Empathy List Archives

S

skyung@nrel.gov

Mon, Jul 25, 2022 6:23 PM

Hi, I am using USRP N310 connected via a switch with 10Gbps interface.

I am using it to set up OAI 5G RAN but I keep getting the following message:

[HW] reived 16384 samples out of 23040

[HW] ERROR_CODE_TIMEOUT (or ERROR_CODE_OVERFLOW)

I wonder if it is a sync problem or something else.

Has anyone had this problem? What is the cause of this problem and how do I resolve the issue?

Hi, I am using USRP N310 connected via a switch with 10Gbps interface. I am using it to set up OAI 5G RAN but I keep getting the following message: \[HW\] reived 16384 samples out of 23040 \[HW\] ERROR_CODE_TIMEOUT (or ERROR_CODE_OVERFLOW) I wonder if it is a sync problem or something else. Has anyone had this problem? What is the cause of this problem and how do I resolve the issue?

MD

Marcus D. Leech

Mon, Jul 25, 2022 6:27 PM

On 2022-07-25 14:23, skyung--- via USRP-users wrote:

Hi, I am using USRP N310 connected via a switch with 10Gbps interface.

I am using it to set up OAI 5G RAN but I keep getting the following
message:

[HW] reived 16384 samples out of 23040

[HW] ERROR_CODE_TIMEOUT (or ERROR_CODE_OVERFLOW)

I wonder if it is a sync problem or something else.

Has anyone had this problem? What is the cause of this problem and how
do I resolve the issue?

USRP-users mailing list -- usrp-users@lists.ettus.com
To unsubscribe send an email to usrp-users-leave@lists.ettus.com

My guess is the switch is dropping frames.

You could use the "rx_samples_to_file" example, and use "/dev/null" as
the file, and do this at the same sample rate as your application. Does
this produce TIMEOUT errors?

On 2022-07-25 14:23, skyung--- via USRP-users wrote: > > Hi, I am using USRP N310 connected via a switch with 10Gbps interface. > > I am using it to set up OAI 5G RAN but I keep getting the following > message: > > > [HW] reived 16384 samples out of 23040 > > [HW] ERROR_CODE_TIMEOUT (or ERROR_CODE_OVERFLOW) > > > I wonder if it is a sync problem or something else. > > Has anyone had this problem? What is the cause of this problem and how > do I resolve the issue? > > > _______________________________________________ > USRP-users mailing list -- usrp-users@lists.ettus.com > To unsubscribe send an email to usrp-users-leave@lists.ettus.com My guess is the switch is dropping frames. You could use the "rx_samples_to_file" example, and use "/dev/null" as the file, and do this at the same sample rate as your application. Does this produce TIMEOUT errors?

S

skyung@nrel.gov

Mon, Jul 25, 2022 8:13 PM

Hi Marcus,

Thanks for your response.

I did run the “rx_sample_to_file” example but did not get any timeout error.

The following is the part of output I get when I run the OAI UE:

Found USRP n310

[HW] device_init() sample_rate:46080000

[HW] cal 0: freq 3500000000.000000, offset 0.000000, diff 119200000.000000

[HW] cal 1: freq 2660000000.000000, offset 0.000000, diff 959200000.000000

[HW] cal 2: freq 2300000000.000000, offset 0.000000, diff 1319200000.000000

[HW] cal 3: freq 1880000000.000000, offset 0.000000, diff 1739200000.000000

[HW] cal 4: freq 816000000.000000, offset 0.000000, diff 2803200000.000000

[HW] RX Gain too high, lower by 35.000000 dB

[HW] RX Gain 0 110.000000 (0.000000) => 110.000000 (max 75.000000)

[HW] USRP TX_GAIN:65.00 gain_range:65.00 tx_gain:0.00

[HW] Actual master clock: 122.880000MHz...

[HW] Actual clock source internal...

[HW] Actual time source internal...

[HW] RF board max packet size 364, size for 100µs jitter 4608

[HW] rx_max_num_samps 364

[HW] setting rx channel 0

[HW] RX Channel 0

[HW] Actual RX sample rate: 40.960000MSps...

[HW] Actual RX frequency: 3.619200GHz...

[HW] Actual RX gain: 75.000000...

[HW] Actual RX bandwidth: 100.000000M...

[HW] Actual RX antenna: RX2...

[HW] TX Channel 0

[HW] Actual TX sample rate: 40.960000MSps...

[HW] Actual TX frequency: 3.619200GHz...

[HW] Actual TX gain: 65.000000...

[HW] Actual TX bandwidth: 100.000000M...

[HW] Actual TX antenna: TX/RX...

[HW] Actual TX packet size: 364

[HW] Device timestamp: 1.283773...

[HW] [RRU] has loaded USRP N300 device.

[PHY] [SCHED][UE] Check absolute frequency DL 3619200000.000000, UL 3619200000.000000 (RF card 0, oai_exit 0, channel 0, rx_num_channels 1)

[PHY] Starting sync detection

[PHY] [UE thread Synch] Running Initial Synch (mode 6)

[PHY] [UE] nr_synchro_time: Sync source = 1, Peak found at pos 434940, val = 1998314 (63 dB) avg 61 dB, ffo 0.003443

