0

I changed the number of nodes (cars) in erlangen.rou.xml :

<flow id="flow0" type="vtype0" route="route0" begin="0" period="3" number="50"/>

I used getCreationTime() and simTime to calculate the message delay (end-to-end delay) in TraCIDemo11p and in TraCIDemoRSU11p :

void TraCIDemo11p::onData(WaveShortMessage* wsm) {
    simtime_t msgDelayCar = simTime() - wsm->getTimestamp(); 
    msgDelayVectorCar.record(msgDelayCar);
....
void TraCIDemoRSU11p::onData(WaveShortMessage* wsm) {
    simtime_t msgDelayRSU = simTime() - wsm->getCreationTime();
    msgDelayVectorRSU.record(msgDelayRSU);
...

It gives me approximatively the same delay !

Why the number of nodes doesn't influence the end-to-end delay ?

henry townshend
  • 23
  • 4
  • Why should the number of nodes influence end-to-end delay? Surely there are some effects that might cause end-to-end delay to go up, but I am not sure the conditions in your network can give rise to one of these effects. Did you check if they do? – Christoph Sommer Sep 12 '16 at 11:05
  • @ChristophSommer I implemented ECDSA in veins. RSU sign his message and sends it to vehicle, then the vehicle verifies the signature. I want to get some results who reflect the performance used by this security scenario. I read in many scientific articles, that the message delay increased with the number of nodes so I try to prove it in this scenario. If the delay doesn’t change, is there another parameter who can influence the performance of the network? I don’t know on what focusing ! – henry townshend Sep 13 '16 at 22:23
  • When you say "I implemented ECDSA in veins", did you implement just something that produces the correct output or did you implement a computer simulation model of ECDSA? If so, does this model also capture aspects of timing? If so, does it capture aspects of concurrency? – Christoph Sommer Sep 19 '16 at 11:54

0 Answers0