Optical Routing in Massively Dense Networks: Practical Issues and Dynamic Programming Interpretation

A form of routing based on an analogy with geometrical optics has recently been proposed for use in wireless networks with a very large number of nodes. The analogy hinges on taking a macroscopic view of the network, and describing it solely in terms of a cost function, that is analogous to the refractive index of optics. In this setting, optimal routes resemble rays of light propagating through a properly defined optical medium. This method has intuitive appeal and leads to insightful results; nevertheless many issues remain unresolved. Two of these issues are treated here. First, we discuss the problem of route discovery, also considering the challenging cases where the optimal routes starting from a given node intersect themselves, and also fail to cover the whole network in a reasonable manner. Then, we show how optical routing can be derived using dynamic programming. The derivation illuminates the connection between this form of routing and traditional routing algorithms that are also based on dynamic programming.