Connection Quality Test Frequently Asked Questions

:?: What is the difference between the Connection Quality Indicator and the Speed Test?

The Connection Quality Indicator involves a quick speed test that is automatically conducted after a connection is established using Open Mobile. This quick test involves a small amount of data, and it is only good for a general assessment of the connection. The Speed Test is a user-initiated test that uses a larger amount of data for a more precise measurement. This would be too large a test to automatically generate after each connection, as it would compete with the user's real traffic.

:?: When I run a speed test from an Internet site, why does it show better performance than when running the speed test in Open Mobile?

There are many factors that can cause variances in performance reporting, including the user's location in relation to the test server and Internet congestion between those two points. Also, some Internet performance tests use a different method of testing than Open Mobile. Some sites flood the line with multiple transfers with the goal of getting a theoretical maximum throughput for the connection. Open Mobile uses a simple file download and ping test in order to gauge the connection performance, because we feel this test would better match user behavior and what they could expect when trying to download a file or watch streaming video. Because of this difference in methodology, test websites may show better numbers than the test built into Open Mobile.

:?: Why did iPass not partner with one of the popular performance testing web sites instead of re-inventing the wheel and building their own speed test?

We tried to find a good partner, but we ran into an obstacle with each one we evaluated. One leading vendor's test used Flash, which would not have worked on our iOS and Android platforms. Another would have been extremely cost-prohibitive. It is possible we will re-evaluate this again in the future, but for right now, we are moving forward with our own test.

:?: Why does the connection quality indicator never change over the timeline of the connection?

Currently the connection quality test is conducted just once, at the beginning of the connection. The trade-off of a test like this is that the download needed to do the test can compete with the real work the end user is conducting. For the first release, we decided on a one-time test. After we have some data to judge the impact on the user, we will consider implementing an intelligent algorithm which will update this indicator periodically.

:?: Is the video indicator for video streaming, or for two-way video conferencing?

The video assessment is for video download only, not two-way conferencing. We considered both, but decided that video streaming is the far more common use case.

:?: Why does it show my connection is suitable for video streaming, but not VoIP (Voice over Internet Protocol)?

VoIP can be very latency sensitive. So even if download speed is good, a connection with high latency will not be good for VoIP. For video streaming, this latency can be handled by the application. If the indicator was for two-way video conferencing, then a connection assessed as not suitable for VoIP would also not be suitable for Video Conferencing.

:?: Where are the Test Servers located?

iPass provides two test locations in the United States (Santa Clara, California and Atlanta, Georgia), one in England (London), and one in China (Hong Kong). Customers and Partners can add their own servers. For customers not in Western Europe, the United States, or Australia, we would recommend they include their own test server for best results.

:?: How does the software determine which Test server location to use?

The software issues a ping to each of the test servers configured. The one that responds first is the one that is chosen to be the test server.

:?: Will iPass add any more test servers in the future?

Possibly but nothing is planned in the short term.

:?: Can a customer use their own test server?

Yes, and for many customers, this is the more expected use case than using the iPass test servers.

:?: Is it better for a customer to just have one test server at their main location and not use the iPass test servers?

In some circumstances, it may be better. For instance, if a customer has a lot of VoIP traffic that is destined for their home location, then a server at their location would give the best result for assessment of VoIP suitability for that connection. If split tunneling is disabled and VPN usage enforced, then a test server at the same location as the VPN server may make the most sense. But in other situations, where general Internet access is the most common function that is going to be used by the individual user, the more test servers the better.

:?: What are the requirements for a test server?

The test server must answer http requests and hold a couple of files that are used by the client to test the connection. The server must also be able to be pinged from the Internet. See more information here: Connection Profiler
 

©2015 iPass Inc. All rights reserved. Terms of Use