A Rapid Probabilistic Potential Solver and its Implications for Fast Electromagnetic Field Solutions

This paper shows how the Monte Carlo method can be used to solve probabilistic potential problems in a fraction of the time required until now. Previous methods have relied on fast sequential processing using mainframe computers. The method presented here exploits the natural concurrency of the Monte Carlo approach. Computation is by a low cost transputer system linked to an PC-AT. The concurrent methodology is powerful and is applicable to many other modelling techniques used for electromagnetic field modelling. A microstrip geometry is considered since this exhibits most of the features required in a general problem. The considerations for designing the parallel algorithm are given. These considerations show the inherent strength of the concurrent approach and its flexibility in application to a broad class of passive electromagnetic field problems. The implications of adopting the concurrent methodologies suggested in this paper are of strategic importance in determining the direction in which passive microwave modelling should proceed in the future.