SCOPE: Problem with Multiple BS with basic 1009 scenario

[MUHAMMAD ENAYETUR RAHMAN]

Hello Researchers and @Leonardo Bonati,

I carefully read all the previous discussions and I was able to run 1 BS with multiple UEs, it works perfectly fine. I can see the application data transfers using 'tmux a -t scope'. For example: running 1 BS with 4 or, 5, or, 6 UEs. 

But I was facing problems when there are multiple BS [eg. 2 BS and each one has 4 UEs. so, total 10 SRNs]

- I was trying many many times to run a very basic scenario with almost default configurations as the SCOPE base image (from /share/nas/common).

- I modified each of the SRNs radio_interactive.conf file (given below) and run it in both BS and UEs. 

Configured the radio_interactive.conf file as follows [assuming ]:

Here: 

if SRN-33 ---> enb

SRN-34, SRN-35, SRN-36, SRN-37 are ---> ue as I see after running the 'cd radio_api && python3 scope_start --config-file radio_interactive.conf' in each of the SRNs.

{
"capture-pkts": "False",
"colosseumcli": "True",
"iperf": "True",
"users-bs": "4",
"write-config-parameters": "True",
"network-slicing": "True",
"global-scheduling-policy": "0",
"slice-scheduling-policy": "[1, 1, 1]",
"tenant-number": "4",
"slice-allocation": "{0: [0, 5], 1: [6, 11], 2: [12, 16]}",
"slice-users": "{0: [33], 1: [34, 35], 2: [36]}",
"custom-ue-slice": "True",
"force-dl-modulation": "False",
"heuristic-params": "{'buffer_thresh_bytes': [1000, 2000], 'thr_thresh_mbps': [0.25, 0.75]}",
"bs-config": "{'dl_freq': 980000000, 'ul_freq': 1020000000, 'n_prb': 50}",
"ue-config": "{'dl_freq': 980000000, 'ul_freq': 1020000000, 'force_ul_amplitude': 0.9}"

} 

For BS 2:

Here: 

if SRN-41 ---> enb

SRN-42, SRN-43, SRN-44, SRN-45 are ---> ue as I see after running the 'cd radio_api && python3 scope_start --config-file radio_interactive.conf' in each of the SRNs.

{
"capture-pkts": "False",
"colosseumcli": "True",
"iperf": "True",
"users-bs": "4",
"write-config-parameters": "True",
"network-slicing": "True",
"global-scheduling-policy": "0",
"slice-scheduling-policy": "[1, 1, 1]",
"tenant-number": "4",
"slice-allocation": "{0: [0, 5], 1: [6, 11], 2: [12, 16]}",
"slice-users": "{0: [42], 1: [43, 44], 2: [45]}",
"custom-ue-slice": "True",
"force-dl-modulation": "False",
"heuristic-params": "{'buffer_thresh_bytes': [1000, 2000], 'thr_thresh_mbps': [0.25, 0.75]}",
"bs-config": "{'dl_freq': 980000000, 'ul_freq': 1020000000, 'n_prb': 50}",
"ue-config": "{'dl_freq': 980000000, 'ul_freq': 1020000000, 'force_ul_amplitude': 0.9}"

} 

I am getting data transfer problems from UEs: 42, 43, 44, 45: it says:

'Unable to connect to server'

Note:

- I ran the same configuration (as mentioned above) files for both BS and UE SRNs.

Please let me know how can I run multiple BS with multiple UEs in a specific scenario.

Thank you.

[Leonardo Bonati]

Dear Muhammad,

This is most likely due to the interference between base stations. Indeed, in the scenario that you are using, all the nodes are at the same distance from each other.

I would suggest you to either place your base stations at different frequencies (and their users, of course), or to use a scenario with some degree of separation between the nodes. As an example, the 1017 scenario clusters nodes in groups of 5, meaning that nodes 1-5 are emulated to be closer to each other than to the other nodes (and so on for nodes 6-10, 11-15, and so on). You can find more information here https://colosseumneu.freshdesk.com/a/solutions/articles/61000295793.

