[srslte-users] pdsch_enodeb > file and synch_file example

Ismael Gomez ismael.gomez at softwareradiosystems.com
Mon May 16 21:43:11 UTC 2016


Hi Douglas,

The problem could be that the DFT size for the 5 MHz signal is 384 instead
of 512. See srslte/lib/common/phy_common.c function srslte_symbol_sz() to
see the DFT sizes when 3/4 sampling is enabled (enabled by default).

So try again this synch_file program with FFTLEN=384 and SLOTLEN=15*384

cheers



On Mon, 16 May 2016 at 23:07 Anderson, Douglas J. <danderson at its.bldrdoc.gov>
wrote:

> Hi,
>
> I've been trying to figure out why I can't get any srslte library functions
> to work on output from pdsch_enodeb for greater than 6 PRBs.
>
> For example here I create 100 frames at 6 PRBs, and synch_file detects
> all 100 frames (if I reduce the threshold). When I try the exact same
> thing with 25 PRBs (or any other bandwidth), I get nothing.
>
> Peaking around with GDB, I can see that the peak to side-lobe ratio
> returned by the pss_synch function is ~1 for all N_id_2 above 6 PRB.
>
> I'm concerned that it has something to do with the output of
> pdsch_enodeb, since I get the same results (no PSS correlation) even if I
> resample the output file down to 1.92 MHz before running synch_file.
>
> Here's an example of what I'm talking about, running on srslte HEAD:
>
> $ ./pdsch_enodeb -o 6prb -p 6 -n 100 -c 123
> linux; GNU C++ version 5.3.1 20160413; Boost_105800;
> UHD_003.010.git-202-g9e0861e1
>
>  - Resource Allocation Type:        Type 0
>    + Resource Block Group Size:        1
>    + RBG Bitmap:            0x3f
>  - Modulation and coding scheme index:    1
>  - HARQ process:            0
>  - New data indicator:            No
>  - Redundancy version:            0
>  - TPC command for PUCCH:        --
>  - PRB Bitmap Assignment 0st slot:
> 0, 1, 2, 3, 4, 5,
>  - PRB Bitmap Assignment 1st slot:
> 0, 1, 2, 3, 4, 5,
>  - Number of PRBs:            6
>  - Modulation type:            QPSK
>  - Transport block size:        208
> Type new MCS index and press Enter: Using Volk machine: avx2_64_mmx
> Done
>
> $ FFTLEN=128; SLOTLEN=960; FRAMELEN=$((10*SLOTLEN)); ./synch_file -l
> $FRAMELEN -s $FFTLEN -t 5 -i 6prb
> linux; GNU C++ version 5.3.1 20160413; Boost_105800;
> UHD_003.010.git-202-g9e0861e1
> Initializing...Using Volk machine: avx2_64_mmx
> done in 0 s 638 ms
>
>     Fr.Cnt    N_id_2    N_id_1    Subf    PSS Peak/Avg    Idx    m0
> m1    CFO
>
> ===============================================================================
>     1    0    41    5    6.413        960    13    11    -0.000
>     2    0    41    0    7.155        960    11    13    -0.000
> [...]
>     98    0    41    0    9.678        960    11    13    -0.000
>     99    0    41    5    8.089        960    13    11    -0.000
>
> Average exec time: 1.553 ms / frame. 6.182 Msamp/s (31.056% CPU)
> Average CFO: -0.000
> Done
>
> $ ./pdsch_enodeb -o 25prb -p 25 -n 100 -c 123
> linux; GNU C++ version 5.3.1 20160413; Boost_105800;
> UHD_003.010.git-202-g9e0861e1
>
>  - Resource Allocation Type:        Type 0
>    + Resource Block Group Size:        2
>    + RBG Bitmap:            0x1fff
>  - Modulation and coding scheme index:    1
>  - HARQ process:            0
>  - New data indicator:            No
>  - Redundancy version:            0
>  - TPC command for PUCCH:        --
>  - PRB Bitmap Assignment 0st slot:
> 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20,
> 21, 22, 23, 24,
>  - PRB Bitmap Assignment 1st slot:
> 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20,
> 21, 22, 23, 24,
>  - Number of PRBs:            25
>  - Modulation type:            QPSK
>  - Transport block size:        904
> Type new MCS index and press Enter: Using Volk machine: avx2_64_mmx
> Done
>
> $ FFTLEN=512; SLOTLEN=3840; FRAMELEN=$((10*SLOTLEN)); ./synch_file -l
> $FRAMELEN -s $FFTLEN -t 5 -i 25prb
> linux; GNU C++ version 5.3.1 20160413; Boost_105800;
> UHD_003.010.git-202-g9e0861e1
>
> Initializing...Using Volk machine: avx2_64_mmx
> done in 2 s 709 ms
>
>     Fr.Cnt    N_id_2    N_id_1    Subf    PSS Peak/Avg    Idx    m0
> m1    CFO
>
> ===============================================================================
>
> Average exec time: 9.369 ms / frame. 4.098 Msamp/s (46.847% CPU)
> Average CFO: 0.000
> Done
> _______________________________________________
> srslte-users mailing list
> srslte-users at lists.softwareradiosystems.com
> http://www.softwareradiosystems.com/mailman/listinfo/srslte-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.srsran.com/pipermail/srsran-users/attachments/20160516/92e03a0e/attachment.htm>


More information about the srsran-users mailing list