Abstract

In a wireless network, the main objective of designers is to manage the available resources to maintain good quality of service for all users. As a result, engineers put extra effort in evaluating and improving the performance of wireless systems. Geometric modeling of wireless networks is drawing significant attention with regards to analytically evaluating the performance. In this dissertation, stochastic modeling of users is used to emulate their distribution in the wireless network. The wireless network is analyzed as a single-hop network where there is only one direct link connecting the base station to each user in the cell. Next, probability density function (PDF) of signal to interference ratio (SIR), signal to noise ratio (SNR) and signal to interference plus noise ratio (SINR) are evaluated. Given that outage probability is one of the important factors in studying the performance of a wireless network, outage probability is evaluated based on path loss, SIR, SNR and SINR. Simulation is performed in order to validate the analytical results. To show the application of the work done related to outage probability, a dual-hop wireless system is considered.

Wireless networks face some challenges, such as the presence of users in severe shadowing regions, or obstacles which may diminish the link quality. To overcome these limitations, dual-hop relaying system is considered. Dual-hop system is used as a cost effective solution in improving the performance of the wireless system. In such a solution, a number of relays is deployed over the cell to help transmission of signals from the base station to mobile stations. In the dual-hop system, a low-quality long-distance link is broken into two better quality links, achieving the required outage probability and enhancing the system performance. The number of deployed relays and their location over the studied area affect the performance of a dual-hop wireless system. Therefore, the required number of relays and their placement over the studied area are investigated, and a desired outage probability is achieved.

Uniform distribution of users cannot describe the real distribution for all cases. Gaussian distribution has malleability, as clustering tendency of users can be controlled in the network using the standard deviation of the distribution. Consequently, Gaussian distribution is considered for users in the studied area to evaluate PDF of SIR and the outage probability.