Optimising the flow of information within a C3I network

It is proposed that problems of efficiently responding to a situation in a Command, Control, Communications, and Intelligence (C3I) network can be investigated using dynamic programming techniques. An initial model in the form of an integer linear program is decomposed to reveal a recurrence relation, which can then be solved using a method of successive approximations. e the recurrence relation cannot be directly realised as an algorithm, a modified problem is proposed that considers only those candidate solutions having at most some prescribed number of transmission and transform operations. Solving this for successively increasing numbers of acceptable transforms eventually yields the optimum for the original model. he understanding that the set of all possible allocations of information to nodes of the network may be infinite, some way of ensuring eventual termination of the algorithm is required. Two possibilities for limiting the state space are explored here; the modified problem can be solved for some suitably large maximal number of operations, or a known upper bound on the optimal cost can be assumed. rk presented here is intended principally as a theoretical basis for future development. Practical application of this work will further require consideration for stochastic uncertainty, reliability measures, and reduced exploration of the state-space.