Radio networks are important in rural Intelligent Transportation Systems (ITS) because they can provide communication from field elements back to a Transportation Management Center (TMC). The designs for these radio networks have specific goals they need to satisfy while attempting to minimize the overall cost of building and maintaining the network. For example, goals might be to connect a set of Road Weather Information System (RWIS) sites for weather monitoring or to provide connectivity for Vehicle Infrastructure Integration (VII) applications from an important section of roadway back to the TMC. Designing such networks in open urban areas is already well studied. However, in remote rural mountainous regions design is much harder due to rugged terrain, heavy vegetation, and sparse availability of wired backbone networks that can offer connections to the TMC. Manual evaluation of all possible designs is infeasible, and an exhaustive computational evaluation is infeasible because of the enormous number of candidate solutions. Instead, our research develops an Artificial Intelligence approach to design these networks, specifically using a technique called Ant-Colony Optimization. This approach can quickly find network designs that meet required goals while still minimizing the overall cost of building the network. In addition, our research enhances propagation analysis by using high-resolution 10-meter USGS elevation data instead of the currently accepted 100-meter data.
Abstract