Problem with collection of data

[Obinna Nnamani]

Hi...

Please I need help... I am new to ns3 and trying to implement my thesis on it using LTE network... However, i am encountering some issues... I have read other thread but do not seem to understand how to resolve my issue...

I want to plot throughput against distance... I have a ueNode moving away from an eNBNode... I want to plot the throughput against distance as the user moves and subsequently, get the point where the ueNode stop receiving bytes from them the eNBNode. This will give me the coverage radius of the node with my set parameters. The problem is that Flowmonitor, RLCStat etc gives end to end results so I decided to schedule app to be created and sent after 1s while also checking the user position. This works fine except that each time I call the app generation at different distance, the throughput is the same... But when I send more packets at a scheduled time, the throughput would vary...

Please what am I doing wrongly...

I have attached my codes to this post...

[Konstantinos]

Hi Obina,

What is the problem of FlowMonitor and RLC stats? I do not understand what you mean by end-to-end results.

And what exactly is your application doing? Isn't it generating a packet from one end and you receive it on the other end, i.e. end-to-end...

For both methods (FlowMonitor and LTE stats) you can get results over time and by correlating the time with the position(*), you can have results over position.

In LTE stats that comes as default in the files. For FlowMonitor you should schedule the calculation periodically (discussed in the list quite a few times).

Indeed if you only send one packet, the throughput will not change dramatically. What you could see is only reception or no reception. 

However if you have a continuous transmission you should be able to see differences in performance.   

(*)By scheduling the movement of the node to go X meters away after Y seconds, when you plot the results over-time, you can group each Y seconds to be at the same distance. 

Regards,

K.

[Obinna Nnamani]

Hello Konstantinos,

Thank you for your reply... I am sorry for my late response, I was trying to clarify myself on my problem... I failed again at that... Please I will try to explain to you my understanding of ns3 and thus seek further clarification... Thank you...

I think I have succeeded in setting up a typical LTE network upon which I would implement my work... However, I need to obtain output from this network and compare with other works as a means of validating the network. Currently, the output I am interested in is throughput, packet loss and delay upon distance between user and ENB...

Flowmonitor is IP based (layer 3) because its statistics are based on packet interaction at that layer... However, though it can be installed during the simulation, its data can only be accessed at the end of the simulation. It follows flow distribution. This means that it can not be scheduled during the simulation. In other to schedule flowmonitor effectively, I think it is best to schedule your flow of packets... That way, flowmonitor can also be called back at the scheduled time... That is what I have implemented in my previous attached... I tried to schedule application based on distance traveled by the user... This gives me results but my question is based on my results... Each time I schedule a flow, the same results are obtained but increasing the number of packets generated at a particular distance shows variation in different flow statistics... Why is this the case??? I expected the reverse...
These are my understanding... please correct me where I am wrong... Thank you...

Please if there are suggestions on other better ways to achieve obtain results... I would appreciate...