Municipal Wi-Fi Network Surveying and Testing: Part I
January 02, 2008
The first in a three-part series on how to effectively survey and test a municipal Wi-Fi deployment.
Most planners of municipal Wi-Fi implementations focus their energy on how to fund the network, but the technical sidenetwork design and installationmust also be sound or else the economic model cannot succeed. Delivering a sub-par network with slow or spotty service simply wont do.
The first step toward success is to draw up realistic technical specifications and designs early in the planning phasesand the second step is to meet them. Since Wi-Fi wasnt developed and tuned for the large-scale, congested situations, and applications of city implementations, this is especially crucial.
Since municipal deployments are relatively new, industry and network designers dont have many years of doing these installs under their belts. Adding to the challenge of muni network deployments is the tendency to use mesh networking, which is proprietarythe intelligence behind the scenes that enables these large and robust networks is uniquely developed by each vendor and not standardized.
The majority of the information and tips were gathered from our own muni wireless experiences assisting with surveying and testing of the city-wide networks in Miami Beach, FL, San Francisco, CA and the Denver, CO area.
Surveying and testing phases
Pre-installation survey. This should be done before any network design or installation work is performed. The findings given in such a report can help the municipality (or other network operator) and solution-providers deliver better plans. Major barriers or issues to implementing the network can be found and more accurate numbers and figures can be tallied. This initial surveying and evaluation may require two to four days on the street, and about the same amount of time in the office to compile a report. Tasks performed in this phase include:
- Scan and analyze the airwaves to discover existing networks and RF interference.
- Discover and analyze typical mounting assets, such as light and traffic light poles.
- Report on the environment, such as tall buildings and dense foliage or trees.
- Discover population concerns, such as densely (or lightly) populated areas.
- Note any possible priority areas for the networks, such as the business district or high-usage areas.
Pilot testing. To help choose between competing solution providers, pilot locations of a one-half- to one-square-mile area may be set up within representative areas of the city. This testing can include the following tasks performed for each pilot network:
- Review and analyze network designs and installation plans for suitability and efficiency.
- Verify that wireless coverage and signal quality is up to the defined levels.
- Test network performance, including throughput and latency.
- Verify the networks back-up power capabilities.
- Evaluate and test the technical aspects of any proposed solutions for city, public, or safety applications.
Pilot testing in the field can take a few days, with several days to create a report on the evaluation of the pilot networks.
Acceptance testing. This testing and surveying should be performed after a municipal network is completely installed but before the network is used and officially accepted by the city (or other network owner). This is to make sure the network is up to the defined specscomparable to post-site surveys of wireless networks within buildings. Acceptance testing includes similar tasks used for pilot testing, however more in-depth. Field testing could take two to three weeks, with an additional two weeks for a report to be generated.
These testing and surveying tasks should be performed by companies or organizations not involved, and with no stake, in the projectin other words, an independent consultant with wireless networking (and preferably muni wireless) experience. This reduces the chance of intentional false and exaggerated reports. If a city or other organization is spending millions to unwire a city, the design and final product should be verified by a third-party.
Gathering information and documents
The first step in performing a pre-installation survey of municipal Wi-Fi networks is for the independent tester to gather information and documents, including the following:
- Wireless standards/frequencies. In order to choose the correct testing and analyzer equipment, the particular wireless standards (802.11 a or g) or frequencies used by both the backhaul and client access must be identified.
- Coverage areas. The areas where Wi-Fi coverage is sought after must be identified, including any priority areas.
- Mounting asset information. Any existing information or maps of street and traffic lights, water towers, and any other possible mounting locations for the client access nodes and backhaul transceivers should be gathered. Information on ownership and available power supplies or sources at these locations is particularly helpful. A city engineer should be able to provide some of this information.
- Environment information. Talking with city staff, in addition to driving around for the initial surveying, should uncover any environmental concerns, such as tall buildings and dense foliage and trees.
Before the Pilot or Acceptance Testing is started, the independent tester needs to gather the following:
- Coverage specifics. Agreements and contracts should have defined the minimum amount of coverage required in the general coverage areas. For instance, this may be 95% outdoor and 85% indoor. Additionally, indoor coverage should be defined by a certain amount of feet into the establishment from the exterior walls, and how many levels or floors up. Because its nearly impossible to ensure 100% coverage (with acceptable performance) throughout the entire coverage area, these coverage definitions offer a real-world requirement.
- Performance requirements. The performance requirements should have also been stated within agreements and contracts, such as:
- Throughput. For example 1 Mbps download and 0.5 Mbps upload.
- Latency. For instance, 100 ms or less to support voice or video applications.
- SNR (Signal-Noise-Ratio). For example, a minimum of 25dB SNR to support voice applications.
- Back-up power. For instance, at least two hours of battery life once the main power source of a mesh node or backhaul link is lost.
Stay tunedthe next installment will discuss the testing and surveying tasks for each phase.
Eric Geier is computing and wireless networking author and IT professional. His latest books include 100 Things You Need to Know about Upgrading to Windows Vista (Que 2007) and Wi-Fi Hotspots: Setting up Public Wireless Internet Access (Cisco Press 2006).