Base stations represent the core contributor to the energy consumption of a mobilernnetwork. Since traffic load in mobile networks significantly fluctuates during a day andrnnight time, it is important to quantify the influence of these variations on the base stationrnenergy consumption. Therefore, this thesis proposes energy-saving technique to reducernenergy consumption at the base station level. One of the approaches is the base stationrnsleeping technique which takes advantage of varying traffic patterns on a daily or weeklyrnbasis and selectively switches some lightly loaded base stations to the sleep mode. rnVarious literatures have investigated the potential of energy-saving technique. However,rnthose studies have made a simplifying assumptions that leads to inaccuracy in overallrnresults. This thesis considers the effect of trafï¬c load on energy consumption instead ofrnmaking assumptions. Ethio telecom deployed UMTS network on two separate frequencyrnbands: 900MHz and 2100 MHz. While applying energy-saving techniques, data servicernand radio coverages provided by the remaining active base stations need to guaranteernthe requirement over the whole area at all times. To this end, optimization of the SINRrnvalue of the coverage area is carried out by considering received signals of the basernstations using genetic algorithms. As input data, real traffic statistics are used on a fullyrnoperated base station traffic load along with its power consumption. The tools utilized forrnanalysis include MATLAB, MapInfo and WinProp. rnPerformance results show that the proposed energy saving technique and algorithm canrnachieve energy savings of up to 25%, corresponding to the annual reduction of over 2,520rnKWh per-site. Considering the selected case study area of ethio telecom mobile network,rnthe use of energy-saving solutions allowed annual savings of more than 73,990.00 Birrrnper 12 square kilometers.