurs in p paths then so does every image of e under
the automorphisms, and there are at least n/2 distinct images by nodetransitivity. This gives
µ ¶pn/2 edges, but the total number of edges of paths
n
in P is at most Dg
, which proves the claim.
2
Now let S ⊆ V (G), |S| = s ≤ |V (G)|/2. The number of paths in P
connecting S to V (G) \ S is exactly gs(n − s). On the other hand, this
number is at most |∇(S)| · Dg(n − 1), and hence
|∇(S)| ≥
s n−s
gs(n − s)
=
·
.
Dg(n − 1)
D n−1
Hence the conductance of S is
n|∇(S)|
1
n
1
≥
>
.
ds(n − s)
n − 1 dD
dD
This proves the first assertion. The second follows by a similar argument.
Let us use Theorem 5.7 to estimate the mixing rate of Cnk (where n
is odd). This graph has an edge-transitive automorphism group, and its
diameter is k(n − 1)/2. Hence its conductance is more than 2/(kn), and so
its mixing rate is at most
1
1− 2 2
2k n
We see that the bound is worse than the coupling and eigenvalue bounds;
in fact, depending on the relative value of n and k, the mixing rate may be
close to either the upper or the lower bound in Theorem 5.3.
If we look in the proof of Theorem 5.7 at all paths connecting a given
pair {u, v} of nodes, and take each such path with weight 1/n2 g, we get
Random Walks on Graphs: A Survey
31
2
a flowµfrom
¶ u to v with value 1/n . The little argument given shows that
n
these
flows load each edge with at most D(n − 1)/n2 . The rest of the
2
argument applies to any graph and shows the following:
Proposition 5.8. If we can construct in G a flow fuv of value ¡π(u)π(v)
¢
from u to v for each u 6= v, and the maximum total load of these n2 flows
on any edge is at most γ, then the conductance of G is at least 1/(2mγ).
This proof method has many applications (Jerrum and Sinclair [36],
Diaconis and Stroock [21], Fill [34]). But what are its limitations, i.e., how
close can we get to the true conductance? An important theorem of Leighton
and Rao [47] shows that we never lose more than a factor of O(log n).
Theorem 5.9. Let G be a graph with conductance Φ. Then there exists
a system of flows fuv of value π(u)π(v) from u to v for each u 6= v, loading
any edge by at most O(log n)/mΦ.
There are many refinements and extensions of this fundamental result
(see e.g. Klein, Agraval, Ravi and Rao [45]; Leighton et al [46]), but these
focus on multicommodity flows and not on conductance, so we do not discuss
them here.
Shortcuts
In the last paragraphs we have sketched the following steps in estimating
the mixing rate:
mixing rate → eigenvalue gap → conductance → multicommodity flows.
It is possible to make “shortcuts” here, thereby obtaining bounds that are
often sharper and more flexible.
Diaconis and Stroock [21] and Fill [34] prove the following lower bound
on the eigenvalue gap, P
shortcutting the notion of conductance. We define
the cost of a flow f as e f (e).
Theorem 5.10. Assume that there exists a flow fuv of value π(u)π(v)
¡ ¢ from
u to v for each u 6= v, such that the maximum total load of these n2 flows
on any edge is at most γ, and the cost of each flow fuv is at most βπ(u)π(v).
Then
1
λ2 ≤ 1 −
.
2mβγ
32
L. Lovász
Lovász and Simonovits [50,51] introduced a method that estimates the
mixing rate directly using conductance or related parameters, without the
use of eigenvalue techniques. This makes the method more flexible. We
formulate one application that is implicit in these papers:
Theorem 5.11. Let t ∈ Z+ and assume that for each 0 ≤ s ≤ t and
0 ≤ x ≤ 1, every level set A = {v ∈ V : P s (v) ≥ x} has conductance at
least ψ. Then for every S ⊆ V ,
¶t
µ
p
ψ2
|P (S) − π(S)| ≤ |V | 1 −
.
4
t
In other words, if the convergence P t → π is slow, then among the level
sets of the P t there is one with small conductance. Other applications of
this method include results where sets S with “small” measure π(S) are
allowed to have small conductance.
6. Sampling by random walks
Probably the most important applications of random walks in algorithm
design make use of the fact that (for connected, non-bipartite graphs) the
distribution of vt tends to the stationary distribution π as t → ∞. In most
(though not all) cases, G is regular of some degree d, so π is the uniform
distribution. A node of the random walk after sufficiently many steps is
therefore essentially uniformly distributed.
It is perhaps surprising that there is any need for a non-trivial way
of generating an element from such a simple distribution as the uniform.
But think of the first application of random walk techniques in real world,
namely shuffling a deck of cards, as generating a random permutation of 52
elements from the uniform distribution over all permutations. The problem
is that the set we want a random element from is exponentially large (with
respect to the natural size of the problem). In many applications, it has in
addition a complicated structure; say, we consider the set of lattice points
in a convex body or the set of linear extensions of a partial order.
