Skip to content
Projects
Groups
Snippets
Help
Loading...
Help
Support
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in / Register
Toggle navigation
O
OpenXG-RAN
Project overview
Project overview
Details
Activity
Releases
Repository
Repository
Files
Commits
Branches
Tags
Contributors
Graph
Compare
Issues
0
Issues
0
List
Boards
Labels
Milestones
Merge Requests
0
Merge Requests
0
CI / CD
CI / CD
Pipelines
Jobs
Schedules
Analytics
Analytics
CI / CD
Repository
Value Stream
Wiki
Wiki
Snippets
Snippets
Members
Members
Collapse sidebar
Close sidebar
Activity
Graph
Create a new issue
Jobs
Commits
Issue Boards
Open sidebar
常顺宇
OpenXG-RAN
Commits
1d765027
Commit
1d765027
authored
Jul 06, 2018
by
Elena Lukashova
Browse files
Options
Browse Files
Download
Email Patches
Plain Diff
Extracting SISO llr computations in separate fucntions
qpsk_llr, qam16_llr, qam64_llqam64.
parent
0dfcff75
Changes
2
Hide whitespace changes
Inline
Side-by-side
Showing
2 changed files
with
165 additions
and
113 deletions
+165
-113
openair1/PHY/LTE_UE_TRANSPORT/dlsch_llr_computation.c
openair1/PHY/LTE_UE_TRANSPORT/dlsch_llr_computation.c
+149
-113
openair1/PHY/LTE_UE_TRANSPORT/transport_proto_ue.h
openair1/PHY/LTE_UE_TRANSPORT/transport_proto_ue.h
+16
-0
No files found.
openair1/PHY/LTE_UE_TRANSPORT/dlsch_llr_computation.c
View file @
1d765027
...
...
@@ -681,19 +681,28 @@ int dlsch_qpsk_llr(LTE_DL_FRAME_PARMS *frame_parms,
dlsch_llr,
llr32);
*/
//printf("ll32p=%p , dlsch_llr=%p, symbol=%d, flag=%d \n", llr32, dlsch_llr, symbol, first_symbol_flag);
for
(
i
=
0
;
i
<
len
;
i
++
)
{
*
llr32
=
*
rxF
;
//printf("llr %d : (%d,%d)\n",i,((int16_t*)llr32)[0],((int16_t*)llr32)[1]);
rxF
++
;
llr32
++
;
}
//*llr32p = (int16_t *)llr32;
qpsk_llr
((
short
*
)
rxF
,
(
short
*
)
llr32
,
len
);
return
(
0
);
}
void
qpsk_llr
(
int16_t
*
stream0_in
,
int16_t
*
stream0_out
,
int
length
)
{
int
i
;
for
(
i
=
0
;
i
<
2
*
length
;
i
++
)
{
*
stream0_out
=
*
stream0_in
;
//printf("llr %d : (%d,%d)\n",i,((int16_t*)stream0_out)[0],((int16_t*)stream0_out)[1]);
stream0_in
++
;
stream0_out
++
;
}
}
int32_t
dlsch_qpsk_llr_SIC
(
LTE_DL_FRAME_PARMS
*
frame_parms
,
int32_t
**
rxdataF_comp
,
int32_t
**
sic_buffer
,
//Q15
...
...
@@ -817,45 +826,21 @@ void dlsch_16qam_llr(LTE_DL_FRAME_PARMS *frame_parms,
int16_t
**
llr32p
,
uint8_t
beamforming_mode
)
{
int32_t
*
rxF
=
(
int32_t
*
)
&
rxdataF_comp
[
0
][(
symbol
*
frame_parms
->
N_RB_DL
*
12
)];
int32_t
*
ch_mag
=
(
int32_t
*
)
&
dl_ch_mag
[
0
][(
symbol
*
frame_parms
->
N_RB_DL
*
12
)];
int32_t
*
llr32
;
#if defined(__x86_64__) || defined(__i386__)
__m128i
*
rxF
=
(
__m128i
*
)
&
rxdataF_comp
[
0
][(
symbol
*
frame_parms
->
N_RB_DL
*
12
)];
__m128i
*
ch_mag
;
__m128i
llr128
[
2
];
uint32_t
*
llr32
;
#elif defined(__arm__)
int16x8_t
*
rxF
=
(
int16x8_t
*
)
&
rxdataF_comp
[
0
][(
symbol
*
frame_parms
->
N_RB_DL
*
12
)];
int16x8_t
*
ch_mag
;
int16x8_t
xmm0
;
int16_t
*
llr16
;
#endif
int
i
,
len
;
int
len
;
unsigned
char
symbol_mod
,
len_mod4
=
0
;
#if defined(__x86_64__) || defined(__i386__)
if
(
first_symbol_flag
==
1
)
{
llr32
=
(
uint32_t
*
)
dlsch_llr
;
}
else
{
llr32
=
(
uint32_t
*
)
*
llr32p
;
}
#elif defined(__arm__)
if
(
first_symbol_flag
==
1
)
{
llr16
=
(
int16_t
*
)
dlsch_llr
;
llr32
=
(
int32_t
*
)
dlsch_llr
;
}
else
{
llr
16
=
(
int16
_t
*
)
*
llr32p
;
llr
32
=
(
int32
_t
*
)
*
llr32p
;
}
#endif
symbol_mod
=
(
symbol
>=
(
7
-
frame_parms
->
Ncp
))
?
