Virtual Leashing: Creating a computational foundation for software protection

We introduce Virtual Leashing,1 a new technique for software protection and control. The leashing process removes small fragments of code, pervasive throughout the application, and places them on a secure server. The secure server provides the missing functionality, but never the missing code. Reverse engineering the missing code, even with full tracing of the program's execution and its communication with the server, is computationally hard. Moreover, the server provides the missing functionality asynchronously: the application's performance is independent (within reason) of the secure server's speed. For example, the server might reside on a slow inexpensive chip or a remote Internet server. Leashing makes only modest demands on communication bandwidth, space, and computation.