It can take a while depending on the size of the document..please wait
Discuto
Starting: 22 Aug
Ending
0 days left (ends 29 Aug)
Go to discussion, participate and give your opinion
description
Further info
LATEST ACTIVITY
LEVEL OF AGREEMENT
MOST DISCUSSED PARAGRAPHS
No activity yet
LATEST COMMENTS
No activity yet
MOST ACTIVE USERS
No activity yet
Status:
Closed
Privacy:
Public
CONTRIBUTORS (2)
Share:
_
filter:
All Most voted (0) Most comments (0) Controversial (0) Decided (0)
In voting (0)
Supported (0) My contributions (0)
view:
<< Previous paragraphs
P19
As proposed in the 2012 BEREC NN QoS Guidelines [2], the measurement methods shall encompass both the IAS as a whole, as well as individual applications using it. The methodology supports both IPv4 and IPv6 - this topic is further discussed where necessary.
Add comment
P20
The aim of this chapter is to specify the measurement methodology best practices with the combined goal of maximising measurement accuracy and to ensure that the measurement results are comparable between different member states.
Add comment
P22
• Empowering the end user to validate the commitments made to them from their IAS provider.
Add comment
P23
• Monitoring the general IAS quality and confirming that the performance of IAS is developing sufficiently over time when taking into account technological evolution.
Add comment
P24
• To support the detection of any prioritisation and/or throttling of selected applications compared to other applications running over IAS.
Add comment
P25
• NRAs may also use the data to increase transparency (e.g. interactive maps showing performance in a geographic area).
Add comment
P26
According to BEREC NN guidelines [3] paragraph 166, speed should be calculated “based on IP packet payload, e.g. using TCP as transport layer protocol” and according to the NN guidelines paragraph 140, an ISP should define the speed on the basis of the IP packet payload or transport layer protocol payload.
Add comment
P27
This methodology is targeted to measure IAS quality in both the upload and download direction. It is worth noting that IAS speed is just one component of the performance experienced by the end users, since different applications have different protocol overheads and different requirements related to IAS delay, delay variation and packet loss.
Add comment
P28
For both measurement tasks - IAS as a whole and individual applications using IAS - the fundamental precondition is that measurements are performed at the edge of the network which provides the IAS (i.e. end user premises for fixed access or via the radio access for Mobile IAS).
Add comment
P29
Where measurements are performed against a test server, this server should be located outside the IAS network. It should have adequate connectivity between the server and the IAS provider to avoid influencing measurements. In general it is recommended that the measurement server should be located at the national Internet exchange point (IXP) unless there is specific reason for its placement elsewhere.
Add comment
P30
Monitoring mechanisms should mitigate, to the extent possible, confounding factors which are internal to the user environment. Examples of these factors include existing cross-traffic and the usage of Wi-Fi based interfaces. This topic is discussed separately in chapter 5.
Add comment
P31
The assessment of measurement results is discussed further in chapter 6 and the certified monitoring mechanism is further discussed in chapter 7.
Add comment
3.1.1Speed measurement overall methodology
P32
In order to maximise compatibility in a real world environment, it is recommended to measure upload/download speeds based on the time to execute a set of controlled file transfers over HTTP.
Add comment
P33
This methodology is supported by the broadest range of platforms, and can be implemented within a web browser or within the restricted sandbox of an on-device app. As such it is considered to be the best compromise between the competing demands of accuracy, platform agnosticism, ease of implementation and transparency.
Add comment
P34
Another reason to recommend the use of HTTP is to mitigate any firewall based restrictions which may result from the choice of a less commonly used protocol/port. The use of HTTPS also prevents manipulation from proxies[1].
Add comment
P35
In order to saturate the path, it is recommended to use 3-5 HTTP connections. Furthermore, these connections should all have completed the TCP slow start phase to maximise throughput and ensure that the measurement is as representative as possible. The test is stopped after a pre-defined interval and the transfer speed is calculated by the recipient based on the data transferred over that interval.
Did you know you can vote on comments? You can also reply directly to people's comments.