Random Walks on Graphs: A Survey
33
Enumeration and volume computation
The following general scheme for approximately solving enumeration problems, called the product estimator, is due to Jerrum, Valiant and Vazirani
[39], and also to Babai [10] for the case of finding the size of a group. Let
V be the set of elements we want to enumerate. The size of V is typically
exponentially large in terms of the natural “size” k of the problem. Assume
that we can find a chain of subsets V0 ⊂ V1 ⊂ . . . Vm = V such that for each
i,
(a) |V0 | is known (usually |V0 | = 1);
(b) |Vi+1 |/|Vi | is polynomially bounded (in k);
(c) m is polynomially bounded;
(d) we have a subroutine to generate a random element uniformly distributed over Vi , for each 1 ≤ i ≤ m.
Then we can estimate the ratios |Vi+1 |/|Vi | by generating a polynomial
number of elements of Vi+1 and counting how often we hit Vi . The product
of these estimates and of |V0 | gives an estimate for |V |. This scheme leads
to a randomized polynomial time approximation algorithm (provided (a),
(b), (c) and (d) are satisfied and the subroutine in (d) is polynomial).
The crucial issue is how to generate a random element of Vi in polynomial time. We discuss this question for Vm = V ; in virtually all applications
of the method, every Vi itself is of the same type as V , and so the same
arguments apply (this phenomenon is called “self-reducibility”).
As mentioned above, random walks provide a general scheme for this.
We define a connected graph G = (V, E) on which a random walk can be
implemented, i.e., a random neighbor of a given node can be generated (most
often, the nodes have small (polynomial) maximum degree). By adding
loops, we can make the graph regular and non-bipartite. Then we know
that if we stop after a large number of steps, the distribution of the last
node is very close to uniform. Our results about the mixing rate tell us how
long we have to follow the random walk; but to find good estimates of the
mixing rate (on the spectral gap, or on the conductance) is usually the hard
part.
This method for generating a random element from a combinatorial
structure was initiated by Broder [14] for the problem of approximating the
number of perfect matchings in a graph. A proof of the polynomiality of
the method was given by Jerrum and Sinclair [36] for the case of graphs
34
L. Lovász
with minimum degree at least n/2. Whether the method can be modified
to handle the case of sparse graph is an open problem.
Let us sketch this important result. Let G be a graph on n nodes with
all degrees at least n/2. We want to generate a random perfect matching
of a graph G on n nodes (n even), approximately uniformly. Therefore, we
want to define a graph whose nodes are the perfect matchings, and do a
random walk on this graph. However, there is no easy way to step from one
perfect matching to another; therefore, we extend the set we consider and
include also all near-perfect matchings (i.e., matchings with n/2 − 1 edges).
We connect two near-perfect matchings by an edge if they have n/2−2 edges
in common, and connect a perfect matching to all near-perfect matchings
contained in it, to obtain a graph H. The degrees in H are bounded by 3n;
we add loops at the nodes to make H regular of degree 3n.
Now one can construct a multicommodity flow (basically following the
tranformation of one matching to the other by alternating paths) to show
that 1/Φ(H) is polynomially bounded in n. Hence we can generate an
essentially uniformly distributed random node of H by walking a polynomial
number of steps. If this node corresponds to a perfect matching, we stop.
Else, we start again. The assumption about the degrees can be used to show
that the number of perfect matchings is at least a polynomial fraction of
the number of near-perfect matchings, and hence the expected number of
iterations before a perfect matching is obtained is polynomially bounded.
Other applications of this method involve counting the number of linear extensions of a partial order (Khachiyan and Karzanov [41]), eulerian
orientations of a graph (Mihail and Winkler [59]), forests in dense graphs
(Annan [7]), and certain partition functions in statistical mechanics (Jerrum
and Sinclair [37]). See Welsh [66] for a detailed account of fully polynomial
randomized approximation schemes for enumeration problems.
As another example, consider the fundamental problem of finding the
volume of a convex body. The exact computation of the volume is difficult,
which can be stated, and in some sense proved, in a mathematically exact
way. Dyer and Frieze [26] and Khachiyan [43] proved that computing the
volume of an n-dimensional convex polytope is #P-hard. Other results by
Elekes [29] and Bárány and Füredi [12] show that for general convex bodies
(given by, say, a separation oracle; see (Grötschel, Lovász and Schrijver [35])
for background information on the complexity of geometric algorithms) even
to compute an estimate with bounded relative error takes exponential time,
and the relative error of any polynomial time computable estimate grows
exponentially with the dimension.
