Wireless Networks Topology Optimization – Decision for Directional Antennas

Abstract

Wireless networks have become an important part of the communication infrastructure, which providing mobile customers with access to a variety of services. The process of developing the wireless network topology is associated with the solution of multi-criteria optimization problems. This is a computationally complex procedure, especially for directional antennas, because three parameters must be defined for each network node: antenna’ coordinates and radiation angle. Directional antennas in network nodes provide a longer signal transmission range than Omni-directional antennas and reduce interference. The paper estimates the possibility of using a parallel implementation of the brute force algorithm for building an optimal topology of a wireless network with directional antennas at its nodes. The criterion of the optimal solution is the number of residents, which entering the coverage area and having access to network services. The presenting results of computational experiments illustrate the scalability and performance of the parallel implementation of the brute force algorithm. The evaluation of parallel application efficiency by means of Locks and Waits analysis is produced in Intel Parallel Studio environment. As a computing platform, we used in experiments a workstation with two 6-core processors E5-2630 v2 (2.6 GHz), supporting the execution of multithreaded applications.