Planning Telecommunication Access Network (TAN) infrastructure is a real time problem that suffers from uncertainty and ambiguity. Fuzzy system is a discipline that proved its capability to deal with vague problems. Converting traditional data to suit fuzzy system calculations is called fuzzification process which is a crucial step affecting the whole system accuracy? Generating appropriate membership function for fuzzy variables is one of the most challenging issues in fuzzy systems design. This paper proposes a solution to generate membership function automatically. An integrated hybrid model (PSO-TE) is introduced which benefit the Particle Swarm Optimization (PSO) and the information theory measures (entropy and mutual information) as the fitness function to adjust particles (membership function parameters). The proposed algorithm is tested using realistic planning information to show its effectiveness and efficiency. During the comparison between the PSO-TE and the fuzzy C-mean algorithm, the proposed model proved its ability to produce stable membership function for the telecommunication data barriers.

No abstract