The future Square Kilometre Array (SKA) radio telescope is an interferometer array that will use a variety of collector types, including approximately 2500 dishes distributed with separations up to a few thousand kilometres, and about 250 aperture array (AA) stations located within 200 km of the core. The data rates associated with each individual collector are vast: around 10 GBytes/s for each dish and 2 TBytes/s for an AA station. As each of these must be connected directly to a central correlator, designing a cost-effective cabling and trenching infrastructure presents a great engineering challenge. In this paper we discuss approaches to performing this optimisation. In graph theory, the concept of a minimum spanning tree (MST) is equivalent to finding the minimum total trench length joining a set of n arbitrary points in the plane. We have developed a set of algorithms which optimise the infrastructure of any given telescope layout iteratively, taking into consideration not only trenching but also cabling and jointing costs as well. Solutions for few example configurations of telescope layout are presented. We have found that these solutions depend significantly on the collectors’ output data rates. When compared to a “traditional” MST-based approach which minimises trenching costs only, our algorithms can further reduce total costs by up to 15–20%. This can influence greatly the SKA infrastructure related costs.
展开▼