Regards,

Leonardo

--
You received this message because you are subscribed to the Google Groups "Colosseum Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to colosseum-use...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/colosseum-users/f7590cbe-8c9b-4688-9f66-0e9da1a76a71n%40googlegroups.com.

[MUHAMMAD ENAYETUR RAHMAN]

Dear Leonardo,

I tried to run scenario 1017 with 2 BS and 3 UE for each BS.

- I changed only the  "slice-users" parameter of 'radio_interactive.conf'

- I run the same configuration files for both BS and UE.

But still facing problem:

for BS 1:  one UE iperf3 - error: unable to unable to connect to server: network is unreachable.

for BS 2: no UEs are transmitted data. saying each one: 
one UE iperf3 - error: unable to unable to connect to server: network is unreachable.

[Leonardo Bonati]

Dear Muhammed,

Are you assigning the SRNs to nodes in different clusters when you start the scenario? Otherwise cluster 1 will have 2 base stations and cluster 2 will have 0. Please refer to the following page for instructions on how to assign SRNs to specific nodes of the scenario: https://colosseumneu.freshdesk.com/support/solutions/articles/61000253367-scenario-json-file-format

As an alternative, you can start a reservation with 10 nodes (2 BSs and 8 UEs) and the node positions should be ok as each cluster has 5 nodes.

Regards,

Leonardo

To view this discussion on the web visit https://groups.google.com/d/msgid/colosseum-users/7bd3b0ad-0617-47a3-b98f-5166db4bf436n%40googlegroups.com.

[MUHAMMAD ENAYETUR RAHMAN]

Hello, Leonardo and Colosseum users,

I have tried scenario 1018 with 10 SRNs (2 BS, 4 UEs each) and it worked well.

I have some questions based on our previous discussion and I have tried some scenarios that arose some questions as well.

1. Where the radio mapping .json file should be located? Should it be any place on any specific SRN? Should I create .json inside each of the SRNs?

2. Maybe it's a silly question: what is meant by RadioA and RadioB in the radio map file?

Tried with below scenario configuration:

===============================        1 BS with 9 UEs, Total 10 SRNs    =================================

I tried with 10 SRNs with the following configuration: [suppose: SRN-001 to SRN-010]

I tried only with the 2 slices: 0, and 1. Below is the modified configuration file (bold means modified) and run it all SRNs.

Scenario number : 1018

{
"capture-pkts": "False",
"colosseumcli": "True",
"iperf": "True",

"users-bs": "9",

"write-config-parameters": "True",
"network-slicing": "True",
"global-scheduling-policy": "0",

"slice-scheduling-policy": "[1, 1]",
"tenant-number": "9",

"slice-allocation": "{0: [0, 5], 1: [6, 11]}",

"slice-users": "{0: [2, 3, 4, 5], 1: [6, 7, 8, 9, 10]}",

"custom-ue-slice": "True",
"force-dl-modulation": "False",
"heuristic-params": "{'buffer_thresh_bytes': [1000, 2000], 'thr_thresh_mbps': [0.25, 0.75]}",
"bs-config": "{'dl_freq': 980000000, 'ul_freq': 1020000000, 'n_prb': 50}",
"ue-config": "{'dl_freq': 980000000, 'ul_freq': 1020000000, 'force_ul_amplitude': 0.9}"

} 

Observation:

 

After running the tmux a -t scope, I observed 

- data transfer with iperf for SRN-002 to SRN-005 was successful. 

- SRN-006 to SRN-010 could not connect to the server, so, no data transfer.

3. Could you please let me know why?

4. Am I doing the correct configuration?

===============================        4 BS with 9 UEs each, Total 40 SRNs     =================================

I tried with 40 SRNs with the following configuration: [suppose: SRN-001 to SRN-040]

I tried for making a network with 4 BS and 9 UEs each = Total 40 SRNs.

