Weak interaction limits for one-dimensional random polymers (Q1396753): Difference between revisions

A series of limit theorems are established for a whole class of one-dimensional self-repellent random walks; these walks being considered in the limit of vanishing self-repellence or in the limit of diverging step variance. The strength of these results lies in their universality: The authors prove e.g. that the height of the walk after \(n\) steps satisfies a law of large numbers, and that the scaling coefficients appearing in this LLN do not depend on the details of the walk other than its step variance. These scaling coefficients may all be expressed in terms of the Edwards model (or self-repellent Brownian motion), and most of the proofs in the present paper rely in an essential way on the results obtained in the companion paper [Ann. Probab. 31, 2003--2039 (2003)]. More precisely, approximative rate functions are introduced for the empirical drift of the path, and it is shown that these rate functions scale to the rate function of the Edwards model. This approach proves to be very flexible in the present context, and it yields improved convergence results where the probabilities of rare events are being controlled exponentially.