Abstract. Simple random walk is well understood. However, if we condition a random walk not to intersect itself, so that it is a self-avoiding walk, then it is much more difficult to analyse and many of the important mathematical problems remain unsolved. This paper provides an overview of some of what is known about the critical behaviour of the self-avoiding walk, including some old and some more recent results, using methods that touch on combinatorics, probability, and statistical mechanics.2010 Mathematics Subject Classification. Primary 60K35, 82B41. Keywords. Self-avoiding walks, critical exponents, renormalisation group, lace expansion.
Self-avoiding walksThis article provides an overview of the critical behaviour of the self-avoiding walk model on Z d , and in particular discusses how this behaviour differs as the dimension d is varied. The books [29,40] are general references for the model. Our emphasis will be on dimensions d = 4 and d ≥ 5, where results have been obtained using the renormalisation group and the lace expansion, respectively.An n-step self-avoiding walk from x ∈ Z d to y ∈ Z d is a map ω : {0, 1, . . . , n} → Z d with: ω(0) = x, ω(n) = y, |ω(i + 1) − ω(i)| = 1 (Euclidean norm), and ω(i) = ω(j) for all i = j. The last of these conditions is what makes the walk self-avoiding, and the second last restricts our attention to walks taking nearestneighbour steps.Let d ≥ 1. Let S n (x) be the set of n-step self-avoiding walks on Z d from 0 to x. Let S n = ∪ x∈Z d S n (x). Let c n (x) = |S n (x)|, and let c n = x∈Z d c n (x) = |S n |. We declare all walks in S n to be equally likely: each has probability c −1 n . See Figure 1. We write E n for expectation with respect to this uniform measure on S n .
What it is not:• It is not the so-called "true" or "myopic" self-avoiding walk, i.e., the stochastic process which at each step looks at its neighbours and chooses uniformly from those visited least often in the past -the two models have different critical behaviour (see [28,50] for recent progress on the "true" self-avoiding walk).• It is by no means Markovian.• It is not a stochastic process: the uniform measures on S n do not form a consistent family.