Following growth in infrastructure, number of subscribers and availability of smart devices andrnapplications, the aggregate cellular data traffic in Addis Ababa cityâ€™s cellular network isrnincreasing exponentially. Moreover, traffic growth follows non-uniform distribution both in spacernand time. To accommodate this non-uniformly growing data traffic, ethio telecom, for now, thernsole telecom service provider in the city, deploy single-layer homogeneous macro base stationsrn(MBSs). Macro-cell densification has been used to increase capacity of the radio access networkrn(RAN). However, excess densification increases the RAN energy consumption, which is becomingrna concern for cellular network operators like ethio telecom. rnDeploying small cells overlaid with macro BSs, named as the heterogeneous network (HetNet), isrnan energy-efficient (EE) approach capable of meeting the high capacity demand and also keepsrnnetwork deployment costs low. Many studies have analyzed the HetNet planning and deploymentrnscenario. However, user usage scenarios and their mobility pattern based on realistic data are notrnconsidered for the selection of initial small cell candidate locations. Their results differ from onernanother depending on the environment, cell size, data set, and technology, or the methodology onrnwhich the research is made. rnThis research investigates a genetic algorithm (GA) based multi-objective optimization based onrnsystem capacity and EE maximization to provide a set of optimal solutions for HetNet selection.rnIn doing so, based on a dataset collected from Addis Ababa cellular network, existing macro BSsrndata traffic, user usage scenarios, and spatial data traffic demand distribution are generated tornidentify hotspot areas, and are given as input parameters for the GA for optimal small cellrnselection. Then, layered planning and deployment is carried out in an interference-limited LongTermrnEvolutionrn(LTE)rnnetworkrnbased onrntherntargetrnrequirement.rnFinally,rnperformance gainrnofrnthernrnoptimizedrnlayeredrnapproachrnisrnevaluatedrnwithrnsystemrncapacityrnandrnEErnasrnperformancernmetricsrnandrnrncomparedrnrnwith a uniform topology which is resulted from unplanned small cell deploymentrnthrough network simulation tool. The simulation results mainly show that both EE (up to 22%)rnand up to 16% capacity gain with cell edge performance gain (52%) of the target area are,rnimproved by the deployment of optimized small cells, over the uniform (unplanned) deployment.