Application of Financial Risk-reward Theory to Link and Network Optimization

In this work, we introduce a general framework for analysis and optimization of adaptive transmission systems in information-unstable channels. In information-unstable channels, the information density does not converge to a one-point measure and the maximum achievable transmission rate is seen as a random variable because it depends on the actual channel state. For that reason, instead of conventional channel capacity, we propose to use new performance indicators such as expected utility and riskiness that are commonly used to order probability density functions in axiomatic decision theory. We show the analogies between decision-theoretic problems and information-theoretic problems in adaptive transmission systems. We present different single-user, multi-user, and multi-terminal communication scenarios and map them to various rationality concepts and uncertainty models in decision theory. To the author’s best knowledge, adaptive transmission systems and networks have not been yet analyzed within the framework of rational decision theory. Distribution A: Approved for public release; distribution is unlimited.