Random Walks on Graphs: A Survey
35
It was a breakthrough in the opposite direction when Dyer, Frieze and
Kannan [27] designed a randomized polynomial time algorithm (i.e., an algorithm making use of a random number generator) which computes an
estimate of the volume such that the probability that the relative error is
larger than any prescribed positive number is arbitrarily small. Randomization reduces the relative error of a polynomial time approximation algorithm
from exponentially large to arbitrarily small!
Several improvements of the original algorithm followed; here are some
contributions and their running time estimates (we count the number of calls
on the separation oracle; the ∗ after the O means that we suppress factors
of log n, as well as factors depending on the error bounds): Dyer, Frieze
and Kannan [27] O∗ (n27 ), Lovász and Simonovits [50] O∗ (n16 ), Applegate
and Kannan [8] O∗ (n10 ), Lovász [49] O∗ (n10 ), Dyer and Frieze [28] O∗ (n8 ),
Lovász and Simonovits [52] O∗ (n7 ), Kannan, Lovász and Simonovits [40]
O∗ (n5 ).
Here is the general idea. Let K be a convex body in Rn . Using known
techniques from optimization, we may assume that K contains the unit ball
and is contained in a ball with radius R ≤ n3/2 . Let Ki be the intersection
of K and the ball about 0 with radius 2i/n (i = 0, 1, . . . , m = ⌈2n log n⌉).
Then K0 ⊆ K1 ⊆ . . . ⊆ Km = K, vol(Ki+1 )/vol(Ki ) ≤ 2, and vol(K0 )
is known. Thus the general scheme for enumeration described above can
be adapted, provided we know how to generate a random point uniformly
distributed in a convex body.
For this, we use random walk techniques. There is some technical
difficulty here, since the set of points in a convex body is infinite. One
can either consider a sufficiently fine grid and generate a random gridpoint
in K, or extend the notions and methods discussed above to the case of an
infinite underlying set. Both options are viable; the second takes more work
but leads to geometrically clearer arguments about mixing rates.
We define the random walk as follows. The first point is generated
uniformly from the unit ball B. Given vt , we generate a random point
u uniformly from the ball vt + B ′ with center vt and radius δ (here the
parameter δ depends on the version of the algorithm, but typically it is
√
about ε/ n with some small positive constant ε; B ′ = δB). If u ∈ K
then we let vt+1 = u; else, we generate a new point u and try again. This
procedure corresponds to the random walk on the graph whose vertex set
is K, with two points x, y ∈ K connected by an edge iff |x − y| ≤ δ.
The stationary distribution of this random walk is not the uniform
distribution, but a distribution whose density function is proportional to
36
L. Lovász
the “degrees” ℓ(x) = vol(K ∩ (x + B ′ ))/vol(B ′ ). This quantity ℓ(x) is also
called the “local conductance” at x; it is the probability that we can make
a move after a single trial. If the stepsize is sufficiently small then this
quantity, however, is constant on most of K, and the error committed is
negligible.
(In several versions of the algorithm, the graph is padded with “loops”
to make it regular. More exactly this means that if u is chosen uniformly
from vt + B ′ and u ∈
/ K, then we set vt+1 = vt . So the two random walks
produce the same set of points, but in one, repetition is also counted. It
turns out that for the description as given above, the conductance can be
estimated in a very elegant way as in Theorem 6.2 below, while in the other
version, points with small local conductance cause a lot of headache.)
Putting these together, we have the outline of the volume algorithm.
The analysis of it is, however, not quite easy. The main part of the analysis is
the estimation of the conductance of the random walk in K. The proof of the
following theorem involves substantial geometric arguments, in particular
isoperimetric inequalities.
Theorem 6.2. The conductance of the random walk in a convex body K
√
with diameter D is at least const · δ/( nD).
This implies that it takes only O∗ (nR2 /δ 2 ) steps to generate a random
point in K.
This theorem suggests that one should choose the stepsize as large as
possible. In fact, choosing δ = R would give us a random point in K in a
single step! The problem is that if δ is large, we have to make too many
trials before we can move to the next point. It is easy to calculate that in
a stationary walk, the average “waiting time”, i.e., the average number of
points u to generate before we get one in K is
ÁZ
vol(K)
ℓ(x) dx
K
√
One can prove that this quantity is bounded from above by 1/(1 − δ n),
√
and hence it is O(1) if δ is chosen less than 1/(2 n). This means that the
number of unsuccessful trials is only a constant factor more than that the
number of steps in the random walk, which is O∗ (R2 n2 ) for this choice of
the stepsize.
The issue of achieving an R that is as small as possible is crucial but
does not belong to this survey. With somewhat elaborate tricks, we can
Random Walks on Graphs: A Survey
37
√
achieve R = O( n) and hence the cost of generating a random point in
K is O∗ (n3 ). One has to generate O∗ (n) points to estimate each ratio
vol(Ki )/vol(Ki+1 ) with sufficient accuracy, and there are O∗ (n) such ratios.
