WLAN Design and Validation Project

Project	WLAN Design and Validation
Environment	Academic building, one floor
Team	2 students
Final design	54 proposed APs
Primary band	5 GHz
Channel width	40 MHz
Channels	40, 48, 153, 161
Coverage target	≈ −65 dBm
SNR target	≥ 20 dB
Tools	ExtremeCloud IQ, VisiWave Site Survey, Excel, Word

Overview

Two-person WLAN design and validation project completed for a Wireless Networking Fundamentals course at SAIT Polytechnic, based on a course-provided RFP for an academic building. We were assigned the second level (M200) of the Stan Grad Centre — the MB, MC and MD wings around the Irene Lewis Atrium, covering classrooms, computer labs, e-learning rooms, offices, study areas and the atrium walkways.

We ran a walkthrough site survey, built a predictive WLAN design in ExtremeCloud IQ, worked out capacity and channel planning, put together an implementation plan and Bill of Materials, and passively surveyed the existing WLAN for validation.

Note: Original course materials, SAIT floor plans, and screenshots containing account information are not included in this public repository. The diagrams below were recreated to illustrate the actual technical decisions and results from the project.

Project Outcome

The project produced a complete WLAN design and implementation proposal for the M200 floor: capacity planning, AP placement, channel planning, a Bill of Materials, and a validation analysis. One of the main outcomes was revising an initial estimate of 83 access points down to a final proposed design of 54 APs, after reconsidering our active-client assumptions against what we saw during the site survey. The proposed network was never physically deployed — the validation phase surveyed the existing WLAN instead (see Validation below).

Project Goals

• Reliable 5 GHz coverage (2.4 GHz optional) throughout occupied areas
• RSSI target of approximately −65 dBm
• Minimum SNR of 20 dB
• 40 MHz channels
• Enough capacity for classrooms, labs, offices and the atrium walkways
• Controlled channel reuse with reduced co-channel interference (CCI)
• A workable implementation plan and Bill of Materials
• A post-deployment validation plan

Tools Used

• ExtremeCloud IQ — predictive WLAN design, AP placement, heat maps
• VisiWave Site Survey — passive validation survey of the existing WLAN
• Microsoft Word / Excel — report writing and BOM calculations

Design Process

Site Walkthrough → Capacity Planning → Predictive Design → AP Placement → Channel Planning → Implementation / BOM → Passive Validation → Final Analysis

We walked the M200 floor twice with a printed floor plan, recording room function, estimated occupancy, wall/ceiling materials, possible interference sources, and AP mounting options for every room. Some rooms were locked on both visits; for those we used door signage and the RFP's default assumptions instead of guessing, and noted clearly which values were observed versus assumed. That survey data fed into the predictive design in ExtremeCloud IQ, where we modelled the building materials (drywall, brick, concrete, glass, elevator shafts) and placed access points across the MD, MC and MB wings.

Capacity Planning

This was the most important lesson from the project. Our first pass at the design came out to 83 access points, based on assuming two devices per person (laptop + phone) and effectively treating every associated device as active at the same time.

After comparing that against what we actually saw during the walkthrough — most existing rooms only had one AP — we reconsidered the assumption. We changed our estimate to about 60% of associated devices being active at once (a design assumption we made for this project, not an industry standard), with roughly 50 active clients supported per AP. Recalculating with that assumption, and then checking the result against coverage and capacity requirements for every room again, brought the final design down to 54 proposed access points: 13 in MD, 19 in MC, and 22 in MB and the atrium walkways.

It wasn't a matter of just deleting APs to hit a smaller number — the whole capacity model was recalculated, and the reduced AP count was then checked against the −65 dBm / 20 dB SNR targets room by room before we called it final.

Illustrative recreation based on the actual project calculations — not an original design-tool screenshot.

Channel Planning

The design uses 5 GHz as the primary band with four non-DFS 40 MHz channels: 40, 48, 153, and 161. We stuck to non-DFS channels to keep the plan simpler and more predictable, since the RFP didn't require DFS.

With 54 APs on one floor, some channel reuse was unavoidable. After placing APs, we reviewed the layout in ExtremeCloud IQ and adjusted assignments so that same-channel adjacent APs were avoided where practical, and so that multiple APs inside the same room used different channels where possible. Transmit power was also reduced in some denser, multi-AP rooms to limit overlap between neighbouring cells. The interference heat map showed no major co-channel interference with this plan, though some overlap between same-channel cells was still expected given how many APs shared four channels.

Illustrative reuse pattern — not the actual floor plan or AP layout.

Illustrative recreation of the coverage/roaming concept behind the −65 dBm design target.

Validation

For validation we used VisiWave Site Survey to passively survey the existing eduroam network on the M200 floor — not our proposed 54-AP design, since that design was never physically installed. We walked the accessible parts of all three wings and the atrium, comparing 5 GHz and 2.4 GHz coverage, expected data rates, channel usage, and co-channel interference.

The evaluation version of VisiWave flags that its results may be altered, so we interpreted the survey mainly for general patterns rather than exact numbers. It was still useful: it let us compare our design assumptions against real RF conditions in the building and see how the existing network's channel usage and interference looked next to our predictive plan.

Qualitative summary of passive survey observations, not exact measured values. Survey represents the existing eduroam network, not the proposed 54-AP design.

Challenges

• Locked/inaccessible rooms during the walkthrough — used door signage and RFP defaults where we couldn't get in, noted clearly as assumptions.
• Incomplete data on the first pass — gaps in wall-material notes and survey measurement points meant returning to the site more than once.
• Initial AP overestimate — the first capacity calculation produced 83 APs, which prompted us to revisit our active-client assumptions after comparing the result with our site observations.
• Channel reuse with 54 APs and only four planned channels — required repeated review of the channel/interference maps, since changing one AP's channel could affect channel reuse and interference conditions for nearby APs.

Key Takeaways

• More access points doesn't automatically mean a better WLAN — adding APs can add interference just as easily as it adds coverage.
• The number of associated devices isn't the same as the number of active devices; capacity assumptions need to reflect how people actually use the network.
• Site surveys often take more than one visit, especially in a building with locked rooms.
• A predictive design is only as good as its inputs, and a validation survey — even a passive one on a different network — is a useful reality check.
• Coverage, capacity, channel reuse, transmit power and building materials all have to be considered together, not one at a time.

Skills Demonstrated

• WLAN site surveying
• Predictive WLAN design
• Wireless capacity planning
• AP placement and channel planning
• RSSI / SNR analysis
• Co-channel interference analysis
• Passive WLAN validation
• Technical documentation / Bill of Materials planning

My Contributions

This was a two-person project. My teammate and I worked together on the site surveys, major WLAN design decisions, and reviewing each other's work.

I focused primarily on:

• Walkthrough observations and documentation
• Predictive WLAN design and AP placement
• Capacity calculations and assumptions
• The 83 → 54 AP design revision
• 5 GHz channel planning
• Transmit-power adjustments
• Implementation planning
• Bill of Materials review and cost checking
• Final report integration and consistency review
• Passive validation surveying using VisiWave Site Survey
• Passive survey data collection across accessible areas of the M200 floor
• Review of existing eduroam 5 GHz and 2.4 GHz coverage
• Expected data rate and channel usage analysis
• Co-channel interference analysis
• Comparison of observed WLAN conditions with our predictive design assumptions