Realistic Network Algorithms

About

Market Overview: This algorithm, known as MUSKETEER, is used to simulate a realistic multiscale network model, enabling the formulation of algorithms for realistic network generation. Currently, the market for big data is over $6 billion and is expected to increase in the coming years. However, high-quality, big data is often not available because of economic, legal, or technological obstacles. For example, the development of security systems (including cybersecurity) requires testing across diverse threat scenarios and validation across network structures that are not yet known, in anticipation of the computer networks of the future. Realistic network models are vital for evaluation of different algorithms, scenarios, methods for situational awareness, anonymization, and anomaly/threat detection. Such models can be used for training, verification, and missing data completion. There is a need for multi-scale realistic network generation for national security, creation of power grids, determining optimal public health policies, and simulation of differing procedures within industries. Clemson University researchers have developed an algorithm that can be used for generating realistic networks to represent relationships between entities in complex systems. The network representation can then reveal the evolution, structure, and dynamics of those systems. The broad applications and industries that this multiscale network generation has the potential to transform are significant. Application High performance computing; Cyberinfrastructure and security; National security; Operations research Stage of Development Prototype and proof-of-concept demonstrated Advantages Generates a network while keeping the data "anonymous,” solving the problem of modeling confidential data that is not readily accessible Provides user control of the magnitude of the changes to the input data, allowing diversity of the output network Can be used to simulate dynamics of the data Technical Summary This Clemson University invention produces artificial networks that can realistically model a variety of network properties by using a specific algorithm, MUSKETEER. This strategy first creates a hierarchy of combined representations of the original network, and then reformulates the network generation problem at all levels of this hierarchy in order to take into account properties at multiple scales of the system. The network is then edited at any scale, depending on the desired variability in the ensemble of synthetic networks. The most important properties that distinguish multiscale network generation from all other existing strategies are the realism and diversity of a synthetic network that can obtained using this method. (2015-058)