|
|
|
|
Index table +Intro +Prephase +TimePlan +Outline +Research Phase +-28 Feb.- +-5 Mar.- +-12 Mar.- +-Odysee +-19 Mar.- +-26 Mar.- +-2 Apr.- +Develope Tests Spec. +- 9 Apr.- +Test Phase +- 16 Apr.- +- 23 Apr.- +- 30 Apr.- +Review Results +- > 7 May.- +Prep. Report +- 14 May.- +- 21 May.- +After Project |
Last Test seriesNow after the comprehensive test, and the look at the delay graphs, the question is, if the high number of peaks is effecting the VoIP conversation. To test this a similar test scenario like the multi channel test, but with a real VoIP software is needed. I will try a test with three PC involved. Two having a Netmeeting conversation, while one of this PCs and another laptop is going to run the software to emulate the traffic of a larger number of VoIP conversations in the background. The quality of the Netmeeting conversation should be comparable to the quality of one of the channels, although the difference in this testsetup is, that the Netmeeting conversation has his own receiver and transmitter, while the emulated conversations share one receiver and transmitter.The result was, that the number of peaks can be seen as a quality indicator. The Netmeeting conversation was not possible at normal conditions (about -65dBm) and 4 other conversation running simultanously. Interesting was the that at one conversation per Laptop (means one Netmeeting and one emulated conversation only). A Netmeeting conversation could be started, but had problems, if the other conversation where sending 258 byte packets. If both (Netmeeting and the other conversation) sending 84 byte packets, both are fine. I connect this to the way the media access is handled. At 258 byte packets a RTS/CTS is needed, this means a handshake and correct transmission control is done. While the 84 byte Netmeeting trys to send in the breaks and sometimes fails - the conversation is not all the time good. At both 84 bytes, both are sending without RTS/CTS, then the chance that they are colliding or transmissions succeeds is equal - hence both conversations are alright. ReportThe rest of the time was used to write the last two parts of the report. |
| © and created by Jörg Abendroth | |