NET 211 - Wireless Networking

Chapter 8, Conducting a Site Survey

Objectives:

This lesson discusses site surveys, tools used to do them, and procedures to follow when doing them. Objectives important to this lesson:

  1. What is a site survey?
  2. Tools for site surveys
  3. Procedures
Concepts:
Chapter 8

The author introduces the topics of this chapter by remarking that the average homeowner does not expect as much from wireless coverage at home as that same homeowner or his/her boss may expect in a workplace. Wireless LAN coverage in all parts of a building is a nice thing to have in the home, but it is more of a necessity in an office environment, a school, a hospital, and anywhere people may be when they need to access data on a mobile device.

This chapter addresses the idea of conducting a coverage survey of a site, with the idea in mind that we would like to improve the wireless LAN coverage where users need it to be better. The author introduces a related concept that we should consider whenever possible. In addition to having adequate coverage, usable bandwidth, and no dead spots, we should try to keep our wireless LAN inside our boundaries. It is nice thing for a coffee shop to extend their complementary wireless across their parking lot. It is a different matter when that kind of extended coverage exposes our network to eavesdropping and compromise.

The text lists some of the factors that can affect our wishes to provide ideal coverage for a wireless LAN:

  • devices that emit RF on 2.4 or 5 GHz - cordless phones, fluorescent lights, and microwave ovens are common sources of interference
  • walls and structures - steel beams, concrete pillars, file cabinets, and heavy machinery can obstruct signals
  • doors - fire doors in particular can obstruct signals
  • rearrangements - changes like adding cubicles to an office, moving machinery in a factory, and other operational changes can make a working wireless LAN stop working
  • interference from other WLANs - in urban settings in particular, there can be intrusions from WLANs belonging to other entities

The text presents a list of goals of a site survey on page 280. Some are obvious. We should take notice of some of the less obvious ones:

  • Determine the best location for each Access Point
  • Establish good coverage and operation
  • Find any unauthorized APs
  • Map nearby networks that intrude on ours
  • Minimize interference to and from our network
  • Secure the network

The text also tells us about some events that should cause us to do a site survey:

  • changes to a building
  • changes to a network
  • changes to network needs
  • changes in personnel

The text remarks that some network products can manage themselves to a degree, increasing signal strength, for example, when  reception changes in the environment. This solution has obvious limitations, since transmission strength is limited by FCC regulations. Sometimes there are problems that have to be discovered in order to solve them. A survey is the best first move.

Types of surveys:

  • manual site survey - a person carries a wireless client around the site, taking measurements
    • passive manual survey - packets are not sent from the client; measurements are take on signal strength, noise level, and signal to noise ratio
    • active manual survey - packets are sent with the client; the measurements above are taken along with bandwidth, packet delay and data on APs, such as MAC addresses and SSIDs
  • predictive site survey - a simulation is loaded with data about the site to predict what problems and performance it should have

The text discusses tools to use in a manual survey or predictive survey. As explained above, a predictive survey requires  floor plans and structural information about your site. The actual file formats needed will vary with the software you are using.

A manual survey can be done with a number of tools. The text discusses several types in general terms of their functions:

  • wireless device tools - The text recommends that we start a manual survey at the location of each AP, and the power settings used on them, if yours have such settings. A technician should carry a wireless device, like a laptop, tablet, or smart phone that can access the AP and measure signal strengths at various locations and distances from it. The text warns that standard tools can vary from one device to another, so the same device and tool should be used, or choose a better tool from the next discussion.
  • dedicated site analyzer - The text discusses features to expect in a site analyzer tool. An example is the Acrylic Wi-Fi Professional tool, which comes in a free version for schools, and in a trial version/licensed version for individuals. This is a 2015 article from Network World about free tools to stumble and survey wireless networks. Acrylic is one of their choices. Read the article, and decide if one of the products would be useful to you.
  • spectrum analyzer - This analyzer measures the frequency, voltage, period, and shape of the many waves used in wireless. Because we use lots of frequencies, lots of power levels, and lots of frame methods, this is more complicated than it is on a wired network. The text explains that we should run an analyzer to detect interference with the operation of APs, and move them or the interfering devices.


    The technician in the video above is very precise and correct, but he may be a little old fashioned for some. Take a look at this video from a more media friendly technician.