This gives the total of O∗ (n5 ) steps (oracle calls).
In virtually all applications of this method, the key issue is to estimate
the conductance of the appropriate graph. This is usually a hard problem,
and there are many unsolved problems. For example, is the conductance
of a “matroid basis graph” polynomially bounded from below? (A matroid
basis graph has all bases of a matroid (E, M) as nodes, two being connected
iff their symmetric difference has cardinality 2.) This is proved for graphic
matroids (Aldous [2], Broder [15], cf. the proof of Theorem 6.6), and for
a larger class of matroids called balanced (Mihail and Feder [30]). It is
interesting to note that the property of graphic matroids that allows this
proof to go through is inequality (4.1) for the number of spanning trees.
Metropolis filter
In many applications of random walks, the distribution we want to generate a random element from is not uniform. For example, a randomized
optimization algorithm may be considered as a method of generating a random feasible solution from some probability distribution Q that is heavily
concentrated on optimal and near-optimal solutions. To be more specific,
let f : V → R+ be the objective function; then maximizing f over V is
just the extreme case when we want to generate a random element from a
distribution concentrated on the set of optimum solutions. If, instead, we
generate a random point w from the distibution Q in which Q(v) is proportional to (say) exp(f (v)/T ), where T is a very small positive number, then
with large probability w will maximize f .
The elegant method of random walk with Metropolis filter (Metropolis,
Rosenbluth, Rosenbluth, Teller and Teller [58]) describes a simple way to
modify the random walk, so that it converges to an arbitrary prescribed
probability distribution.
Let G = (V, E) be a graph; for simplicity, assume that G is d-regular.
Let F : V → R+ , and let v0 be any starting point for the random walk. Let
vt be the node where we are after t steps. We choose a random neighbor u
of vt . If F (u) ≥ F (vt ) then we move to u; else, we flip a biased coin and
move to u only with probability F (u)/F (vt ), and stay at v with probability
1 − F (u)/F (vt ).
It is clear that this modified random walk is again a Markov chain; in
fact, it is easy to check that it is also time-reversible (and so it can be con-
38
L. Lovász
sidered as a random walk on a graph with edge-weights). The “miraculous”
property of it is the following:
Theorem 6.3. The stationary distribution QF of the random walk on a
graph G filtered by a function F is given by the formula
F (v)
.
w∈V F (w)
QF (v) = P
An additional important property of this algorithm is that in order to
carry it out, we do not even have to compute the probabilities QF (v); it
suffices to be able to compute the ratios F (u)/F (vt ) = QF (u)/QF (vt ). This
property of the Metropolis filter is fundamental in some of its applications.
Unfortunately, techniques to estimate the mixing time (or the conductance) of a Metropolis-filtered walk are not general enough, and not too
many succesful examples are known. One notable exception is the work
of Applegate and Kannan [8], who proved that random walks on the lattice points in a convex body, filtered by a smooth log-concave function,
mix essentially as fast as the corresponding unfiltered walk. They applied
this technique to volume computation. Diaconis and Hanlon [22] extended
certain eigenvalue techniques to walks on highly symmetric graphs, filtered
by a function which is “smooth” and “log-concave” in some sense. Some
negative results are also known (Jerrum [38]).
Exact stopping rules
Let us start with the following funny fact.
Fact 6.4. Let G be a circuit of length n and u any starting node. Then
the probability that a random walk starting at u visits every node before
hitting v is the same for each v 6= u.
Clearly, if we replace the circuit with the complete graph, we get a
similar result. Answering a question of Graham, it was proved by Lovász
and Winkler [53] that no other graph has such a property. This follows
from the next result, which verifies in a sense the intuition that the last
node visited is more likely to be “far” than “near”. Let p(u, v) denote the
probability that a random walk starting at u visits every node before v.
Theorem 6.5. If u and v are two non-adjacent nodes of a connected graph
G and {u, v} is not a cutset, then there is a neighbor w of u such that
p(w, v) < p(u, v).
Random Walks on Graphs: A Survey
39
Consequently, if G is e.g. 3-connected, then for each v, the nodes u for
which p(u, v) is minimal are neighbors of v.
As another result leading up the question of “exact stopping rules”,
let us describe a method due to Aldous [2] and Broder [15], generating a
random spanning tree in a graph, so that each spanning tree is returned
with exactly the same probability.
Theorem 6.6. Consider a random walk on a graph G starting at node u,
and mark, for each node different from u, the edge through which the node
was first entered. Let T denote the set of marked edges. With probability
1, T is a spanning tree, and every spanning tree occurs with the same
probability.
Of course, only the second assertion needs proof, but this is not quite
trivial. Our discussion below contains a proof based on a certain coupling
idea; for a more direct proof, see Lovász [48], problem 11.58 (or work it out
yourself!)
