In this chapter, we continue the exploration of dualities for independent random walkers. By taking the “many particle” limit, a deterministic process (a system of ODEs) arises as a dual process, with a duality function which is also obtained by a proper rescaling of the self-duality function of Chap. 3 . This deterministic process is in turn self-dual with a very simple duality function. From the algebraic perspective, we show that behind all these dualities there is always the same abstract object, which is written in terms of the Heisenberg algebra generators. The dualities then arise by considering several representations of the algebra. From an analytic perspective we introduce generating functions to show the equivalence of all these dualities. When interpreted as Poisson averaging, the generating function also serves as an intertwining operator. Furthermore, the use of generating functions gives a full classification of all product self-dualities for independent random walkers, which can essentially be of two types: either the triangular single-site self-dualities of Chap. 3 or self-dualities involving Charlier polynomials. We close the chapter with two applications. First, by using the duality with the deterministic system, we complete the ergodic theory of independent random walkers on the infinite lattice \(\mathbb Z^d\) , removing the restrictive assumption of an appropriate moment growth condition that was necessary in Chap. 3 . Second, for asymmetric random walkers, we state duality with the reversed process and use this to compute the joint moment generating function of currents along edges.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Duality for Independent Random Walkers

  • Cristian Giardinà,
  • Frank Redig

摘要

In this chapter, we continue the exploration of dualities for independent random walkers. By taking the “many particle” limit, a deterministic process (a system of ODEs) arises as a dual process, with a duality function which is also obtained by a proper rescaling of the self-duality function of Chap. 3 . This deterministic process is in turn self-dual with a very simple duality function. From the algebraic perspective, we show that behind all these dualities there is always the same abstract object, which is written in terms of the Heisenberg algebra generators. The dualities then arise by considering several representations of the algebra. From an analytic perspective we introduce generating functions to show the equivalence of all these dualities. When interpreted as Poisson averaging, the generating function also serves as an intertwining operator. Furthermore, the use of generating functions gives a full classification of all product self-dualities for independent random walkers, which can essentially be of two types: either the triangular single-site self-dualities of Chap. 3 or self-dualities involving Charlier polynomials. We close the chapter with two applications. First, by using the duality with the deterministic system, we complete the ergodic theory of independent random walkers on the infinite lattice \(\mathbb Z^d\) , removing the restrictive assumption of an appropriate moment growth condition that was necessary in Chap. 3 . Second, for asymmetric random walkers, we state duality with the reversed process and use this to compute the joint moment generating function of currents along edges.