TABLE OF CONTENTS The material in this appendix was originally written by one of the authors as a contribution to the BSI report [14]. The permission of the chairman of the PNO-ISC to reproduce it here is gratefully acknowledged.
Consider the class of overload controls defined in section 7.6.4. It is simple to analyse the steady state behaviour of all controls in this class. To do this, some notation is required. Let there be n source nodes, each being a point of restriction. Let ? i denote the rate at which the restriction process at source i associated with the route to the overloaded node admits calls which are then offered to the overloaded node. Let:
denote the total calling rate offered to the overloaded node, and let B( ?) denote the probability that the overloaded node rejects a call from any source due to internal overload. The rate at which calls from source i are rejected by the overloaded node is given by:
In equilibrium, when control is active at all sources, this reject rate equals the locally configured target reject rate at source node i, denoted by r i:
From equation (7.1), equation (7.2) and equation (7.3), it follows that:
where R denotes the sum of the individual source node target reject rates.
If the overloaded node's call rejection probability B( ?) is a continuous...
