This section presents some specific aspects related to cellular network planning for mobile communications services (GSM, UMTS, PMR, TETRA, etc.).

The planning process is based on common steps specified in the section access network planning, although some aspects are presented here to better adapt to the type of network.

Then, steps of access network planning process are listed, noting in each one when necessary, the points to which the user must pay special attention to the planning of cellular networks for mobile communications.

NEEDS ANALYSIS - CONFIGURATION

•Cartography. Depending on the service area for the network on project, this type of studies may require the use of rural, urban or mixed cartography.

When the stations are located in rural areas, terrain models of 100 and 25 meter resolution are often used. For simulating the effect of multipath losses in the coverage of these stations over urban areas, it is usually used a morphography layer in which the user can include additional losses associated with each urban area.

If the stations are located in an urban area, is advisable to use urban cartography for radio planning. It is adequate a terrain model of between 1 and 2 meter resolution. In this case it would not be added any additional losses to propagation.

Finally, it is possible to use multiple altimetry layers to perform mixed calculations, where the station is located near an urban center and the user wants to analyze the propagation toward its interior.

•Calculation methods. Different mobile communication technologies work in very different frequency bands and require many customisations when defining a calculation method applicable to their planning. Here are some proposals for specific environments and thecnologies:

oRural environment. Planning in such environments is usually performed using deterministic methods such as ITU-R Rec. 526 or Deygout method, provided that cartography of rural area where the network deployment area is available. In these cases it is necessary to consider that the simulation results assure levels exceeded in 50% of time and locations, so that, in order to ensure higher percentages it will be necessary to use the "fading margin", which is configurable in parameters of the calculation method.

Together with these methods, and if morphographic layers are available, it is common to introduce additional losses associated with the terrain, especially to see the effects of multipath in urban environments.

In cases where no precise morphographic layers from the service area are available, it is possible to use empirical methods such as Okumura-Hata particularized for rural settings.

oUrban environment. Planning for urban or mixed areas (rural-urban) can be done by empirical methods (the Okumura-Hata method is an adequate approximation of the signal rangel into a radius from the base station) or deterministic if precise enough urban cartography is available (1 or 2 meters resolution). As a deterministic method, can be used Deygout method to obtain a simulation of propagation in city taking into account diffractions on the buildings roofs.

Whether using empirical methods or detailed urban cartography, it is not advisable to use clutter because a redundant correction over propagation estimates is being introduced.

oUMTS in urban environment. The Xia method implemented in Xirio is a semi-empirical method specifically designed for the estimation of signal propagation in urban mobile environments in the 2 GHz band and is used for planning of UMTS networks provided that urban cartography is available sufficiently accurate.

Quality objectives for these networks require to ensure a minimum coverage for a certain percentage of time and locations. These minimum percentages are translated into signal margin over the estimated average value in simulations, which is obtained from a statistical distribution that simulates the effects of fading of the received signal by the mobile terminal in a given propagation environment. Therefore, in order to ensure higher percentages it will be necessary to use the "fading margin" configurable in parameters of the calculation method.

PARAMETERIZING STATIONS

•Link budget. Before the network planning, the user must perform their link balance calculation and determine the critical path between uplink and downlink. Because path losses are identical in both directions, Xirio simulates only the downlink. However, if the user notices that the uplink is the critical path, he can simulate it by simply exchanging the power and sensitivity parameters of the Base Station and the Mobile Station in the transmitter objects and the receiver radio parameters.

•Stations and sectors. Mobile communications networks are typically composed of a distribution of stations in the form of cellular mesh. Each station can be composed of one or more sectors. The way to define them in Xirio is selecting a site as station location and as many transmitters as sectors.

NETWORK PLANNING

•Best server by signal. The user has to give the criteria that Xirio will use to determine the desired transmitter at each location. This will be the sector to which the mobile would be linked at every location in normal conditions. The usual approach is to work with a better server by signal, that is, the mobile would be linked to the transmitter which provides the highest signal level.