symbol
-
(
7
-
frame_parms
->
Ncp
)
:
symbol
;
#if defined(__x86_64__) || defined(__i386__)
ch_mag
=
(
__m128i
*
)
&
dl_ch_mag
[
0
][(
symbol
*
frame_parms
->
N_RB_DL
*
12
)];
#elif defined(__arm__)
ch_mag
=
(
int16x8_t
*
)
&
dl_ch_mag
[
0
][(
symbol
*
frame_parms
->
N_RB_DL
*
12
)];
#endif
if
((
symbol_mod
==
0
)
||
(
symbol_mod
==
(
4
-
frame_parms
->
Ncp
)))
{
if
(
frame_parms
->
nb_antenna_ports_eNB
!=
1
)
len
=
(
nb_rb
*
8
)
-
(
2
*
pbch_pss_sss_adjust
/
3
);
...
...
@@ -881,18 +866,45 @@ void dlsch_16qam_llr(LTE_DL_FRAME_PARMS *frame_parms,
len
>>=
2
;
// length in quad words (4 REs)
// printf("len>>=2=%d\n", len);
len
+=
(
len_mod4
==
0
?
0
:
1
);
// printf("len+=%d\n", len);
for
(
i
=
0
;
i
<
len
;
i
++
)
{
qam16_llr
((
short
*
)
rxF
,
(
short
*
)
ch_mag
,
(
short
*
)
llr32
,
len
);
// printf ("This line in qam16_llr is %d.\n", __LINE__);
}
void
qam16_llr
(
int16_t
*
stream0_in
,
int16_t
*
chan_magn
,
int16_t
*
llr
,
int
length
)
{
int
i
;
#if defined(__x86_64__) || defined(__i386__)
__m128i
*
rxF_128
=
(
__m128i
*
)
stream0_in
;
__m128i
*
ch_mag_128
=
(
__m128i
*
)
chan_magn
;
__m128i
llr128
[
2
];
int32_t
*
llr32
=
(
int32_t
*
)
llr
;
#elif defined(__arm__)
int16x8_t
*
rxF_128
=
(
int16x8_t
*
)
stream0_in
;
int16x8_t
*
ch_mag_128
=
(
int16x8_t
*
)
chan_magn
;
int16x8_t
xmm0
;
int16_t
*
llr16
=
(
int16_t
*
)
llr
;
#endif
// printf ("This line in qam16_llr is %d.\n", __LINE__);
for
(
i
=
0
;
i
<
length
;
i
++
)
{
#if defined(__x86_64__) || defined(__i386)
xmm0
=
_mm_abs_epi16
(
rxF
[
i
]);
xmm0
=
_mm_subs_epi16
(
ch_mag
[
i
],
xmm0
);
xmm0
=
_mm_abs_epi16
(
rxF_128
[
i
]);
xmm0
=
_mm_subs_epi16
(
ch_mag_128
[
i
],
xmm0
);
// lambda_1=y_R, lambda_2=|y_R|-|h|^2, lamda_3=y_I, lambda_4=|y_I|-|h|^2
llr128
[
0
]
=
_mm_unpacklo_epi32
(
rxF
[
i
],
xmm0
);
llr128
[
1
]
=
_mm_unpackhi_epi32
(
rxF
[
i
],
xmm0
);
llr128
[
0
]
=
_mm_unpacklo_epi32
(
rxF
_128
[
i
],
xmm0
);
llr128
[
1
]
=
_mm_unpackhi_epi32
(
rxF
_128
[
i
],
xmm0
);
llr32
[
0
]
=
_mm_extract_epi32
(
llr128
[
0
],
0
);
//((uint32_t *)&llr128[0])[0];
llr32
[
1
]
=
_mm_extract_epi32
(
llr128
[
0
],
1
);
//((uint32_t *)&llr128[0])[
1
];
llr32
[
1
]
=
_mm_extract_epi32
(
llr128
[
0
],
1
);
//((uint32_t *)&llr128[0])[
0
];
llr32
[
2
]
=
_mm_extract_epi32
(
llr128
[
0
],
2
);
//((uint32_t *)&llr128[0])[2];
llr32
[
3
]
=
_mm_extract_epi32
(
llr128
[
0
],
3
);
//((uint32_t *)&llr128[0])[3];
llr32
[
4
]
=
_mm_extract_epi32
(
llr128
[
1
],
0
);
//((uint32_t *)&llr128[1])[0];
...
