In stochastic models for queues and their networks, random events evolve in time. A process for their backward evolution is referred to as a
time‐reversed process
. If some property is unchanged under time reversal, we may better understand the model. A concept of reversibility is invented for this invariance. Local balance for a stationary Markov chain has been used, but it is still too strong for queueing applications.
We are concerned with a continuous‐time Markov chain, but do not assume it has a stationary distribution. We define reversibility in structure as an invariant property of a family of the set of models under a certain operation. Any member of this set is a pair of transition rate functions and its supporting measure, and each set represents dynamics of queueing systems such as arrivals and departures. We use a permutation Γ of the family members to describe the change of the dynamics under time reversal. This reversibility is called Γ
‐reversibility in structure
.
To apply these definitions, we introduce new classes of models, called
reacting systems
and
self‐reacting systems
. Using them, we give a unified view for queues and their networks. They include symmetric service, batch movements, and state‐dependent routing.