Systems and methods for securing information
A system may be configured to prepare and use prediction models for predicting existence of fingerprints among encrypted traffic. Some embodiments may: obtain a machine learner configured to identify statistical differences between pseudo-randomness associated with encrypted user data and higher-entropy randomness associated with a set of other data; determine at least a portion of a path traversed by the encrypted user data in the network based on the identification; and secure the network based on the determination.
1. A computer-implemented method comprising:
generating a data set configured to have a first amount of entropy;
before cryptographic encryption of user data, combining the user data with the generated data set;
obtaining, from a network, (i) a first ciphertext resulting from cryptographic encryption of the combined data and (ii) a plurality of second ciphertexts located in the network;
determining, via a machine learner, an entropy level of the first ciphertext;
determining, via the machine learner, an entropy level of each of the plurality of second ciphertexts;
determining, via the machine learner, whether the entropy level of the first ciphertext is greater than the entropy level of each of the second ciphertexts;
responsive to the determination that the first ciphertext has the greater amount of entropy, identifying a portion of the network traversed by the first ciphertext; and
outputting, to a user remote from the network, the portion.
2. The computer-implemented method of claim 1 , wherein the determination is based on a classifier that implements at least one of a clustering analysis, support vector machine (SVM), linear discriminant analysis, time frequency pattern analysis, singular value decomposition (SVD), artificial neural network, deep neural network (DNN), recurrent neural network (RNN), convolutional neural network (CNN), hidden Markov model (HMM), and Bayesian network (BN).
3. The computer-implemented method of claim 1 , wherein the combination comprises concatenating the data set with the user data.
4. The computer-implemented method of claim 1 , wherein the user data is a portion of a message being communicated over the network.
5. The computer-implemented method of claim 1 , wherein the user data and the data set are separate files stored in at least one database.
6. The computer-implemented method of claim 1 , further comprising:
identifying a second portion of a second network through which the first ciphertext is detected to traverse, the second network being different from the network.
7. The computer-implemented method of claim 1 , wherein the identified portion forms part of a route established from a node of the network, to another node of a second, different network, and
wherein the user data is encrypted within the node.
8. The computer-implemented method of claim 1 , wherein the identified portion is an entire path starting from a node in the network at which the data set is combined with the user data, the entire path ending at another node in the network or in another network at which the user data is consumed.
9. The computer-implemented method of claim 1 , wherein the combination is performed on demand by the user.