Search

In this paper, we present two methods which generalize static team theory to dynamic team theory, in the context of continuous-time stochastic nonlinear differential decentralized decision systems, with relaxed strategies, which are measurable to different noisy information structures. For both methods we apply Girsanov's measure transformation to obtain an equivalent dynamic team problem under a reference probability measure, so that the observations and information structures available for decisions, are not affected by any of the team decisions. The first method is based on function space integration with respect to products of Wiener measures, and generalizes Witsenhausen's [1] definition of equivalence between discrete-time static and dynamic team problems. The second method is based on stochastic Pontryagin's maximum principle. The team optimality conditions are given by a "Hamiltonian System" consisting of forward and backward stochastic differential equations, and a conditional variational Hamiltonian with respect to the information structure of each team member, expressed under the initial and a reference probability space via Girsanov's measure transformation. Under global convexity conditions, we show that that PbP optimality implies team optimality. In addition, we also show existence of team and PbP optimal relaxed decentralized strategies (conditional distributions), in the weak$^*$ sense, without imposing convexity on the action spaces of the team members. Moreover, using the embedding of regular strategies into relaxed strategies, we also obtain team and PbP optimality conditions for regular team strategies, which are measurable functions of decentralized information structures, and we use the Krein-Millman theorem to show realizability of relaxed strategies by regular strategies., Comment: 50 pages

In this paper we consider non convex control problems of stochastic differential equations driven by relaxed controls. We present existence of optimal controls and then develop necessary conditions of optimality. We cover both continuous diffusion and Jump processes., Comment: Pages 23, Submitted to SIAM Journal on Control and Optimization

Decentralized optimization of distributed stochastic differential systems has been an active area of research for over half a century. Its formulation utilizing static team and person-by-person optimality criteria is well investigated. However, the results have not been generalized to nonlinear distributed stochastic differential systems possibly due to technical difficulties inherent with decentralized decision strategies. In this first part of the two-part paper, we derive team optimality and person-by-person optimality conditions for distributed stochastic differential systems with different information structures. The optimality conditions are given in terms of a Hamiltonian system of equations described by a system of coupled backward and forward stochastic differential equations and a conditional Hamiltonian, under both regular and relaxed strategies. Our methodology is based on the semi martingale representation theorem and variational methods. Throughout the presentation we discuss similarities to optimality conditions of centralized decision making., Comment: 39 pages, 1 figure Submitted to IEEE Transactions on Automatic Control

In this second part of our two-part paper, we invoke the stochastic maximum principle, conditional Hamiltonian and the coupled backward-forward stochastic differential equations of the first part [1] to derive team optimal decentralized strategies for distributed stochastic differential systems with noiseless information structures. We present examples of such team games of nonlinear as well as linear quadratic forms. In some cases we obtain closed form expressions of the optimal decentralized strategies. Through the examples, we illustrate the effect of information signaling among the decision makers in reducing the computational complexity of optimal decentralized decision strategies., Comment: 39 pages Submitted to IEEE Transaction on Automatic Control

The relation between nonanticipative Rate Distortion Function (RDF) and filtering theory is discussed on abstract spaces. The relation is established by imposing a realizability constraint on the reconstruction conditional distribution of the classical RDF. Existence of the extremum solution of the nonanticipative RDF is shown using weak$^*$-convergence on appropriate topology. The extremum reconstruction conditional distribution is derived in closed form, for the case of stationary processes. The realization of the reconstruction conditional distribution which achieves the infimum of the nonanticipative RDF is described. Finally, an example is presented to illustrate the concepts., Comment: 41 pages,5 figures, Provisionally Accepted Version of the manuscript submitted to IEEE Transactions on Automatic Control; Part of this work was presented in 20th International Symposium on Mathematical Theory of Networks and Systems (MTNS '12)