Consider a spanning tree T with root u, and draw a (directed) edge
to each spanning tree T ′ with root v if uv ∈ E(G) and T ′ arises from T
by deleting the first edge on the path from v to u and adding the edge
uv. Let H denote the resulting digraph. Clearly each tree with root v has
indegree and outdegree d(v) in H, and hence in the stationary distribution
of a random walk on H, the probability of a spanning tree with a given
root is proportional to the degree of the root (in G). If we draw a spanning
tree from this distribution, and then forget about the root, we get every
spanning tree with the same probability.
Now observe that a random walk on G induces a random walk on H as
follows. Assume that we are at a node v of G, and at a node (T, v) in H,
where T is a spanning tree. If we move along an edge vw in G, then we can
move to a node (T ′ , w) in H by removing the first edge of the path from w
to v and adding the edge vw to the current spanning tree.
Also observe that by the time the random walk in G has visited all
nodes (or at any time thereafter), the current spanning tree in H will be
the tree formed by the last exits from each node, and the root is the last
node visited. To relate this procedure to Theorem 6.6, let us consider the
random walk on G for N steps (where N is much larger than the cover time
of G. Viewing this backward is also a legal random walk on G, since G
is undirected. If we follow that corresponding random walk on H, then it
ends up with a rooted tree (T, vN ), which is the tree of first entries for this
reverse walk, unless not all nodes of G were visited during the N returns to
40
L. Lovász
v0 . Letting N → ∞, the probability of this exception tends to 0, and the
distribution of (T, vN ) tends to the stationary distribution on H which, for
fixed vN , is uniform on spanning trees. This proves Theorem 6.6.
Looking at this proof, it is natural to ask: can we get rid of the small
error arising from the possibility that not all nodes are visited during N
steps? After all, this is easily recognized, so perhaps in these cases we
should walk a bit longer. More generally, given a random walk on a graph
(or a Markov chain), can we define a “stopping rule”, i.e., a function that
assigns to every walk on the graph (starting at a given node u) either
“STOP” or “GO”, so that (a) with probability 1, every random walk is
stopped eventually and (b) the distribution of the node where the random
walk is stopped is the stationary distribution. We also consider randomized
stopping rules, where coins may be flipped to determine whether we should
stop.
Our first example above shows that for circuits and complete graphs,
the ”last node visited” rule provides an answer to the problem (we have to
modify it a bit if we want to include the starting node too). In the case of
the second example, we want to make the stopping time N dependent on the
history: we only want to stop after we have seen all nodes of the graph G,
but also want to maintain that the walk backward from the last node could
be considered a random walk. Such a rule can be devised with some work
(we omit its details). In what follows, we give some general considerations
about this problem.
Of course, one has to be careful and avoid trivial rules like generating
a node v from the stationary distribution, and then stopping when we first
visit v. I don’t know of any clean-cut condition to rule out such trivial
solutions, but one should aim at rules that don’t use global computations,
in particular, don’t make use of an a priori knowledge of the stationary
distribution.
Stopping rules exist for quite general Markov chains. Asmussen, Glynn
and Thorisson [9] describe a randomized algorithm that generates an element from the stationary distribution of a finite irreducible Markov chain,
which needs only the number of states and a “black box” that accepts a state
as an input and then simulates a step from this state. Lovász and Winkler
[54] have found a randomized stopping rule that generates an element from
the stationary distribution of any irreducible Markov chain, and only needs
to know the number of states. This rule can be made deterministic under
the assumption that the chain is aperiodic.
41
Random Walks on Graphs: A Survey
To indicate the flavor of the result, let us describe the case when the
Markov chain has two states. The general case follows by a (not quite
trivial) recursive construction (similarly as in the work of Asmussen, Glynn
and Thorisson [9]).
So let
v0 , v1 , v2 , . . .
(6.1)
be an irreducible aperiodic Markov chain on states {u, v}. Irreducible means
that the transition probabilities puv , pvu are positive; aperiodocity means
that at least one of puu and pvv is also positive. It is easy to check that the
stationary distribution is given by
π(u) =
pvu
,
puv + pvu
π(v) =
puv
,
puv + pvu
The following randomized stopping rule generates a random element from
π, without knowing any value pij or π(i), only looking at the sequence (6.1):
Rule 1. Flip a coin. If the result is head, let i = 0; else, let i be the first
index for which vi 6= v0 . If vi+1 6= vi then output vi+1 ; else, discard the first
i + 1 elements and repeat.
If you don’t like that we use coin flipping, you can use the Markov chain
itself to simulate it, making the rule entirely deterministic.
Rule 2. Wait for the first pair i < j with the following properties: (i)
vj = vi , (ii) vj+1 6= vi+1 , (iii) vj+2 6= vj+1 , and moreover, (iv) the state vi
occurs an even number of times before vi and (v) not at all between vi and
vj . Output vj+2 .
