Real-world testing of Wi-Fi hotspots
The first hotspots were small-office/home-office (SOHO)-class access points, generally used for residential connectivity, with a simple Wi-Fi connectivity process and coverage that was designed for household use. While some businesses still try to leverage this approach, the method lacks the performance required for today’s public hotspots. This increased demand for bandwidth has left those offering hotspot connectivity with a choice: either deal with poor performance and frustrated customers or install enterprise-class equipment to support use expectations.
Connectivity is so important to consumers, that it’s not uncommon for them to select a destination or method of transport based on the cost and quality of Wi-Fi Internet access. It is also not uncommon for them to select one coffee shop over another based on high-speed Internet access. But, what do these consumers think about when looking for hotspot connectivity? And, how can a business ensure a positive experience for their customers?
First let’s look at some of the common hotspot features in-depth and how they are facilitated.
Ease of use is a feature
Hotspots use the 802.11 open authentication method, meaning no authentication process at Layer 2 – at all. The customer’s client device (laptop, iPad, smartphone, etc.) joins the hotspot’s SSID, and is forwarded to the DHCP service, and the client device receives an IP address, default gateway and DNS. This, in its purest form, is hotspot connectivity.
At this point the client is now ready to access the Internet. One option is to just allow direct access. This is the easiest of all systems. It causes no difficulty with devices, because there is no user interaction.
However, most hotspot providers opt for a captive portal solution – whereby any attempt by the client device to either load a browser-based Internet session, check e-mail, etc., will all be redirected to an HTTP web page. By capturing all possible outbound ports, the customer’s experience is changed from what they would get at home.
On this captive portal page, the customer can choose to accept the terms of service, and/or pay for Internet usage. The use of a captive portal makes accessing the Internet via a hotspot quite difficult for devices that do not have native web browsing capabilities. The more “hoops” a customer has to go through, the lower their valuation of the hotspot service.
Bandwidth/throughput
The next feature that is on the top of customer’s mind is the actual throughput of the connection. If Internet access is slow or inconsistent, customer complaints rise. Gone are the days when a 100-bed hotel could utilize a single T-1 line (1.5 MBs) being shared between all the guests.
With the advent of streaming audio and video services — like Spotify, Pandora, Hulu and Netflix — users expectations of throughput have increased faster than most hotspot providers have increased bandwidth. A business can have the best Wi-Fi system available, with fantastic data-rates going over the RF medium, but without an adequately sized backhaul, end users will still complain.
Advanced features
Many customers are getting increasingly more sophisticated in their IT skills and use of technology, such as with public IP addresses, VPN support, and even higher in-bound needs. Many smaller hotspots won’t need to address these more advanced features — but in airports, conference centers and hotels, the ability to offer access to these features will be paramount to those users who need them.
Now that we have a better understanding of the consumer expectation, how can a business measure and analyze hotspots to ensure performance?
There are two parts to every Wi-Fi hotspot service. The most obvious is the Wi-Fi component — the ability to use radio frequencies to transmit packets from the client devices to the Internet. The second, and just as important, is the backhaul to the Internet.
Many of the earlier pioneers of Wi-Fi systems mistakenly thought the main goal of designing Wi-Fi was all about the RF coverage, specifically the measure of received signal strength indicator (RSSI). This measurement is usually captured in decibels compared to one thousandth of a watt (dBm). Client devices have a calibrated receive sensitivity at different data rates, and they need a certain amount of RF signal above the ambient RF noise floor in able to operate.
Though measuring the RSSI in any given target area is certainly important and necessary, simply focusing on this alone is not sufficient. In order to capture, analyze, and report the performance of any given hotspot, it’s necessary to measure the actual throughput of data, not merely the RF energy. To do this, we need tools that can consistently replicate and collect data in a known method and repeatable format.
The first way to evaluate a hotspot is to use a tool to capture active Wi-Fi signals throughout the facility’s footprint to verify adequate RF coverage and performance. Professional tools, like Fluke Networks’ AirMagnet Survey PRO, will also provide helpful, visually appealing heat maps and full reporting capabilities.