I tried only with the 2 slices: 0, and 1. Below is the modified configuration file (bold means modified) and run it all SRNs.

Scenario number : 1018

For each SRN-001 to SRN-010:  (I imagined, if I configure then, it will be grouped with 10, 1 BS and 9 UEs)

{
"capture-pkts": "False",
"colosseumcli": "True",
"iperf": "True",

"users-bs": "9",

"write-config-parameters": "True",
"network-slicing": "True",
"global-scheduling-policy": "0",

"slice-scheduling-policy": "[1, 1]",
"tenant-number": "9",

"slice-allocation": "{0: [0, 5], 1: [6, 11]}",

"slice-users": "{0: [2, 3, 4, 5, 6], 1: [7, 8, 9, 10]}",

"custom-ue-slice": "True",
"force-dl-modulation": "False",
"heuristic-params": "{'buffer_thresh_bytes': [1000, 2000], 'thr_thresh_mbps': [0.25, 0.75]}",
"bs-config": "{'dl_freq': 980000000, 'ul_freq': 1020000000, 'n_prb': 50}",
"ue-config": "{'dl_freq': 980000000, 'ul_freq': 1020000000, 'force_ul_amplitude': 0.9}"

} 

And I did the same way for all the groups [for example, SRN-011 to SRN-020, and so on]

Observation:

After running the tmux a -t scope, I observed:

- SRN-008, SRN-018, SRN-028, SRN-038 became the enb. Usually, each group's first SRN should be the enb, isn't it?

5. Could you please clarify for me, why it's behaving like this?

Please let me know if I am doing the correct configuration. 

With Best Regards,

Muhammad Enayetur Rahman

[Leonardo Bonati]

Dear Muhammad,

I'm glad the scenario worked as expected. Please find the answers to your questions below.

1. Where the radio mapping .json file should be located? Should it be any place on any specific SRN? Should I create .json inside each of the SRNs?

The radiomap file is passed to the Colosseum CLI command when you start the scenario, you only need it on the node from which you start the scenario (more info here: https://colosseumneu.freshdesk.com/support/solutions/articles/61000253397-colosseum-cli).

 

2. Maybe it's a silly question: what is meant by RadioA and RadioB in the radio map file?

Radio A and B are the two daughtercards of the USRP X310 (more info here: https://colosseumneu.freshdesk.com/support/solutions/articles/61000253367-scenario-json-file-format).

Did you assign users to the slices you are creating? This should be specified in the radio_code/scope_config/slicing/ue_imsi_slice.txt file (more info here: https://github.com/wineslab/colosseum-scope)

This depends on the SRNs you were assigned, my guess is that SRN 8 was your first SRN, 18 the second and so on.

Regards,

Leonardo

[MUHAMMAD ENAYETUR RAHMAN]

Dear Leonardo,

Thank you for answering my queries. 

For my below requirements (2 slices, 9 users, 1 BS), what files do I need to change?

No, I did not. I only modified the radio_interactive.conf file.

Question: For my above requirements (2 slices, 10 users, 1 BS), what files do I need to change?

Best Regards,

Muhammad 

[Leonardo Bonati]

Hi Muhammad,

The default configuration assumes you have users spread over 3 slices. It should be fine to use 2 slices only (I was thinking for a moment that you were configuring the whole thing manually instead of using the startup script). Can you verify the users have been assigned to the right slices as specified in your configuration file? (For instance in the file ue_imsi_slice.txt after running the startup script.)

The configuration you are passing to the script looks fine to me. As a side note, you don't need to modify "tenant-number" as this is the maximum number of slices supported, and not of users.

Regards,

Leonardo

--
You received this message because you are subscribed to the Google Groups "Colosseum Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to colosseum-use...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/colosseum-users/73649080-5f04-41dd-b7b4-eb0ddf778292n%40googlegroups.com.