If this sounds mysterious, note that for each of the first, second, etc.
occurence of a pair of indices with (i), (ii), (iv) and (v), vj+1 can be either
of the states with probability 1/2.
The stopping rule sketched above takes a lot of time; we don’t even know
how to make the expected number of steps of the random walk polynomial
in the maximum access time, let alone comparable with the mixing time
(that we know may be logarithmic in n). On the other hand, if we allow
global computation, we can get a stopping rule which needs, on the average,
at most twice as many steps as the mixing time τ . We follow the random
walk for τ steps, then “flip a biased coin”; with probability π(vτ )/2Pτ (vτ ),
we stop; with probability 1 − π(vτ )/2Pτ (vτ ), we forget about the past and
start from vτ a random walk of length τ etc. It is easy to see that the
42
L. Lovász
probability that we stop at v after k rounds is 2−k π(v), which adds up to
π(v). Also, the expected number of steps is 2τ .
A threshold rule is a (relatively) simple kind of stopping rule. It is
specified by a function t : V → R+ , depending on the staring point v0 , and
works as follows:
if t(vk ) ≤ k, then stop;
if t(vk ) ≥ k + 1, go on;
if k < t(vk ) < k + 1 then “flip a biased coin” and move with probability
t(vk ) − k but stop with probability k + 1 − t(vk ).
Lovász and Winkler [55] have shown that there is a function t that gives
a threshold rule that is optimal among all stopping rules in a very strong
sense: it minimizes the expected number of steps among all randomized
stopping rules (for a fixed starting node). It also minimizes the expected
number of times any given node is visited. Every threshold rule is of course
finite, in the sense that there is a finite time T such that it is guaranteed
to stop within T steps (in fact, T ≤ maxi t(i)). The optimal threshold rule
minimizes this bound among all finite rules.
The expected number of steps for the optimal threshold rule, starting
at node v, is
X
π(u)H(u, v).
τ ∗ = max H(u, v) −
u
u
It follows from the description of the stopping rule using the mixing time
that
τ ∗ ≤ 2τ.
Since the definition of the mixing time τ has an arbitrarily chosen constant
1/2 in it, while the definition of τ ∗ is “canonical”, it should be more natural
to call the quantity τ ∗ the mixing time.
Since this optimal stopping rule has many nice properties, it would be
good to have an efficient implementation. The threshold function is polynomially computable; but this is not good enough since we want to apply these
rules to exponentially large graphs. However, one can describe simple, easily
implementable stopping rules with comparable expected time that achieve
approximate mixing on the exponentially large graphs of interest discussed
above.
Random Walks on Graphs: A Survey
43
References
[1] D. J. Aldous, Lower bounds for covering times for reversible Markov chains and
random walks on graphs, J. Theoretical Probability 2(1989), 91–100.
[2] D. J. Aldous, The random walk construction for spanning trees and uniform labelled
trees, SIAM J. Discrete Math. 3(1990), 450–465.
[3] D. J. Aldous and J. A. Fill, Reversible Markov Chains and Random Walks on Graphs
(book in preparation).
[4] R. Aleliunas, R. M. Karp, R. J. Lipton, L. Lovász, C. W. Rackoff, Random walks,
universal travelling sequences, and the complexity of maze problems, in: Proc. 20th
Ann. Symp. on Foundations of Computer Science (1979), 218–223.
[5] N. Alon, Eigenvalues and expanders, Combinatorica 6(1986), 83–96.
[6] N. Alon and V. D. Milman, λ1 , isoperimetric inequalities for graphs and superconcentrators, J. Combinatorial Theory B 38(1985), 73–88.
[7] J. D. Annan, A randomised approximation algorithm for counting the number of
forests in dense graphs, Combin. Probab. Comput. 3(1994) 273–283.
[8] D. Applegate and R. Kannan, Sampling and integration of near log-concave functions,
Proc. 23th ACM STOC (1990), 156–163.
[9] S. Asmussen, P. W. Glynn and H. Thorisson, Stationary detection in the initial transient problem, ACM Transactions on Modeling and Computer Simulation 2(1992),
130–157.
[10] L. Babai, Monte Carlo algorithms in graph isomorphism testing, Université de
Montréal Tech. Rep. DMS 79-10, 1979 (42).
[11] L. Babai and M. Szegedy, Local expansion of symmetrical graphs, Combinatorics,
Probability, and Computing 1(1992), 1–11.
[12] I. Bárány and Z. Füredi, Computing the volume is difficult, Proc. of the 18th Annual
ACM Symposium on Theory of Computing (1986), 442–447.
