The rapid proliferation of smartphones over the last decades presents new challenges for military communication and commanders. Besides dealing with device and application vulnerabilities, the intentional and unintentional information transmitted by every device can reveal important information about troop locations and movements. This research seeks to address some of these challenges by using decoys to increase operational security. Using Wi-Fi data from the University of Colorado Boulder as a surrogate for cellular data and using models generated with Nonnegative Matrix Factorization (NMF), three Wi-Fi decoys are deployed at three locations across campus to simulate network activity from a military scenario. By populating alternate locations with decoy devices to mimic real activity, this research shows how NMF can be used to create models that can effectively be translated into real network activity to obfuscate real operations. This enhances operational security by dispersing network signatures and confusing adversarial detection.
