Android Open Source

  NN ware Interesting topic. Neural networks in information transformation. In encryption. Encryption algorithm generators are interesting. Synchronous. Something like this: each message is encrypted with a new algorithm (not an algorithm with new parameters, but in a completely different way). That is, the application must create thousands of new algorithms... But the question immediately arises of how to evaluate their reliability Entropy Tech  

  Synchronous random number generator Synchronous generators can be used in recursive encryption algorithms. When a received decrypted message or part of a message is "mixed" with some combination of bytes and used as the source of a newly generated array of numbers for encryption. You can completely dispense with books, magazines or news reports to obtain the same initial values to generate. Moreover, there is no need to store the decoded received message at all. You can immediately generate a new array of numbers based on it and store this array. But these are all rather old technical ideas. Which use a stable, constant algorithm. Imagine an application in which each message is encoded with a different algorithm (plus all sorts of permutations, packaging, compression, replacement), not similar to the previous one. It's like encrypting the first message using XOR, the second message using RSA, and so on. And not a single repetition. Those. Each message must be h

  Synchronous random number generator A synchronous random number generator performs the same functions as some thick book or reference book for the sender and recipient of coded messages. Only unlike the book, the encryption process can be automated. It is not at all difficult to write an application that will implement, for example, the Richard Sorge or Vernam cipher. And even significantly complicate them. But another problem will remain. The problem of secret data transmission. It is impossible to secretly send data directly to the server and even implant it into a JPG image. And a request like http://cnn.com?453872654076958341023078659845   will look suspicious to control systems. However, there is more to come from this technology. In this case, only publicly familiar technologies are used.     Sieve of Eratosthenes

  Digital spices   If you place the application in the public domain, you can decode it and understand the algorithm of operation. But the result of the application (an array of numbers) very much depends on the input source text. However, you can add some algorithms and input data to the generator, which will further change the resulting array. How spices change the taste of soup. Because the sender will have to enter additional data, this will increase the complexity of use and increase labor intensity. But it will also improve data protection. There are so many to choose from. If your application copies (for the sender and recipient) are made under an individual contract (custom), then data protection will be significantly higher. Java APP      

  Synchronous random number generator Fixed some bugs and improved entropy a bit. It may seem that 60 bytes of numbers at a time is not enough. More can be done. But you can use two or three phrases. And this improves subsequent encryption. Moreover, there are many options. https://drive.google.com/file/d/1in6sX45yDGty7NyDGCskZhJYJ4wo8V_T/view?usp=sharing   Android Java  

  Bulletin   Errors have been corrected in Application Czech Entropy . Now the application normally generates several keys (arrays of random numbers) in a row from different texts. Text can be added via Copy/Paste. We still need to fix some little things in the User Interface. Then you can improve the generation mechanism. Synchronous random number generator  

  Entropy Application The Android application Czech Entropy can generate the same set of 60 “random” numbers if the same text is entered as the initial data and the same smartphones with the same versions (build) of the operating system are used on the sender and recipient sides. In the future (not quickly) a version will be presented without dependence on the platform, with dynamic algorithms for generating arrays of “random” numbers (formulaless and non-reversible).   Android APP Entropy Generation

  Android APP Fixed some errors in the interface, slightly redesigned the user interface (it will remain minimalistic). The application currently generates a set of numbers once (you need to exit it and run it again to make the next code). This will be fixed shortly.   Therefore, version 0.0 for now Then QR code generation will be added (possibly). After the errors are corrected, the algorithm for generating “random” numbers will be upgraded - new transformation layers will be added. Czech Vernam Entropy

  Application     Synchronous random number generator for Android smartphones. Uses an algorithm for expanding the original data. Android 10 Tested   Download CzechEntropy APK Czech Entropy P.S. I started studying neural networks to improve the quality of generating “random” numbers. There are interesting ideas. We need to make prototypes and test them.            

Android APP Version 0.0 of the application CzechEntropy is available at  Download CzechEntropy.APK       When installing an application, you must have permission to work with memory (the application creates temporary files). The application generates an array of 60 "random" numbers based on the entered text (in the top field of the application). The text should be longer than 100-120 characters. If you use several identical smartphones with the same version of Android (build), then the same 60 “random” numbers will be generated for the same entered text. This version of the application uses one type of formless, irreversible transformation. Czech Entropy In the future, the quality of the generated synchronous sets of “random” numbers will be improved by adding transformations using other technologies, and a processing option that does not depend on the hardware and software version will be added. Czech Entropy                         

  Synchronous entropy generators One, even irreversible transformation of the entered text (in fact, a long password), does not guarantee that some dependence will not be found in the sequence of “random numbers”. The first prototype of the application only demonstrates the technical feasibility. Using several consecutive non-reversible transformations will greatly reduce the likelihood of calculating the dependence of “random” numbers (increasing the cracking time to a commercially acceptable level). Such a set of numbers can already be effectively used for the second encryption of already encoded text. The next level of reliability can be achieved by using a software module that generates a variety of non-reversible conversion algorithms while the application is running. Then the only effective way to hack would be to obtain an instance of the application to understand the conversion algorithm used in each specific case. Those. These are already variants of the application create

  Android APP Beta version of the application. It is necessary to enter the text in English without punctuation marks longer than 100 characters and press the "GENERATE" button. After a while, a set of "random" numbers will be shown as INTEGERS and HEX. They can be copied to the clipboard for use, for example, in XOR encryption. The length of the set of numbers is 60 bytes. The given text is converted into a set of numbers using formulaless technology and is not reversed. So far, the minimally working algorithm has been used. If the sender and receiver select and enter the same text (for example, the same news item published on CNN at 12 o'clock), they will receive the same set of "random" numbers. Attention. The application DOES NOT encrypt data. The application only generates a synchronous set of numbers. It can be seen that while the quality of the set of numbers is not high. Algorithms for choosing "random" numbers can be significa

  S ynchronous random number generator The beta version of the synchronous "random" number generator is almost ready. When entering text longer than 100 characters, it generates the same sets of numbers for the message sender and message recipient. Does not use formulas to generate "random" numbers. The mechanism for choosing numbers is the simplest so far. Further, the mechanism will be regularly improved using other non-reversible generation mechanisms. It will be free to download Czech entropy    

Encoding "Intuitive" coding will have to wait. About synchronous generators of "random" numbers. It is possible to create generators of "random" numbers that, without information exchange (secretly), will create the same set of random numbers. You can make it so that it will be really random numbers, for example, based on the thermal noise of the transistor. It's simple. Somewhere there is a hardware server that creates a random digital stream. Two or more application instances listen to a stream and wait for a given combination of bytes to appear. After that, according to some algorithm, as many bytes are selected from the stream as needed to encode information. During the day, you can collect several keys (regulated by the number of bytes to wait). On the receiving side, the same is done and the received message is decremented by one of the found keys. The vulnerability of the method lies in the possibility of identifying users connected to

  Crypto Tech Modern PCs and even smartphones have enough performance to use neural networks to create messages with hidden information. One of the algorithms The system generates the "natural" text of the announcement about the sale, for example, technical literature. The text of the message is formed by the neural network so that the required positions contain the characters that make up the encoded message. The message is then posted on the bulletin board. If you publish PDF files in online libraries, then this method of transferring information is even easier. Another way to use the technology is to design advertising posters, in which the data is included in the background created for the advertised item. But these are the simplest options for which countermeasures can be provided. It is interesting to use global online services or (and) neural networks for "intuitive" coding. Czech Pepper