[13] G. Brightwell and P. Winkler, Maximum hitting time for random walks on graphs,
J. Random Structures and Algorithms 1(1990), 263–276.
[14] A. Broder, How hard is it to marry at random? (On the approximation of the
permanent), Proc. 18th Annual ACM Symposium on Theory of Computing (1986),
50–58.
[15] A. Broder, Generating random spanning trees, Proc. 30th Annual Symp. on Found.
of Computer Science IEEE Computer Soc., (1989), 442–447.
[16] A. K. Chandra, P. Raghavan, W. L. Ruzzo, R. Smolensky and P. Tiwari, The
electrical resistance of a graph captures its commute and cover times, Proc. 21st
ACM STOC, (1989), 574–586.
[17] F. R. K. Chung and S. T. Yau, Eigenvalues of graphs and Sobolev inequalities, (1993)
preprint.
[18] D. Coppersmith, U. Feige, and J. Shearer, Random Walks on Regular and Irregular
Graphs, Technical report CS93-15 of the Weizmann Institute 1993.
[19] D. Coppersmith, P. Tetali and P. Winkler, Collisions among random walks on a
graph, SIAM J. Discr. Math. 6(1993), 363–374.
[20] P. Diaconis, Group Representations in Probability and Statistics, Inst. of Math.
Statistics, Hayward, Californis, 1988.
44
L. Lovász
[21] P. Diaconis and D. Stroock, Geometric bounds for eigenvalues of Markov chains,
Annals of Appl. Prob. 1(1991), 36–62.
[22] P. Diaconis and P. Hanlon, Eigen analysis for some examples of the Metropolis
algorithm, Hypergeometric functions on domains of positivity, Jack polynomials, and
applications, Contemporary Math. 138(1992) Amer. Math. Soc., Providence, RI
[23] P. Diaconis and L. Saloff-Coste, Comparison theorems for random walk on finite
groups, Ann. Prob. 21(1993), 2131–2156.
[24] J. Dodziuk and W. S. Kendall, Combinatorial Laplacians and isoperimetric inequality, in: From Local Times to Global Geometry, Control and Physics (ed.: K. D. Ellworthy), Pitman Res. Notes in Math. Series 150(1986), 68–74.
[25] P. G. Doyle and J. L. Snell, Random walks and Electric Networks, MAA, 1984.
[26] M. Dyer and A. Frieze, On the complexity of computing the volume of a polytope.
SIAM J. Comp. 17(1988), 967–974.
[27] M. Dyer, A. Frieze and R. Kannan, A Random Polynomial Time Algorithm for Approximating the Volume of Convex Bodies, Proc. of the 21st Annual ACM Symposium
on Theory of Computing (1989), 375–381.
[28] M. Dyer and A. Frieze, Computing the volume of convex bodies: a case where
randomness provably helps, in: Probabilistic Combinatorics and Its Applications (ed.:
B. Bollobás), Proceedings of Symposia in Applied Mathematics, 44(1992), 123–170.
[29] G. Elekes, A geometric inequality and the complexity of computing volume, Discrete
and Computational Geometry 1(1986), 289–292.
[30] T. Feder and M. Mihail, Balanced matroids, Proc. 24rd ACM Symp. on Theory of
Comp. (1992), 26–38.
[31] U. Feige, A Tight Upper Bound on the Cover Time for Random Walks on Graphs,
Random Structures and Algorithms 6(1995), 51–54.
[32] U. Feige, A Tight Lower Bound on the Cover Time for Random Walks on Graphs,
Random Structures and Algorithms 6(1995), 433–438.
[33] U. Feige, Collecting Coupons on Trees, and the Analysis of Random Walks, Technical
report CS93-20 of the Weizmann Institute 1993.
[34] J. A. Fill, Eigenvalue bounds on the convergence to stationarity for nonreversible
Markov chains, with an application to the exclusion process, The Annals of Appl.
Prob. 1(1991), 62–87.
[35] M. Grötschel, L. Lovász and A. Schrijver, Geometric Algorithms and Combinatorial
Optimization, Springer-Verlag, 1988.
[36] M. R. Jerrum and A. Sinclair, Approximating the permanent, SIAM J. Comput.
18(1989), 1149–1178.
[37] M. R. Jerrum and A. Sinclair, Polynomial-time approximation algorithms for the
Ising model, Proc. 17th ICALP, EATCS (1990), 462–475.
[38] M. R. Jerrum, Large cliques elude the Metropolis process, Random Structures and
Algorithms 3(1992), 347–359.
[39] M. R. Jerrum, L. G. Valiant and V. V. Vazirani, Random generation of combinatorial
structures from a uniform distribution, Theoretical Computer Science 43(1986), 169–
188.
[40] R. Kannan, L. Lovász and M. Simonovits, Random walks and a faster volume
algorithm, (in preparation).
