The Navy has identified the need for a platform to provide scalable mass data delivery for nodes in a large network. To address this need, Architecture Technology Corporation (ATCorp) proposes an innovative software technology called Optimal Network Clustering for Large Data Dissemination (ONet). The ONet approach will address the three major requirements for building a mass distribution platform: 1) An efficient clustering algorithm for partitioning the network into manageable components; 2) Mechanisms for detecting changes in the network topology in real time; and 3) Efficient methods for distributing data within the clusters.
Benefit: The end result of the ONet development effort will be an automated network cluster generator algorithm. The results of this SBIR will directly support future Navy requirements for using the Communications as a Service (CaaS) concept to support a larger number of nodes for command and control and sensor data messages. In commercial enterprises, ONet can provide a point solution technology for clustering within the Content Delivery Network (CDN) market.
Keywords: graph partitioning, graph partitioning, Clustering, multicast
