Greetings and WELCOME to
Ed Rice's System Engineering Home Page
Electrical Engineer Ed Rice, USAF Retired
23 Years of USAF experience in Weapon Delivery Analysis, Signal Analysis, Research & Development for SIGINT Systems

Technical Notes On
Classifying Arbitrary Bit Streams Using Shannon Entropy, Statistical Distributions and
Euclidean Distance Vector Fitting Technique
15 September 1998
Edward G. Rice, Consultant
Research Associates of Syracuse Inc.

Presents the 2 man-month feasibility analysis obtained from evaluating 14 statistical parameters of 13,762 arbitrary bit stream samples. The results demonstrate the feasibility of accurately discriminating and classifying arbitrary bit streams with the goodness of fit values generated by a Euclidean distance algorithm.

Abstract & TOC

1.0 Classifying Arbitrary Bit Streams Introduction

2.0 Overview of Arbitrary Bit Streams

3.0 A Bit Stream's Statistical Feature Vector

4.0 Maximum Liklihood Subspace Fitting

5.0 Euclidean Distance Vector Fitting Technique

6.0 Classifying Arbitrary Bit Streams Conclusion

7.0 Recommendations for Further Investigation and Development

Appendix A

Appendix B

Appendix C

Briefing Slides Currently Available

• Bit Stream Classifier Intorduction

• Bit Stream Classifier Feature Vector

• Bit Stream Classifier Interum Data

• Bit Stream Classifier Parameters