Survival and effectiveness of a weapon system against a superior force (guerrilla warfare)

Stochastic models are developed for a weapon system which attacks at a certain rate, but withdraws when attacked (guerrilla warfare). The models yield as output the distributions and mean values in closed form of the survival period and number of attacks made. As input are, for the weapon system, attack rate or amount of ammunition and time horizon, and, for the opponent, kill rate and probability of killing the weapon system, given the opponent has been attacked (strike back performance of the opponent). One model allows, however, for strike back also by the weapon system, when attacked by the opponent. This model is then used to determine, when the weapon system should strike back. The models are based on the Poisson and the binomial processes. Consistency among the models is shown and an example is provided.