Aerohive access point, Руководство по развертыванию

Страница: 18 / 221

Deployment Guide 17
P
LANNING
Budgeting Wi-Fi: The Chicken and Egg Problem
The hardware cost of a Wi-Fi solution is generally driven by the number of access points needed, and an Aerohive
network is no exception. Unfortunately, a traditional challenge of budgeting for Wi-Fi is that it is difficult to know
how many access points to plan for until you have deployed and measured them. There are methods of doing site
surveys before a deployment to answer these questions. While doing so is often worthwhile, you might just need a
general idea of what you would need to budget. Fortunately there are some simple guidelines that you can use to
figure out how many access points you need, including the number of access points per square foot, the number of
clients per access point, and the distance between access points.
• Access Points per Square Foot
The simplest and most common way of budgeting access points is per square foot. You simply take the
square footage of a building and divide it by some number. The most common metric used today is one
access point for every 4,000 to 5,000 square feet for standard offices with cubicles. However, if you need to
support voice applications, you need a higher concentration of access points. In this case, the
recommended formula is one access point for every 3,000 square feet, or even as low as one access point
for every 2,000 square feet. In the lightest weight convenience networks, it is possible to use fewer access
points, and densities as low as one access point for every 10,000 to 15,000 square feet can be successful.
Keep in mind that such a deployment often has dead spots and can only support very low client densities.
• Number of Clients for Each Access Point
Another way to determine the number of access points needed is to consider the number of clients you
want each access point to support. In a standard office environment, most enterprises plan to support an
average of 5 to 15 clients per access point. While the specifications of most access points state that they
can support up to about 120 clients, a significantly lower density is recommended to get an acceptable
throughput for standard office applications. If you expect to support voice over Wi-Fi in the enterprise,
account for those phones as well. With the addition of voice, the client density substantially increases,
requiring you to plan for an average of 5 to 10 data clients and 5 to 10 voice clients for each access point.
Remember that voice clients consume virtually zero bandwidth when they are not on a call. However, when
they are on a call, it is imperative that the traffic goes through.
• Distance Between Access Points
In a standard office environment, it is a good idea to ensure that access points are between 30 and 100 feet
from one another. A distance of 30 feet is needed in high-density environments and those with many walls
separating access points. A distance of 100 feet is sufficient in low-density areas with plenty of open space.
The three tips above can help determine how many access points to deploy in a given area. In general, the square
footage estimate provides the best budgeting estimate, with client estimations and the distance between access
points confirming the square footage calculations.
As with all rules, there are exceptions. If certain locations in the network have a higher density of clients, such as
conference rooms or lecture halls, a higher density of access points is required. Conversely if there are large open
areas with few active clients, fewer access points are sufficient.
Planning Tools
If following general guidelines does not provide enough confidence or if the deployment environment is particularly
challenging, you might consider using software planning tools like AirMagnet's Planner or Ekahau’s Site Survey (ESS).
Aerohive also includes a free planning tool with its HiveManager platform and online at
www.aerohive.com . Such
tools are useful in determining the placement of access points without performing a site survey.