The problem of planning local wireless network IEEE
802.11g consists of automatically positioning and setting up
wireless access points (APs) in order to provide access to the local
network with the desired coverage and the required quality of
service (QOS).In addition to the complexity of predicting the
Quality of Service (QoS) of a network from the variables of the
problem (positions, parameters and frequency of the APs), the
planning of WLAN networks faces several difficulties. In
particular, the location of APs and the allocation of frequencies.
There is no single model to solve the problem of designing wireless
local networks. Depending on the situations and the hypotheses
studied, different criteria can be considered and expressed in
terms of constraints to be observed or in terms of objectives to be
optimized. The first distinction is to separate the financial criteria
from the network quality criteria. The nature of these two criteria
being fundamentally different. Then there are a variety of service
quality criteria, but we can still group them into three main
categories: coverage criteria, interference criteria and capacity
criteria.. In this article, we will use an optimization method based
on an algorithm of stochastic optimization, which is also based on
the mechanisms of natural selection and of genetic. It is genetic
algorithm. Our goal consist of minimizing the total interaction
between the APs to perform the good choices when deploying a
network 802.11g in a way that gives users signal-to-interference
ratios (SIR) greater than the required threshold ß.