This article presents deterministic fluid queueing models to characterize and improve the performance of counterterror intelligence systems. We focus on the collection and analysis of intelligence information with the goal of preventing as many terror attacks as possible. Terror plots arise at an exogenous rate, and a subset of these are collected by the intelligence system. However, collection activity also generates fake plots, which can only be distinguished from real plots by intelligence analysts. Collected plots are placed in the Intel queue while awaiting analysis. We model the rate with which terror plots are prevented as a function of the queueing discipline (FIFO, SRO, LIFO) employed, and demonstrate the optimality of LIFO for a given fixed deployment of resources to collection and analysis. For the case of a fixed overall budget governing intelligence collection and analysis, we determine the optimal number of analysts (and hence the optimal collection rate) to maximize the interdiction of terror attacks; our result stands in contrast to the observed state of overcollection relative to analysis. We also consider a stochastic loss model for this problem, and show that the optimal division of resources and terror attack interdiction rates are similar to those obtained earlier.
展开▼