[PHY] [UE0] Initial sync : Estimated PSS position -1, Nid2 1

[PHY] sync_pos -1 ssb_offset 460691

[HW] [recv] received 21212 samples out of 23040

[HW] ERROR_CODE_TIMEOUT

Hi Marcus, Thanks for your response. I did run the “rx_sample_to_file” example but did not get any timeout error. The following is the part of output I get when I run the OAI UE: Found USRP n310 \[HW\] device_init() sample_rate:46080000 \[HW\] cal 0: freq 3500000000.000000, offset 0.000000, diff 119200000.000000 \[HW\] cal 1: freq 2660000000.000000, offset 0.000000, diff 959200000.000000 \[HW\] cal 2: freq 2300000000.000000, offset 0.000000, diff 1319200000.000000 \[HW\] cal 3: freq 1880000000.000000, offset 0.000000, diff 1739200000.000000 \[HW\] cal 4: freq 816000000.000000, offset 0.000000, diff 2803200000.000000 \[HW\] RX Gain too high, lower by 35.000000 dB \[HW\] RX Gain 0 110.000000 (0.000000) => 110.000000 (max 75.000000) \[HW\] USRP TX_GAIN:65.00 gain_range:65.00 tx_gain:0.00 \[HW\] Actual master clock: 122.880000MHz... \[HW\] Actual clock source internal... \[HW\] Actual time source internal... \[HW\] RF board max packet size 364, size for 100µs jitter 4608 \[HW\] rx_max_num_samps 364 \[HW\] setting rx channel 0 \[HW\] RX Channel 0 \[HW\] Actual RX sample rate: 40.960000MSps... \[HW\] Actual RX frequency: 3.619200GHz... \[HW\] Actual RX gain: 75.000000... \[HW\] Actual RX bandwidth: 100.000000M... \[HW\] Actual RX antenna: RX2... \[HW\] TX Channel 0 \[HW\] Actual TX sample rate: 40.960000MSps... \[HW\] Actual TX frequency: 3.619200GHz... \[HW\] Actual TX gain: 65.000000... \[HW\] Actual TX bandwidth: 100.000000M... \[HW\] Actual TX antenna: TX/RX... \[HW\] Actual TX packet size: 364 \[HW\] Device timestamp: 1.283773... \[HW\] \[RRU\] has loaded USRP N300 device. \[PHY\] \[SCHED\]\[UE\] Check absolute frequency DL 3619200000.000000, UL 3619200000.000000 (RF card 0, oai_exit 0, channel 0, rx_num_channels 1) \[PHY\] Starting sync detection \[PHY\] \[UE thread Synch\] Running Initial Synch (mode 6) \[PHY\] \[UE\] nr_synchro_time: Sync source = 1, Peak found at pos 434940, val = 1998314 (63 dB) avg 61 dB, ffo 0.003443 \[PHY\] \[UE0\] Initial sync : Estimated PSS position -1, Nid2 1 \[PHY\] sync_pos -1 ssb_offset 460691 \[HW\] \[recv\] received 21212 samples out of 23040 \[HW\] ERROR_CODE_TIMEOUT

CH

Cédric Hannotier

Tue, Jul 26, 2022 9:22 AM

Hi skyung,

On 2022-07-25 20:13 +0000, skyung--- via USRP-users wrote:

I did run the “rx_sample_to_file” example but did not get any timeout error.

The following is the part of output I get when I run the OAI UE:

[...]

[HW] RF board max packet size 364, size for 100µs jitter 4608

I never worked with N310 nor OAI UE,
but 364 spp seems quite small.
Can this be the reason for the TIMEOUT/OVERFLOW?
What is your NIC MTU?

Regards

Cédric Hannotier

Hi skyung, On 2022-07-25 20:13 +0000, skyung--- via USRP-users wrote: > I did run the “rx_sample_to_file” example but did not get any timeout error. > > The following is the part of output I get when I run the OAI UE: [...] > \[HW\] RF board max packet size 364, size for 100µs jitter 4608 I never worked with N310 nor OAI UE, but 364 spp seems quite small. Can this be the reason for the TIMEOUT/OVERFLOW? What is your NIC MTU? Regards -- Cédric Hannotier

S

skyung@nrel.gov

Tue, Jul 26, 2022 7:51 PM

Hi Cedric,

Thanks for your response.

You are right, the packet size is way too small.. I managed to increase it to 2000 and confirmed that the MTU on interfaces of both USRP and host was already set to 9000.

However, it still give me the “ERROR_CODE_OVERFLOW (Out of sequence error)”.

Currently, the USRP is connected to the host via switch. Do you think I need to configure the switch in any specific way?

Hi Cedric, Thanks for your response. You are right, the packet size is way too small.. I managed to increase it to 2000 and confirmed that the MTU on interfaces of both USRP and host was already set to 9000. However, it still give me the “ERROR_CODE_OVERFLOW (Out of sequence error)”. Currently, the USRP is connected to the host via switch. Do you think I need to configure the switch in any specific way?

CH

Cédric Hannotier

Wed, Jul 27, 2022 8:12 AM

On 2022-07-26 19:51 +0000, skyung--- via USRP-users wrote:

However, it still give me the “ERROR_CODE_OVERFLOW (Out of sequence
error)”.

Currently, the USRP is connected to the host via switch. Do you think
I need to configure the switch in any specific way?

If you have a configurable switch, maybe...
IMO, the best way to find out if the switch is at fault is to remove it,
i.e. trying with a direct connection between the host and the USRP.

Regards

Cédric Hannotier

On 2022-07-26 19:51 +0000, skyung--- via USRP-users wrote: > However, it still give me the “ERROR_CODE_OVERFLOW (Out of sequence > error)”. > > Currently, the USRP is connected to the host via switch. Do you think > I need to configure the switch in any specific way? If you have a configurable switch, maybe... IMO, the best way to find out if the switch is at fault is to remove it, i.e. trying with a direct connection between the host and the USRP. Regards -- Cédric Hannotier

usrp-users@lists.ettus.com

Getting low number of samples (USRP N310)

Regards

Regards