Increasing average download throughput by 170%: The proven capability of our Connectivity Experience Solution
At Fontech, we have developed our Connectivity Experience Solution (CES) so that CSPs can provide an always-best-connected experience to their subscribers. This solution is based on 3 key components, each one geared towards a specific goal:
- Connection Manager: An SDK or app on the end user’s device that seamlessly connects the user to WiFi.
- The QoE Manager: Smart algorithms that select the best connectivity option among multiple WiFi or cellular networks.
- Network Analytics that provides insights on network performance and user experience.
Out of these three components of our Connectivity Experience Solution, we’ve often found that the QoE manager is the one that sparks most interest. How is it capable of identifying the best connectivity option? What implications does this have in real-life deployments? Do users actually perceive improvements? That’s why today, we want to reveal the discoveries of our latest QoE Manager testing.
The QoE Manager is a collection of algorithms that assess the quality and performance of WiFi and cellular networks within the proximity of the device, allowing the application to connect to the most suitable network for the user.
Putting our QoE Manager to the test
This June, we performed a QoE Manager A/B test with real users located in France, Australia, and Belgium. This means that we compared the performance of a set of users with our Connectivity Experience Solution installed on their devices, in which one group (A) had the QoE Manager inactive, and the other group (B) had it active. More than 150k sessions were logged during the testing period to ensure the meaningfulness of the results.
The Connectivity Experience Solution QoE Manager allows for the configuration of different network selection policies that can be more or less conservative in terms of deciding what WiFi networks to select and connect to. This means that an operator can choose desired preferences, whether they be WiFi-first, cellular-first, or balanced. In this test, the selected policy was WiFi-first, which is the least restrictive when it comes to selecting which WiFi network to connect to. This policy makes the device connect to the best available WiFi network according to the QoE algorithms. A cellular connection will only occur when there are no available WiFi networks in range. Among the available operator preferences, this configuration maximizes offload while continuing to ensure an optimized QoE for subscribers.
How did our QoE Manager perform?
Results show a significant improvement in metrics that have a direct impact on user QoE:
- 170% increase in average download throughput
- 15% increase in average upload throughput
- 7.3% reduction in average HTTP session establishment time (related to latency)
- Shorter session duration as a result of acting over sticky client scenarios
It’s important to note that all the metrics collected by the QoE Manager are based on passive measurements and no active tests were performed. The full potential of the QoE Manager improvement for upload cannot be observed by passive measurements but had we carried out active tests, upload throughput could increase as much as download throughput.
While these metrics show an improvement in users QoE, traffic offload remained consistent across both groups. This shows that the QoE Manager is capable of improving user experience while maintaining optimal offload levels for the operator.
These successful results prove the benefits our Connectivity Experience Solution can provide to our client’s subscribers when connecting to WiFi networks. In the future, we’ll be carrying out additional tests to assess more policies and configuration options. Stay tuned! And make sure to follow us on Twitter and LinkedIn for the latest industry insights and Fon | Fontech news!