CharlieTLe / qurinet

Data sets and scripts related to Qurinet, a wireless mesh network testbed located at the Quail Ridge Natural Reserve.
2 stars 3 forks source link

Real-time Visualization of OLSR Topology, OLSR Control Message Flow, Bandwidth between Links, and Ping Data #15

Open DineshUCD opened 8 years ago

DineshUCD commented 8 years ago

Feature: Implement a real-time visualization in JavaScript that represents OLSR topology, bandwidth between links, ping outputs, and tracks the rate and size of OLSR control messages leaving an interface.

Purpose: Use the Web Application to modify the OLSR configuration parameters in acclimating to the Natural Reserve's computer networking requirements.

Outcome: A better understanding of how OLSR functions, finding the root cause of previous issues discovered by former researchers who set-up the test bed, and troubleshooting any potential connectivity issues.

Status Task
In Progress - Near Completion Monitor ping time on Simple XY Line, RTT on parallel bar graph, and packet loss on donut chart.
In Progress - Near Completion Use a network-link diagram to show the types, status, and willingness of routers. Encode bandwidth in links connecting them (drawn as line marks).
In Progress - Very Near Completion Monitor bandwidth output from IPerf for each link on a Multiple XY Line Chart. On new IPerf outputs, append the average bandwidth for each link in a separate Multiple XY Line Chart
Open Show the rate of TC, HELLO, and MID messages going out to a link. Display the to-and-fro information for each packet, as well as the types of messages encoded in each packet. Show the size of each message sent from an interface to its destination (The broadcast address).

More Information: According to what previous researchers told to Todd, the control packets caused some congestion. The task to to parse tcpdump data with pyshark, dpkt, tshark, wireshark, or any other way on the terminal to show how the rate at which TC and HELLO messages leave an interface. Then, we can adjust OLSR parameters to confirm that the rate decreases depending on what parameters we alter. This would validate our OLSR configuration modifications.

It would be great to have numbers relating to the transmission overhead that the dynamic routing adds to the system. This doesn't have to be in real-time, using tcpdump should give you some type of answer. This had been mentioned in the past as a possible cause of poor network performance but that may have been just an excuse given for why the network had bad performance.

"tshark -r -R olsr -2 -T fields [-e field_name]+" prints the fields for every packet captured in the pcap file.

For example "tshark -r test5 -R olsr -2 -T fields -e olsr.message_type" prints out comma separated message types for every packet. If a packet contains one hello message, one TC message, and one HNA message, the command will print out "201,202,4" on a single line.