Term of Award
Master of Science, Electrical Engineering
Document Type and Release Option
Thesis (open access)
Copyright Statement / License for Reuse
Digital Commons@Georgia Southern License
Department of Electrical and Computer Engineering
Committee Member 1
Committee Member 2
This work proposes a neural network architecture that was designed to predict and reverse engineer frequency hopping jamming systems. The neural network was initially optimized for use with a 12th order linear shift feedback register maximum length sequence utilizing a minimal polynomial as the characteristic polynomial. This neural network was then scaled to accommodate 7 different sequences, of orders 6 through 12. The neural network was trained for these sequences using training data that is 10 times the length of the sequence. This information is then used to generate a hopping sequence that reduces the jamming interference to 0 with as few as 4 jammer hopping samples. The model is also capable of determining if the jammer is utilizing a sequence that the model is trained for in as few as 25 jammer hopping samples.
Strickland, Charles, "Predicting Jamming Systems Frequency Hopping Sequences Using Artificial Neural Networks" (2023). Electronic Theses and Dissertations. 2580.
Research Data and Supplementary Material