Dissecting the routing architecture of self-organizing networks

The proper operation of self-organizing networks (SONs) relies on the autonomous behavior of their individual nodes. Routing in such networks has been a challenging task since their conception, due to their nontraditional characteristics and design requirements. Although a large amount of routing architectures and protocols for SONs has been proposed, very little work has been done on the fundamental characteristics that make a routing strategy efficient for a particular network and/or design requirement. Contrary to traditional techniques where the routing architecture is structured as a single unit, we suggest in this article that routing be thought of as a combination of four main architectural components, namely, addressing, dissemination, discovery, and forwarding. This logical decomposition offers significant advantages from both the analysis and the design perspectives. We conclude from our observations that routing architectures should be scenario-driven, in the sense that the configuration parameters are not necessarily universally good for all application scenarios