I have done simulation for a topology of 3 nodes (Node-A, B, and C). Where Node-A is the source and Node-C is the sink, Node-B is the intermediate node between A and C. Now I want to calculate the end-to-end delay for the packets sent from source to sink. For this, I am parsing the trace.json generated after the simulation. However, in trace.json the threadID or the messageID generated for the datagram at the source(Node-A) is not the same when that datagram is forwarded at the intermediate node (Node-B). Thus, when I parse the events pertaining to the sink (Node-C) the threadID or the messageID matches with events pertaining to the intermediate node (Node-B) but not with the source (Node-A). So I want to know, how I can calculate end-to-end delay for the packets transmitted between source and sink. Please let me know, also let me know is there any other approach I can follow to calculate the same. please find the generated trace.json of the simulation here
Calculating end-to-end delay for a multi-hop network simulation using trace.json in UnetStack
Viewed 67
1 Answers
Each node has its own thread of execution. The JSON has stimulus and response message IDs for each entry, allowing you to trace across agents and nodes. To understand this better, you may want to read this blog article.
You may also want to look at the viztrace utility. This utility does the heavy lifting of matching up stimulus-responses to create a coherent trace from the logs, and can even generate sequence diagrams for you. You could modify this utility to meet your needs, or use the output from it for easier analysis.
Related