Testing and measuring layers 1-3: RF, 802.11, and IP connectivity
After RF coverage has been verified, the next step is to use tools to test and measure performance of the hotspot. Visualizing this performance in a “performance weather-map” can be extremely effective (see Figure 1). Below are a few of the metrics we’ll want to look for in our performance analysis:
RSSI: The amount of RF energy received at any given location and time.
Noise: This can be captured with a Wi-Fi network interface controller (NIC) to show packets flowing using RF in the area, and augmented by a spectrum analyzer to see non-modulated RF from other potentially interfering devices.
Signal to noise ratio (SNR): The difference between the RSSI and the noise floor. Higher SNR values are preferred and are indicative of higher data rates. The faster each client gets on and off the wireless medium, the more clients can share the frequency in the same space without causing interference.
Data rates: The total data transfer rate the device can handle.
Throughput: Throughput is usually defined as the amount of actual data that can be transferred across the network in a given amount of time. It is the only metric that can truly represent the true end-user experience for any connection. The actual throughput will always be lower than the data rate.
802.11 association: The 802.11 association is to Wi-Fi what a ‘link light’ is to an Ethernet connection. It is the minimum requirement that shows connectivity between the device and the rest of the local area network. In Wi-Fi connections, we need the basic service set identifier (BSSID) — or the MAC address of the access point we are connected to. This BSSID will be used to help move packets to and from the wireless network.
Dynamic host configuration protocol (DHCP): In order for any device to transmit packets to the Internet, it requires IP address information for the specific subnet it is connected to. Quick, repeatable DHCP responses, with complete answers, including default gateway, domain name system (DNS), and subnet mask information is a hard requirement.
At this point we have Layer 1 – RF, Layer 2 – 802.11, and Layer 3 – IP connectivity. This is the minimum in order to connect via the hotspot to the Internet. These together mean our client device is connected to the AP as well as through the AP back to at least the DHCP server. However, it is also important to test and measure the connection from the default gateway on to the Internet itself.
Testing and measuring default gateway connectivity to the Internet
To test off-site connectivity, use two standard internetworking tools — PING and TraceRoute. These tools provide metrics for how long and for how many hops it takes to get to a site on the Internet. The lower the PING times, the quicker the access. TraceRoute will show the total number of hops via routers on the Internet from the current location, to the designated target. These two show speed and distance, but not bandwidth. For that, additional tools are required.
To test the throughput to a target on the Internet, three other tools can provide more flow testing. The first is file transfer protocol (FTP) which is used to send large chunks of data to/from an FTP server. This is tested by simulating a large file download, and it is the first way to test throughput.
The second is to test HTTP file transfers. Again, this is just a way to force a client to send large amounts of data to a server on the Internet and capture the flow statistics. Finally, test the multimedia download, such as a streaming a video or audio file. These can be implemented to simulate watching a movie via Netflix, or listening to an audio stream via Pandora.
The best approach is to have all of these throughput tests running simultaneously, putting the most load on both the wireless and wired portions of the hotspot.
The toolset
Using a professional toolset that combines all of the features needed to analyze and evaluate a combination of wireless and wired tools can be extremely efficient. The following example gets this functionality from a software solution, AirMagnet WiFi Analyzer PRO.
We used the system’s One-Touch Connection Test feature to perform 802.11 Association, ping, trace, FTP, HTTP, and multimedia testing across multiple locations in a simultaneous manner (see Figure 2). We used the system to generate a written report for customers and staff. The tool can also automate login and testing processes.
Hotspots are increasingly being used throughout business to bolster connectivity with customers, partners and employees. While the technology can (and will) vary, the need to properly evaluate and troubleshoot hotspots to ensure client satisfaction will always require testing and measurement.
If you are responsible for a Wi-Fi Hotspot, you might want to make it a quarterly process to test and evaluate your hotspot to make sure you are meeting the needs of your users. This way you can adjust and adapt your network, keeping it up-to-date with current technology and expectations.
About the Author
Dilip Advani is group product manager for wireless solutions at Fluke Networks.
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :
