@InProceedings{XuTurner89, title = "{A}n {I}rregular {D}istributed {S}imulation {P}roblem with a {D}ynamic {L}ogical {P}rocess {S}tructure", author= "Xu, Ming Q. and Turner, Stephen J. and Pin, Nie", editor= "Wexler, J.", pages = "69--79", booktitle= "{OUG}-11: {D}eveloping {T}ransputer {A}pplications", isbn= "90 5199 020 0", year= "1989", month= "sep", abstract= "This paper describes a modeling problem which exhibits many features of more advanced distributed simulation: the simulation of biological population dynamics — \"Host-parasite interactions\". It is a dynamical simulation in which certain species of hosts and parasites live, move randomly, breed and (as far as parasites are concerned) infect the hosts in a two-dimensional ocean. Apart from its relevance to realistic biological studies, this simulation program does serve to illustrate many crucial ideas in dynamic time and event driven simulations. Our approach to the parallel implementation of this simulation requires the simulation objects (in this case, hosts and parasites) to be organised as LPs (Logical Processes) which can be created and destroyed dynamically at run time to reflect the birth/death of these simulation objects. In addition to their dynamical features, LPs must preserve the temporal aspects of the real world system. In other words, a global ordering of LP interactions (referred to here as actions) must be ensured to preserve the causality principle [1]. Also, the computational load can become imbalanced because the real world system or rather, the distribution of hosts and parasites in the underlying space is changing with time. The methods for counteracting this dynamical load imbalance will be described.We shall begin with the specification of the dynamical rules governing the behaviour of the hosts and parasites during the simulation." }