Planning Robust and Scalable OSPF Domains for Future Battlefield Networks

Automated network planning utilities will potentially perform an important role in optimizing the performance of future battlefield networks (FBNs). Among the network planning goals is the design of scalable routing domains that accommodate 100s or even 1000s of nodes. Of particular interest is the design of scalable routing domains based on the widely-deployed Open Shortest Path First (OSPF) routing protocol which is anticipated to also see deployment in FBNs. In an OSPF-routed network, the OSPF domain is scaled to large node counts through the creation of a two-level area hierarchy. To automate the design of an OSPF area hierarchy, the Dynamic Domain Optimization Agent (DDOA) has been developed as part of the U. S. Army CERDEC (Proactive Integrated Link Selection for Network Robustness) (PILSNER) program. Earlier work reported for DDOA demonstrated its ability to compute OSPF designs that simultaneously satisfy multiple design objectives and constraints. This paper presents novel new heuristics that have been implemented for DDOA to ensure that the multi-objective design challenge is satisfied for the lifetime of a mission even in the presence of time-varying network conditions such as dynamic node arrivals/departures and node mobility. Preliminary proof-of-concept results are reported herein to demonstrate that the DDOA extensions can yield OSPF area designs which are robust in dynamic network environments.