Are the users likely to be highly mobile and therefore roam frequently? If so, over how much of the area are they likely to roam? The answers to the above questions might have an impact on where to place your Wireless Access Points or whether to employ a coverage-base WLAN or a capacity-based WLAN.
It is important to understand that each Access Point will connect to a wired LAN if the devices are to be part of the overall Local Area Network and not isolated as a standalone WLAN. Each Wireless Access Point constitutes a single collision domain, in other words a ‘shared environment’. Commonly today we see switched networks with devices connected to a single port, meaning they are in a single collision domain of their own and not sharing the medium with other devices. In half duplex setup an 11Mbps WLAN will provide a little over half the bandwidth (6Mbps) in theory, but in practice as little as 2 -3 Mbps. A 54 Mbps IEEE 802.11g Wireless Access Point will provide typically around 20 Mbps, but it has to be stressed that these figures are only a rough guide as environmental factors will affect the speed and throughput. Actual wireless speeds vary enormously from the theoretical maximum due to factors such as distance from the access point, physical obstructions that can absorb or reflect wireless signals and interference from other wireless networks. Remember also that a wireless network is a shared environment and the more devices sharing the same wireless medium will havee an impact on the speed of operation of individual devices.
The type of applications operating within the WLAN could have a major impact on its efficiency. For client devices running standard applications such as email and web services then somewhere in the order of a maximum of 25 devices per access point might be about right. If clients are running voice and multimedia applications then the above figure might drop to something like 6-12 devices per access point.
As the name suggests, these type of WLANs are designed to provide coverage of an area with minimum use of Access Points. Coverage Orientated WLANs are most commonly used where the applications being used are standard bursty type applications such as email and Web Services and so packet rates are low. Because of the nature of these applications, users can expect higher throughput with low congestion. Once this type of WLAN has been deployed, maintenance will normally be low provided a good site survey was conducted of the environment prior to deployment of the access points.
With a Capacity-Based WLAN, the objective is to try and maximize the throughput of individual devices within the WLAN because the client devices are expected to be running applications with high packet rates and / or VoIP and Multimedia which require a low latency environment. This type of application would suffer greatly if the WLAN was overcrowded with devices sharing the bandwidth. As previously mentioned, Capacity-Based WLANs might have as little as 6 devices or a maximum of 12. If the applications in use are critical then extensive testing prior to full deployment is essential.
This article was written by David Christie, MD at NSTUK Ltd, who specialise in the delivery of Instructor-Led training courses including Internet Protocol and Voice over IP training courses. Visit www.nstuk.com/training.html