Need vehicle to act as Sender as well as Receiver in D2D Multicast scenario of simuLTE

[Aman]

I have been trying to create a comparable scenario for 802.11p (veins) and LTE-A (simuLTE). In this comparable scenario, I intended to use a highway map with varying vehicle density to analysis Packet reception ratio, Inter Reception time and end to end delay.

I have run into two major problems in order to achieve this : 

1. I want to only compare the performance of broadcast messages (BSM/CAM) in both these frameworks. The way Veins works is that each vehicle takes turns to be the receiver and sender at some point in time during the simulation. However, correct me if I am wrong, but as per my observation in simuLTE, the D2D Multicast configuration does not allow vehicles who have been assigned as senders to act as receivers at some point in time during the simulation and neither can vehicles who have been assigned as receivers as a senders during the simulation. Is there a way to modify the D2D Multicast scenario to allow a vehicles to act as receiver as well as sender?

2. Another problem is that the simulation runs into run-time error "MAC receives feedback on an unexisting H-ARQ tx buffer​". I have created a 1 Km highway map with 4 lanes to run vehicles with various densities ranging from 25 vehicles/km/lane to 100 vehicles/km/lane. For 25 vehicles/km/hr(ie total 100 cars for 4 lanes), the simulation runs smoothly only if I have 25 AlertSenders distributed amongst 75 other AlertReceivers. As soon as I increase the number of Alert Senders, the simulation runs into this runtime error. Also this same ratio of 25:75 for 200 cars throws run-time error. I have raised this issue on github portal for simuLTE (https://github.com/inet-framework/simulte/issues/44). From my observation, increasing packet size or increasing number of vehicles on the road, results in this run-time error. At packet size 10 Bytes, I can run 100 vehicles (50 AlertSenders and 50 AlertReceivers) but by increasing the packet size to 300 bytes, it throws runtime error.

Am I missing some configuration? What could be the reason for this?

I really appreciate someone guiding me to solve this problem.

Thanks!

/Aman

[MK_655]

Le mardi 26 novembre 2019 09:23:46 UTC+1, Aman a écrit :

I have been trying to create a comparable scenario for 802.11p (veins) and LTE-A (simuLTE). In this comparable scenario, I intended to use a highway map with varying vehicle density to analysis Packet reception ratio, Inter Reception time and end to end delay.

2. Another problem is that the simulation runs into run-time error "MAC receives feedback on an unexisting H-ARQ tx buffer​". I have created a 1 Km highway map with 4 lanes to run vehicles with various densities ranging from 25 vehicles/km/lane to 100 vehicles/km/lane. For 25 vehicles/km/hr(ie total 100 cars for 4 lanes), the simulation runs smoothly only if I have 25 AlertSenders distributed amongst 75 other AlertReceivers. As soon as I increase the number of Alert Senders, the simulation runs into this runtime error. Also this same ratio of 25:75 for 200 cars throws run-time error. I have raised this issue on github portal for simuLTE (https://github.com/inet-framework/simulte/issues/44). From my observation, increasing packet size or increasing number of vehicles on the road, results in this run-time error. At packet size 10 Bytes, I can run 100 vehicles (50 AlertSenders and 50 AlertReceivers) but by increasing the packet size to 300 bytes, it throws runtime error.

Am I missing some configuration? What could be the reason for this?

I really appreciate someone guiding me to solve this problem.

>>> I have  the same problem as you.  I observed that it occurs some times where the positions of eNode B and vehicles are so far. When I adjust the distance between nodes and eNodesB it doesn't occur. 

I think it is a bug. I don't think it is related to missing paramaters. 

Thanks!

/Aman

[Aman]

Hey! 
Thanks for reaffirming that this is a problem.

Did you bring the eNodeB close by changing the position in Highway.ned?

[MK_655]

Le mardi 26 novembre 2019 22:56:19 UTC+1, Aman a écrit :

Hey! 
Thanks for reaffirming that this is a problem.

Did you bring the eNodeB close by changing the position in Highway.ned?

>>> I  created my own topology. You can modify   .eNodeB*.mobility.initialX  and 

*.eNodeB*.mobility.initialY  in omnet.ini configuration file

[MK_655]

Le mardi 26 novembre 2019 12:33:52 UTC+1, MK_655 a écrit :

Le mardi 26 novembre 2019 09:23:46 UTC+1, Aman a écrit :

I have been trying to create a comparable scenario for 802.11p (veins) and LTE-A (simuLTE). In this comparable scenario, I intended to use a highway map with varying vehicle density to analysis Packet reception ratio, Inter Reception time and end to end delay.

2. Another problem is that the simulation runs into run-time error "MAC receives feedback on an unexisting H-ARQ tx buffer​". I have created a 1 Km highway map with 4 lanes to run vehicles with various densities ranging from 25 vehicles/km/lane to 100 vehicles/km/lane. For 25 vehicles/km/hr(ie total 100 cars for 4 lanes), the simulation runs smoothly only if I have 25 AlertSenders distributed amongst 75 other AlertReceivers. As soon as I increase the number of Alert Senders, the simulation runs into this runtime error. Also this same ratio of 25:75 for 200 cars throws run-time error. I have raised this issue on github portal for simuLTE (https://github.com/inet-framework/simulte/issues/44). From my observation, increasing packet size or increasing number of vehicles on the road, results in this run-time error. At packet size 10 Bytes, I can run 100 vehicles (50 AlertSenders and 50 AlertReceivers) but by increasing the packet size to 300 bytes, it throws runtime error.

Am I missing some configuration? What could be the reason for this?

I really appreciate someone guiding me to solve this problem.

Can you send me a capture of the error runtime to see from each module it is called? Because I want to check the origin of this problem ?

 

Thanks!

/Aman

[Aman]

Thanks for your comment. I am not sure if only moving the eNodeB helped in my case. I still received the error at a later stage of the simulation than before. To completely eradicate this issue, I switched off the eNode Handover in omnetpp.ini.

I don't have the capture but I can paste the command line output. 

Error: Mac::fromPhy(): Received feedback for an unexisting H-ARQ tx buffer -- in module (LteMacUeRealisticD2D) Highway.car[5].lteNic.mac (id=415), at t=25.051s, event #14073776

Refer to the Github issue link for more details. From what I see, it might be something related to the eNodeB handover. But feel free to trace the origin to the problem with your approach. I will be doing the same in some time. If you happen to find the cause, please post it here.