Exactly resolvable model of random medium 
 
based on the stable law theory
V.V. Uchaikin and G.G. Gusarov 
 
Moscow State University Branch in Ulyanovsk, Ulyanovsk, Russia
 One of the most interesting application of statistical mechanical
 concepts is the large-scale distribution of matter in the universe.
 Most of its  properties have been obtained by a statistical
 correlation analysis for catalogs of galaxies and clusters [1,2].
 The galaxies are considered as the smallest individual entities and
 their distribution is characterized  by points  
 
 in the 3-dimensional Euclidian space.
 The standard analysis consists in the calculation of
 two - point correlation function
 
 
 
Observations show, that C(r) can be approximated by a power law
 
 
This fact is the basis for the assumption about of a fractal structure of the universe [2].
At first the use of walk method for mathematical simulating such distribution of galaxies was offered by Mandelbrot [3]. The process starts from choosing the position where the first galaxy to be, the next galaxy is placed in a random isotropic direction and at a random distance l from the previous one while l is chosen from the distribution
 
 
This procedure repeats many (maybe infinitely large number of) times and every time the value and direction of l are independently chosen. Based on such probability distribution the model cannot be solved exactly.
 In present work it is shown that choice of 3-dimensional spherically
 symmetric stable distribution  
 ,
 having the same asymptotic behaviour as power law  
 ,
  instead of the  
 
 leads to exact
 analytical expressions for  irreducible correlation functions of
 different orders and the cut off probability 1 - q at each step:
 
 
 
Here  
  is the average number of all the particles per
 unit  volume,  
  and  
  means the summation
 over all n!  permutations of the arguments.
 The characteristic exponent of stable law  
  so the
 stable distribution densities form a whole family. The distribution
  
  is a normal (Gauss') law,  
  is Cauchy's law.
 Other distributions  
  cannot be expressed in
 elementary functions and are calculated and tabulated in this work.