The paper gives an explicit expression for the expectation of the maximum attainable fraction of served customers in the long run for the single-server loss system GI/G/1/0, under the assumption of perfect information regarding the sequences {Xi, i = 1, 2, ·· ·} and {Yi, i = 1, 2, ·· ·} of interarrival times and service times, respectively. A heavy traffic result for this fraction is obtained for the system GI/M/1/0. The general result is based on an analysis of the random interval graph corresponding to the random intervals {[Ti, Ti + Yi), i = 1, 2, ·· ·}, in which {Ti} denotes the sequence of arrival epochs.

A new queueing system called G/G/{p} is introduced and studied. In this queue, unlike standard queues, the customers after being served are allowed to become servers themselves. More precisely, at the completion of his service each customer is assumed to become a server with probability p or leave the system with probability 1 – p, independent of everything else. We make some comparisons about the waiting times and queue sizes among different queueing systems. We also study the joint distribution of the queue size, the number of servers and the number of departures at time t for exact and asymptotic behavior for large t.

It is well known that a random set determines its random coverage measure. The paper gives a necessary and sufficient condition for the reverse implication. An equivalent formulation of the condition constitutes a first step in the search for a way to recognize a random measure as being the random coverage measure of a random set.

The Bernoulli/exponential target process is considered. Such processes have been found useful in modelling the search for active compounds in pharmaceutical research. An inequality is presented which improves a result of Gittins (1989), thus providing a better approximation to the Gittins indices which define the optimal search policy.