Welcome to Research

We propose a non intrusive methodology based on mobile devices real measurements. Our objetive is to provide a more realistic test environment using actual mobile devices in real conditions of network load radio propagation while taking into account the intrincis mobility o mobile subscribers

Tool Chain

In order to analyze the RTP traffic performance in mobile networks we have used tool chain shown in figure. Test starts with the video transmision from the Darwin Streaming Server to a streaming cliente locates in the mobile device. We use SymPA tool for capturing incoming traffic in the mobile client and network monitoring. Traffic captured is converted to tcpdump format and it is analyzed with the well-known network protocol analyzer WireShark (previously knwon as Ethereal). We have used a modified version of WireShark. The code has been modified in order to add the capacity of analyzing RTP flows established using RSTP protocol. Using the information about network collected by SymPA and the traffic anaylisis obtained from Wireshark we analyze IP level performance of RTP streams.

Scenario #1 Fixed-to-Mobile video streaming session: GPRS cell-reselection

Scenario #2 Fixed-to-mobile scenario: Handover

Scenario #3 Fixed-to-Mobile static scenario: Sequence errors

Scenario #4 Fixed-to-Mobile staci scenario: Packet losses, low bit rate video

Scenario #5 Mobile-to-Mobile static scenario: Packet losses on the source side

Scenario #6 Mobile-to-Mobile static scenario: Packet losses on the receive side

Scenario #7 Mobile-to-Mobile static scenario: Packet losses, low bit rate video

Scenario #8 Mobile-to-Mobile static scenario: Sequence errors

Scenario #9 Mobile-to-Mobile vehicular scenario: Packet losses, low bit rate video

Scenario #10 Mobile-to-Mobile vehicular scenario, low bit rate video: Jitter

Our future research activies include the use of OPNET simulator in order to compare the practical results obtained using our testing methodology with the results obtained from simulations.

The results obtained will also be published in the following web page: http://www.lcc.uma.es/~pedro/mobile/Research/opnet.html