Abstract—Frequency planning in cellular networks is an extremely onerous task. Manual allocation of frequencies for different cells in a cellular network is even more difficult due to overshooting cells in the network and limiting number of frequencies available in the spectrum; thus, leading to inefficiency in cellular radio systems. The substantial deployment and high growth of cellular networks give rise to the requirement of automatic frequency planning (AFP) tools. Such tools ensure optimum performance and coverage of wireless networks. These tools are based on different evolutionary algorithms which play a key role in the optimization of cellular networks and are frequently used by different cellular service providers for better network frequency planning. In this paper, AFP is implemented in real cellular network of different major cities of Pakistan using Automatic Frequency Optimization System (AFOS) tool. An instant decrease in call drop rate was observed ranging from 14% to 35% after the implementation of frequency plan from AFOS tool using baseband frequency hopping technique. Keywords-automatic freequency planning (AFP); automatic frequency optimization system (AFOS) I. INTRODUCTION In the last couple of years there has been significant growth in wireless services. The service providers have to adapt themselves with this rapid growth while ensuring optimum performance and coverage in existing and new networks. The global system for mobile communication is a digital cellular technology for providing voice and data services to mobile phone users. It is one of the most successful cell phone technologies. The RF engineers spend a lot of time configuring and planning the network. This practice needs to be conducted on regular basis to meet the demands of increased network traffic to achieve satisfactory network performance. As each GSM operator is allocated a limited frequency band having small number of frequencies which need to be assigned to thousands of transceivers (TRXs) of different cells of the network. Therefore there is a reuse of frequencies and automatic frequency plan generation is regarded highly important for present GSM operators for network optimization. In GSM, high interference is observed if same or adjacent frequencies are used in neighboring cells causing handover failures and call drops. Interference has always remained a limiting factor in radio communication systems despite the ongoing enhancement in equipment design. Initially, due to low demands of wireless services, manual frequency planning approach could be employed with confidence. However, with passage of time, tremendous growth in traffic has resulted in an increase in the cell sites density to meet the demands of traffic increase. This has consequently led to high network complexity making it extremely hard to manually assign frequencies to cells for optimum network performance. Therefore automatic frequency planning (AFP) methods have replaced manual frequency planning techniques. These methods use strong optimization algorithm and enhanced techniques for assigning of cells frequencies resulting in near optimum network performance and maximizing network capacity and coverage. Much research has been done in automatic frequency planning such as evolutionary algorithms for AFP [1]-[5], minimizing interference in GSM networks [2], optimizing a nonlinear, multivariate interference function [3], frequency planning using inter-cell dependency matrix (ICDM) [4], using meta heuristics and grid computing for frequency panning [6], etc. Recent developments include the use of Genetic Algorithm for frequency planning [7], use of fuzzy logic for frequency planning and its comparison with Genetic Algorithm based method [8]. In this paper, we have implemented AFP using Automatic Frequency Optimization System (AFOS) Tool in real network on major cities (Peshawar, Islamabad and Rawalpindi). After implementation, various Key Performance Indicators (KPIs) were examined to determine the effect of AFP on network performance in those cities. AFP using AFOS tool has resulted in enhancing network quality and performance as improvement in Call Drop Rate (CDR) has been observed in each city ranging from 14% to 35%. The paper is organized as follows: in section 2, AFOS tool description is shared. In section 3, AFP process using AFOS tool is mentioned. Section 4 covers the performance results and comparison of each city and section 5 provides conclusion. II. AFOS TOOL DESCRIPTION AFOS is an effective, accurate and interface friendly automatic frequency optimization software tool which supports Alcatel Lucent, Huawei and Zhongxing Telecomm (ZTE) vendors. AFOS improves the voice quality, reduce Performance Evaluation of Automatic Frequency Planning Using Automatic Frequency Optimization System Tool M. M. Haider*, J. Yousaf**, N. Qureshi*** * , ***Center for Advanced Studies in Engineering, Islamabad, Pakistan **SungKyunKwan University, Suwon, South Korea *muhammad.mustafa1@ymail.com, **jawad.yousaf@yahoo.com, ***nadeemkureshi@gmail.com,