[41] A. Karzanov and L. G. Khachiyan, On the conductance of order Markov chains,
Order 8(1991), 7–15.
Random Walks on Graphs: A Survey
45
[42] J. Keilson, Markov Chain Models – Rarity and Exponentiality, Springer-Verlag, 1979.
[43] L. G. Khachiyan, The problem of computing the volume of polytopes is NP-hard,
Uspekhi Mat. Nauk 44(1989), 199–200.
[44] L. G. Khachiyan, Complexity of polytope volume computation, in: New Trends in
Discrete and Computational Geometry (ed.: J. Pach), Springer, (1993), 91–101.
[45] P. Klein, A. Agraval, R. Ravi and S. Rao, Approximation through multicommodity
flow, Proc. 31st Annual Symp. on Found. of Computer Science, IEEE Computer
Soc., (1990), 726–727.
[46] T. Leighton, F. Makedon, S. Plotkin, C. Stein, É. Tardos, and S. Tragoudas, Fast
approximation algorithms for multicommodity flow problem, Proc. 23rd ACM Symp.
on Theory of Comp. (1991), 101-111.
[47] F. T. Leighton and S. Rao, An approximate max-flow min-cut theorem for uniform
multicommodity flow problems with applications to approximation algorithms, Proc.
29th Annual Symp. on Found. of Computer Science, IEEE Computer Soc., (1988),
422-431.
[48] L. Lovász, Combinatorial Problems and Exercises, Akadémiai Kiadó, Budapest –
North Holland, Amsterdam, 1979, 1993.
[49] L. Lovász, How to compute the volume? Jber. d. Dt. Math.-Verein, Jubiläumstagung
1990, B. G. Teubner, Stuttgart, (1992), 138–151.
[50] L. Lovász and M. Simonovits, Mixing rate of Markov chains, an isoperimetric inequality, and computing the volume, Proc. 31st Annual Symp. on Found. of Computer Science, IEEE Computer Soc., (1990), 346–355.
[51] L. Lovász and M. Simonovits, On the randomized complexity of volume and diameter,
Proc. 33rd IEEE FOCS (1992), 482–491.
[52] L. Lovász and M. Simonovits, Random walks in a convex body and an improved
volume algorithm Random Structures and Alg. 4(1993), 359–412.
[53] L. Lovász and P. Winkler, On the last new vertex visited by a random walk, J. Graph
Theory 17(1993), 593–596.
[54] L. Lovász and P. Winkler, Exact mixing in an unknown Markov chain, The Electronic
Journal of Combinatorics 2(1995), paper R15, 1–14.
[55] L. Lovász and P. Winkler, Efficient stopping rules for Markov chains, Proc. 27th
ACM Symp. Theory of Comput. (1995), 76–82.
[56] P. Matthews, Covering problems for Brownian motion on spheres, Ann. Prob. 16
(1988), 189–199.
[57] P. Matthews, Some sample path properties of a random walk on the cube, J. Theoretical Probability 2(1989), 129–146.
[58] N. Metropolis, A. Rosenblut, M. Rosenbluth, A. Teller amd E. Teller, Equation of
state calculation by fast computing machines, J. Chem. Physics 21(1953), 1087–
1092.
[59] M. Mihail and P. Winkler, On the number of Eulerian orientations of a graph, Extended Abstract, Proc. 3rd ACM-SIAM Symposium on Discrete Algorithms (1992),
138–145; full version to appear in Algorithmica.
[60] C. St.J. A. Nash-Williams, Random walks and electric currents in networks, Proc.
Cambridge Phil. Soc. 55(1959), 181–194.
[61] G. Polya, Über eine Aufgabe der Wahrscheinlichkeitsrechnung betreffend die Irrfahrt
im Strassennetz, Math. Annalen 84(1921), 149–160.
46
L. Lovász
[62] A. Sinclair and M. Jerrum, Conductance and the rapid mixing property for Markov
chains: the approximation of the permanent resolved, Proc. 20th ACM STOC, (1988),
235–244.
[63] P. Tetali, Random walks and effective resistance of networks, J. Theoretical Probability 1(1991), 101–109.
[64] P. Tetali and P. Winkler, Simultaneous reversible Markov chains, in: Combinatorics,
Paul Erdős is Eighty, Vol. 1 (eds:. D. Miklós, V. T. Sós, T. Szőnyi), Bolyai Society
Mathematical Studies, Budapest, 1(1993), 422–452.
[65] C. Thomassen, Resistences and currents in infinite electrical networks, J. Comb.
Theory 49(1990), 87–102.
[66] D. J. A. Welsh, Complexity: Knots, Colourings and Counting, London Math. Soc.
Lecture Note Series 186, Cambridge University Press, 1993.
László Lovász
Department of Computer Science,
Yale University,
New Haven, CT 06520
USA
e-mail:
lovasz@cs.yale.edu