[MUHAMMAD ENAYETUR RAHMAN]

Hi Leonardo,

For the 1 BS, 9 UEs I am facing the same issue, first 4 UEs are connected to BS, but other 5 UEs couldn't.

 

The default configuration assumes you have users spread over 3 slices. It should be fine to use 2 slices only (I was thinking for a moment that you were configuring the whole thing manually instead of using the startup script). Can you verify the users have been assigned to the right slices as specified in your configuration file? (For instance in the file ue_imsi_slice.txt after running the startup script.)

I checked BS's ue_imsi_slice.txt file after running the startup script, and found below change (I guess, it is IMSI:: SLICE NO.):

001010123456001::0
001010123456002::2    --->    001010123456002::0
001010123456003::0
001010123456004::1   ---->  001010123456004::0
001010123456005::0

if it is (I guess, it is IMSI:: SLICE NO.), then acted according to configuration file. 

But other SRN's from 6-10 are not connected, so maybe did not updated at BS's ue_imsi_slice.txt . 

Questions:

1. Please let me know if I am guessing the correct way or, let me know how it works?

2. Why the other SRN 6 -10 are not connected? Is it because of the clustering as you mentioned before for scenario 1018?

The configuration you are passing to the script looks fine to me. As a side note, you don't need to modify "tenant-number" as this is the maximum number of slices supported, and not of users.

I did not modify this one now.

Thanks

---Muhammad

[Leonardo Bonati]

Dear Muhammad,

The configuration (imsi:slice) looks good. Regarding the clustering, you should have a base station (and only one) on each cluster, otherwise they may interfere with one another. Additionally, also try with different reservations (which will probably assign you different SRNs) in case some of the USRPs are having hardware issues.

Regards,

Leonardo

--
You received this message because you are subscribed to the Google Groups "Colosseum Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to colosseum-use...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/colosseum-users/26106c80-8d3d-4475-92cd-a16bcf8d82d0n%40googlegroups.com.

[MUHAMMAD ENAYETUR RAHMAN]

Hi Leonardo,

I tried for scenario: 1018, 1 BS, 7 UEs with the following modified radio_interactive.conf:

{
"capture-pkts": "False",
"colosseumcli": "True",
"iperf": "True",

"users-bs": "7",

"write-config-parameters": "True",
"network-slicing": "True",
"global-scheduling-policy": "0",

"slice-scheduling-policy": "[1, 1, 1]",
"tenant-number": "3",
"slice-allocation": "{0: [0, 5], 1: [6, 11], 2: [12, 16]}",

"slice-users": "{0: [1, 2], 1: [3, 4], 2: [5, 6, 7]}",

"custom-ue-slice": "True",
"force-dl-modulation": "False",
"heuristic-params": "{'buffer_thresh_bytes': [1000, 2000], 'thr_thresh_mbps': [0.25, 0.75]}",
"bs-config": "{'dl_freq': 980000000, 'ul_freq': 1020000000, 'n_prb': 50}",
"ue-config": "{'dl_freq': 980000000, 'ul_freq': 1020000000, 'force_ul_amplitude': 0.9}"

} 

My radio-mapping file: 

{
  "Node 1":{"SRN":1,"RadioA":1,"RadioB":2},
  "Node 2":{"SRN":2,"RadioA":1,"RadioB":2},
  "Node 3":{"SRN":3,"RadioA":1,"RadioB":2},
  "Node 4":{"SRN":4,"RadioA":1,"RadioB":2},
  "Node 5":{"SRN":5,"RadioA":1,"RadioB":2},
  "Node 6":{"SRN":6,"RadioA":1,"RadioB":2},
  "Node 7":{"SRN":7,"RadioA":1,"RadioB":2},
  "Node 8":{"SRN":8,"RadioA":1,"RadioB":2}
}

Result: But none of them are connected to the BS.

Questions

1. I don't understand how to configure properly so that, I can run multiple BS, and multiple slices?

There is very limited documentation which is not also clear to me.

2. As you suggested previously, I can place Base stations at different frequencies?

How can I do that?

