Soft Computing and Dynamic Tunneling Paradigms in Network Security Attacks

Most cryptanalysis problems are characterised by a feature having one global optimum solution and several local minima, when formulated as an optimisation problem. Among the several heuristic techniques adopted to solve such optimisation problems, it is noticed that a genetic algorithm gets trapped at a local minimum solution. Dynamic tunneling technique has been recently used for obtaining global optimum solution for several problems. Motivated by this fact, we adopt a hybrid version of genetic algorithm and dynamic tunneling technique for the cryptanalysis of a popular stream cipher SEAL (software encryption algorithm). The concepts developed in the paper are useful in the cryptanalysis of block and stream ciphers.