  • protocol analyzer - This is like a packet analyzer, such as Wireshark, but it is also more complicated in wireless because of the number of frequencies, power levels, and frame methods. Like wired network sniffers, these analyzers can be used to monitor traffic on a network, to look for problems, and to watch for particular types and sources of traffic.
  • documentation tools - Some site survey tools do not create documentation, so a tool of this sort can be a useful addition, adding the ability to document the findings of the survey.
  • voice over WiFi (VoWiFi) - As the text explains, voice transmission, whether they are done over a wire or wirelessly, require better performance from a network because users do not want to wait for a stream to be reassembled, to wait for the channel to be turned around, or miss out on part of a conversation due to dropped packets. Some common problems are discussed on page 293. The text recommends that measurements for these problems should be done from a WiFi handset, because only such a device actually experiences the effects discussed below.

    It also recommends that for WiFi installations that support VoWiFi, the cells should overlap, as they do for commercial cell phone service. A useful discussion of common problems can be found on this web site, and at the site that provided the illustration on the right. Note the fact that cells overlap for roaming, but that no cell touching or overlapping other cells is using the same channel as those other cells.

    The left side of the illustration shows the three non-overlapping channels in the 2.4 GHz band, and the right side shows APs using two non-overlapping channels each in the 5 GHz band. Each individual color stands for a particular channel or channel pair.

    These are common problems with packets in a VoWiFi system:
    • packet loss - In a normal data stream, lost packets are requested and replaced, and the end result is that it takes longer for a file to successfully upload or download. The loss of packets in voice transmissions can lead to a message that can't be understood. The standard for packet loss in VoWiFi is "less than one percent".
    • delay - In this case, packets are not lost, some simply arrive later than they should, which often causes an echo effect or other distortion of the signal. The standard for packet delay in VoWiFi is "less than 50 milliseconds".
    • jitter - This is the quality of a conversation or transmission that has too much dead space between packets. The standard for jitter in VoWiFi is "less than 5 milliseconds". The author does not mention it, but research has been done that shows that users on a slow system slow down themselves, adding to the slowdown of the entire process.

The last major part of the chapter discusses actual procedures for performing a site survey. Some are the same as doing any IT project.

  • business requirements - Whether you are doing a survey for an existing wireless installation, a new one, or one that someone wishes to change, it is best to start with the reasons for the installation to exist.
    • Do the users need coverage throughout an entire site, or do they need it in specific locations within the site? This will affect the placement and number of APs required.
    • What are the constraints from other equipment? Do we need to support all protocols due to the use of legacy devices, or can we support only the newer ones?
    • What bandwidth is required by which users? Again, will they need it all over the site, or only in specific locations? If they work only in particular areas, then less capable APs may do in the other areas.
    • How will you gather the information about the business needs? Personal interviews are best, but the most time consuming. Group interviews, such as with committees, can generate information that individuals don't realize they know, but they make it difficult to pursue topics that apply only to some of the attendees. Larger groups tend to cause quiet people to stay quiet. Questionnaires typically have a low participation rate, and they are impersonal, so you are less likely to discover things you did not ask about. Shadowing staff can be very helpful, because you can observe them doing things they do not realize they do.
  • security requirements - Decisions need to be made about required security that will be used on the network. Assume that there will be sensitive data passed on the network, so security is a real concern. The text addresses security in the next two chapters.
  • site documentation - The previous items are mostly logical. You need to consider the physical constraints of the site. Where are there dead zones for wireless coverage? Do we want to change that? If we are relocating APs or other devices, is there power for them in the new locations? Can you move freely around the site to take measurements, or must you have an escort in sensitive areas, such as accounting departments or areas where particularly sensitive data is accessed?
  • existing network characteristics - You need to know what the network is currently used for, what it will be used for, and how many users of various types it will need to support. If the network is at its maximum capacity now, improvements need to be made before you can expect additional services to work. Some facts to gather are shown in a table on page 298. Notice that it should document the wired network characteristics as well as wireless ones.
  • technical requirements - The author presents two other tables on page 298. They summarize the questions and answers that you should have covered in the previous steps.
  • placement of APs - The initial placement of an AP should be considered temporary, because your plan may run into problems you did not anticipate. Placement and configuration may need to take place over time as you encounter actual usage.
  • identify interference and obstruction - As noted previously, when you find that there is interference with an AP, you should consider moving the AP, changing its channel, or changing the orientation of its antenna. Some obstacles to RF transmissions are listed on page 300. Note that you should not expect obstruction from the items at the top of the chart, such as ordinary glass windows, cinder blocks, plaster, and wood. You should expect more obstruction from items as you go down the chart, especially items that contain metal or wire mesh.
  • outdoor surveys - Be aware that outdoor conditions change more frequently than indoor conditions. The seasons change, and leaves grow and fall. Weather can obstruct or affect signals, as can construction between APs and mobile units.

This last link is to a nice guide to site surveys by a vendor. Take a look at it to reinforce the points from this chapter.