Please let me know.

Thank you.

---Muhammad

[MUHAMMAD ENAYETUR RAHMAN]

Sorry, my radio-mapping file is:

{
  "Node 1":{"SRN":1,"RadioA":1,"RadioB":2},

  "Node 2":{"SRN":2,"RadioA":2,"RadioB":2},
  "Node 3":{"SRN":3,"RadioA":3,"RadioB":2},
  "Node 4":{"SRN":4,"RadioA":4,"RadioB":2},
  "Node 5":{"SRN":5,"RadioA":5,"RadioB":2},
  "Node 6":{"SRN":6,"RadioA":6,"RadioB":2},
  "Node 7":{"SRN":7,"RadioA":7,"RadioB":2},
  "Node 8":{"SRN":8,"RadioA":8,"RadioB":2}
}

[MUHAMMAD ENAYETUR RAHMAN]

Hello Leonardo,

Could you please let me know the question I have? 

How can I place BS at different frequencies? 

[Leonardo Bonati]

Dear Muhammad,

You can modify the following configuration option to set different downlink and uplink frequencies: "bs-config": "{'dl_freq': 980000000, 'ul_freq': 1020000000, 'n_prb': 50}".

As an alternative method, you can manually modify the same configuration options here: https://github.com/wineslab/colosseum-scope/blob/b7176a5c334ffbb9f3bbeb61a6e805c58649fddb/radio_code/srslte_config/enb.conf#L62-L64

Regards,

Leonardo

To view this discussion on the web visit https://groups.google.com/d/msgid/colosseum-users/76ac1330-f0b3-4671-935c-0004fb17ca67n%40googlegroups.com.

[MUHAMMAD ENAYETUR RAHMAN]

Hello Leonardo,

Thank you for your answer. I was trying to place BS at different frequencies as you suggested before.

scenario: 1009 with 10 SRNs with the following:

{
"capture-pkts": "True",

"colosseumcli": "True",
"iperf": "True",

"users-bs": "3",

"write-config-parameters": "True",
"network-slicing": "True",
"global-scheduling-policy": "0",

"slice-scheduling-policy": "[1, 1, 1]",
"tenant-number": "3",

"slice-allocation": "{0: [0, 5], 1: [6, 11], 2:[12, 16]}",

"slice-users": "{0: [6], 1: [7], 2:[8]}",

"custom-ue-slice": "True",
"force-dl-modulation": "False",
"heuristic-params": "{'buffer_thresh_bytes': [1000, 2000], 'thr_thresh_mbps': [0.25, 0.75]}",

"bs-config": "{'dl_freq': 500000000, 'ul_freq': 540000000, 'n_prb': 50}",
"ue-config": "{'dl_freq': 500000000, 'ul_freq': 540000000, 'force_ul_amplitude': 0.9}"

} 

The first 3 UEs were connected with the default configuration. For other UEs and BS I modified as above.

For all other UEs it did not work, at UE says:

iperf3 error: unable to connect to server, network is unreachable.

Do I need to change any other fields if I want to place BS at different frequency? 

[Leonardo Bonati]

Dear Muhammad,

You can change the frequencies (as you are doing) but they still need to be tuned at the frequencies supported by the scenario you are using. I.e., your scenario has a central frequency of 1 GHz and you are starting your base station/users at 500 MHz. Additionally the bandwidth supported by MCHEM is 80 MHz (i.e., 40 MHz on each side of the central frequency of the scenario). For more detailed information, please refer to the Colosseum guides that you can find on FreshDesk.

Regards,

Leonardo

To view this discussion on the web visit https://groups.google.com/d/msgid/colosseum-users/33bbfe5f-4ae0-4c66-99fd-e3ab494e91ban%40googlegroups.com.

[MUHAMMAD ENAYETUR RAHMAN]

Dear Leonardo,

I did not find this information on the Colosseum Freshdesk. There is very limited information about conducting or, modifying an experiment.