Optimizing SOM for cell towers distribution

Abstract

The increasing in utilization of cell phones and number of cell phone users in rural and urban areas has requested the network specialists to grow their scope to cover all spots In wireless communication system, the cell towers distribution requires effort such as network planning and cost as it needs expensive equipment's like Global Positioning System (GPS). This paper proposes new technique to determine the minimum number with optimal distribution of cell towers that required for a specific area. The goal of the proposed technique is to guarantee the efficient quality of service that is includes adequate coverage and good call quality with minimum cost. The proposed technique depends on self-organizing map (SOM) neural network with effective modification and used suitable model of path loss propagation. It clusters the subscribers (nodes) by using optimized clustering technique to find minimum appropriate number of cell towers and optimizing their distribution. The optimizing process takes into its consideration achieving maximum coverage without interference and, minimum cost. Suitable path loss propagation model and received power equations are used in applied experiments. The results show improvement in coverage and reduction in required number of towers, resulting in reducing the cost.