Average Likelihood Methods for Code Division Multiple Access (CDMA)
Abstract
This final report summarizes the most significant findings of the work done under the in-house effort titled "Optimal Methods for Modulation Classification" on average likelihood methods. The development began with the exact calculation of the likelihood function using symbolic algebra software. After investigating its properties, it was possible to develop formulas for more general cases. The solution of the likelihood function depends on a joint probability of the spreading code coefficients/elements. Simplification of the likelihood was achieved using the invariance properties of the Total Squared Correlation to permutations. The average likelihood can be expressed as sum of a product of hyperbolic cosine functions. Most these terms vanish after adjusting precision parameter from the joint probability of the code matrix. For a full loaded CDMA signal, the average likelihood depends exclusively on feature vectors or the absolute value of the sum of the columns of the matrix with the lowest Total Square Correlation. The ROC curves were plotted for characterizing the performance of the classifier.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 2014
- Accession Number
- ADA603961
Entities
People
- Alfredo V. Irizarry
Organizations
- Air Force Research Laboratory