A series of encryption algorithms called A5 is used to secure mobile telephone communications, producing a pseudo-random sequence that will be exclusive OR (XORed) with the data flowing in the air interface in order to secure them. These algorithms are essentially composed of shift registers with linear feedback, controlled generally by a function or with another register in order to favor the randomness character of the keystream generated. Evolutionary algorithms are bioinspired calculation methods, whose principle is inspired by the theory of evolution, which consists in evolving a set of solutions to a problem given in order to find better results. This paper presents an improvement of the A5/1 algorithm by an evolutionary approach based on the use of particle swarm optimization algorithm (PSO) in order to limit some weaknesses and drawbacks found in the conventional A5/1 version, which have been cryptanalysed and several attacks have been published such as time memory trade off attacks and guess and determine attacks. Our technique does not alter the A5/1's architecture, but it does help to improve its shifting system by an evolutionary approach, which guarantees the quality of the keystream generated and makes it more complex and more secure.