User experience in public Free-WiFi

Today’s world lives in the paradigm that “WiFi = Internet & Internet = Life”. Therefore it is not surprising that governments around the world consider making WiFi a fundamental right of their citizens. In many ways, WiFi is today what electricity was in the 50s. Many cities around the world already offer Free-WiFi . We even see political parties include Free-WiFi in their election manifestos, especially in developing economies. Free-WiFi is bound to be a ubiquitous presence in our public spaces.


(Picture Courtesty @RealErlich)

However, how free is this Free-WiFi? What kind of a user-experience can one expect from Free-WiFi? Such questions are largely unanswered & unthought of under the pressure to show implementation. It is not very uncommon that one goes to a space which has Free-WiFi, spots a very good WiFi signal, but cannot even connect to the network. Connection reliability and performance requirements of most popular applications are beyond the capabilities of poorly implemented Free-WiFi networks. This results in high costs with near zero value add to the general public.

To address this issue, clearly defined SLAs are needed. These should incentivize integrators, set up expectations for users, and act as guidelines for administrators. Also this will result in proper planning of WiFi coverage and capacity. A closer look at the details on how WiFi works will help to understand the challenge better. WiFi works in 2 ISM bands; 2.4GHz and 5GHz. Each of these bands consists of many channels. Number of channels and channel center frequencies varies slightly from country to country. A typical 802.11n channel is of 40MHz of bandwidth and allows a maximum theoretical through put of 450 Mbps with a 3×3 MIMO antenna, at the physical layer. Practically, applications running on user’s devices can get around 220 to 270 Mbps. Does that imply that 200 users can be served 1 Mbps each on a single channel? Not really. Since users are independent of each other and their actions are not co-ordinated, many clients try to use the channel at the same time and these collisions, brings the efficiency of the channel down. Also typically there are resource limitations in the WiFi Access Point and rest of the WiFi infrastructure, which brings down the number of users that can be served by a single Access point over a single channel.

Graph below shows this more objectively. As the number of clients increases total throughput given by the Access Point decreases rapidly

image-1(measured using SWAT WiCheck multi-client emulator

The degree of deterioration varies between different makes of Access Points and to some extent defines the difference between a consumer grade AP and an enterprise grade AP. Each AP OEM, has their own method to achieve the best result in the most cost effective way.

Solution to improving the efficiency of WiFi networks also needs a broader perspective. Starting point is definitely the allocation of more bandwidth to the ISM bands used by WiFi. World over WiFi ecosystem has been appealing to policy makers in their respective countries for more spectrum. Moving to newer and better standards (802.11ac) that allow much higher throughputs also will help.

Many describe WiFi as the new oxygen. As Free-WiFi networks increase their reach, need for increasing the efficiency of the network becomes more and more important. Each part of the ecosystem, starting from policy makers to OEMs and service provides needs to be sensitive about this and work as a team to find better solutions.


Video Performance over Networks


All of us have seen the familiar picture of “Loading Circle” while watching videos streaming over the internet. Video Streaming is probably the largest consumer of the bandwidth today. Every user would want to avoid the loading circle as much as possible; or the ISP would rather measure and plan the network to make it work smoothly !

Video Streaming, especially real-time Video Streaming is done via RTP over UDP traditionally. This is soon giving way to HTTP streaming (over TCP) .

Performance of Video Streaming over RTP/UDP is measured with Mean Opinion Score (MOS). MOS is done by making a bunch of guys watch the video, ask them to rate their experience on a scale of 1-5 and take the average. Since doing this frequently would be too costly, experts (Ex:Video Quality Experts Group -VQEG) have created models that recreate the video as may have seen by a user to predict the MOS. This would normally be a  function of

  • packet loss (different types of frames )
  • delay
  • jitter

The models are many and sometimes complex and beyond the scope of this blog. These are measured at the receiver using specialised applications and some parameters are derived from packet captures (Ex: tcpdump).

With large amounts of bandwidth available per user, Video Streaming via HTTP/TCP  has become the most popular form of Video Streaming. Since there is no loss involved due to TCP retransmissions, the quality metrics are also greatly simplified. The sample of metrics that are measured now are very intuitive Ex:

  • Initial Buffering Duration (How much time did the user wait after the click)
  • No of interruptions (How many times the user has seen the Loading Circle)
  • Total Buffering time v/s Video Duration (How much time did the user waste waiting due to poor network!)

The above metrics are derived from real/simulated applications with lesser computation than the metrics for RTP/UDP.

Popular sites like Youtube, Netflix use more advanced adaptive streaming technique called Dynamic Adaptive Streaming over HTTP (DASH). DASH brings in changes between bit rates to allow for video to play based on network performance. Performance measurement with DASH is a topic for another blog.



Above chart shows the performance of the network when a number of users simultaneously access a youtube link as measured by Alethea WiCheck. In this test, the clients configured NOT to use DASH (adaptive streaming). The network backbone supported an internet throughput of 80Mbps. This network seems to work well for up to about 50 Clients (5 interrupts). The AP used in the above set up is TP-LINK WDR4300 (802.11n). The test is done using Alethea’s WiFi load test solution SWAT WiCheck (more about the tool at

Customer Matters

In the WiFi industry, we are very focused on advertising performance in terms of megabits and gigabits. Like the processor wars of the 90s and naughties, we are pushing the story that faster is better. And look where it led the processor industry – unsustainable and ultimately costly strategies towards ever faster, ever hotter, ever costlier processors that added very little incrementally to end user experience.

The reality is that 99% of the customers – with the exception of engineers & nerds – do not think of network performance in terms of megabits or gigabits or Real customers see it in use cases: “how fast can I download this file?”, “how good does this movie play on youtube?”, “will my Lync call go well?” or “how easy it is to search something on Google?”. Ultimately, all this depends on throughput, but there is more to user experience than just throughput numbers.

Customer experience includes every step from the point he or she tries to connect to the Access Point to the speed at which the captive portal allows log-ons, to finally to the experience on a specific application. When we as an industry set out to meet customer expectations, we need to understand what our end users judge us on. However well we do with esoteric settings and technologies, at the end of the day, they are all wrapped within the final user experience.

To address this challenge to testing what matters to the customer, we have to look at completely revamping the focus of WiFi infrastructure validation. I see the following as important to each market segment:

– How many devices can be connected simultaneously
– Youtube performance
– Skype / SIPcall quality
– Facebook update speed
– Torrent performance
The one thing that stands out here is how the number of connected devices in a household has increased exponentially in the last few years. It will not be surprising that we need to support 20 – 30 devices in a household TODAY. This number will definitely grow in the future.

– Initial connection time for 32 devices
– Time to captive portal landing page / destination page
– Browsing time
– File download performance
– Video streaming quality

– E-mail
– Lync / Skype for business calls
– Browsing speed
– Video streaming quality
– File download performance
– Number of simultaneous connections to an AP with a pre-defined SLA

These applications / use cases cover 90% of the network activity that we see in each of these user environments and gives us a robust framework to think what is important to the customer. Please share your thoughts on this differentiated approach that we have discussed above.