Muhammad Basit ShahabAbdul Aziz Bhatti2012-09-052012-09-052012Paper Presented at 35th International Conference on Telecommunication and Signal Processing, TSP, Prague, Czech Republic 3-4 July, 2012978-1-4673-1118-2https://escholar.umt.edu.pk/handle/123456789/576This paper proposes neural networks based graph coloring technique to assign Physical Cell Identities throughout the self organized 3GPP Long Term Evolution Networks. PCIs are allocated such that no two cells in the vicinity of each other or with a common neighbor get the same identity. Efficiency of proposed methodology resides in the fact that minimum number of identities is utilized in the network wise assignment. Simulations are performed on a very large scale network, where initially all the cells are without any PCIs assigned. Results of simulations are demonstrated to analyze the performance of the proposed techniqueenPhysical Cell IdentityNeural NetworksSelf Organized NetworksLong Term EvolutionOther