...
@@ -922,14 +934,17 @@ void dlsch_16qam_llr(LTE_DL_FRAME_PARMS *frame_parms,
llr16
[
14
]
=
vgetq_lane_s16
(
xmm0
,
7
);
llr16
[
15
]
=
vgetq_lane_s16
(
xmm0
,
7
);
llr16
+=
16
;
#endif
}
#if defined(__x86_64__) || defined(__i386
__
)
#if defined(__x86_64__) || defined(__i386)
_mm_empty
();
_m_empty
();
#endif
}
void
dlsch_16qam_llr_SIC
(
LTE_DL_FRAME_PARMS
*
frame_parms
,
...
...
@@ -1036,6 +1051,7 @@ void dlsch_16qam_llr_SIC (LTE_DL_FRAME_PARMS *frame_parms,
}
}
//----------------------------------------------------------------------------------------------
// 64-QAM
//----------------------------------------------------------------------------------------------
...
...
@@ -1053,14 +1069,10 @@ void dlsch_64qam_llr(LTE_DL_FRAME_PARMS *frame_parms,
uint32_t
llr_offset
,
uint8_t
beamforming_mode
)
{
#if defined(__x86_64__) || defined(__i386__)
__m128i
*
rxF
=
(
__m128i
*
)
&
rxdataF_comp
[
0
][(
symbol
*
frame_parms
->
N_RB_DL
*
12
)];
__m128i
*
ch_mag
,
*
ch_magb
;
#elif defined(__arm__)
int16x8_t
*
rxF
=
(
int16x8_t
*
)
&
rxdataF_comp
[
0
][(
symbol
*
frame_parms
->
N_RB_DL
*
12
)];
int16x8_t
*
ch_mag
,
*
ch_magb
,
xmm1
,
xmm2
;
#endif
int
i
,
len
,
len2
;
int32_t
*
rxF
=
(
int32_t
*
)
&
rxdataF_comp
[
0
][(
symbol
*
frame_parms
->
N_RB_DL
*
12
)];
int32_t
*
ch_mag
=
(
int32_t
*
)
&
dl_ch_mag
[
0
][(
symbol
*
frame_parms
->
N_RB_DL
*
12
)];
int32_t
*
ch_magb
=
(
int32_t
*
)
&
dl_ch_magb
[
0
][(
symbol
*
frame_parms
->
N_RB_DL
*
12
)];
int
len
,
len2
;
unsigned
char
symbol_mod
,
len_mod4
;
short
*
llr
;
int16_t
*
llr2
;
...
...
@@ -1079,13 +1091,6 @@ void dlsch_64qam_llr(LTE_DL_FRAME_PARMS *frame_parms,
symbol_mod
=
(
symbol
>=
(
7
-
frame_parms
->
Ncp
))
?
symbol
-
(
7
-
frame_parms
->
Ncp
)
:
symbol
;
#if defined(__x86_64__) || defined(__i386__)
ch_mag
=
(
__m128i
*
)
&
dl_ch_mag
[
0
][(
symbol
*
frame_parms
->
N_RB_DL
*
12
)];
ch_magb
=
(
__m128i
*
)
&
dl_ch_magb
[
0
][(
symbol
*
frame_parms
->
N_RB_DL
*
12
)];
#elif defined(__arm__)
ch_mag
=
(
int16x8_t
*
)
&
dl_ch_mag
[
0
][(
symbol
*
frame_parms
->
N_RB_DL
*
12
)];
ch_magb
=
(
int16x8_t
*
)
&
dl_ch_magb
[
0
][(
symbol
*
frame_parms
->
N_RB_DL
*
12
)];
#endif
if
((
symbol_mod
==
0
)
||
(
symbol_mod
==
(
4
-
frame_parms
->
Ncp
)))
{
if
(
frame_parms
->
nb_antenna_ports_eNB
!=
1
)
len
=
(
nb_rb
*
8
)
-
(
2
*
pbch_pss_sss_adjust
/
3
);
...
...
@@ -1115,18 +1120,49 @@ void dlsch_64qam_llr(LTE_DL_FRAME_PARMS *frame_parms,
len2
=
len
>>
2
;
// length in quad words (4 REs)
len2
+=
((
len_mod4
==
0
)
?
0
:
1
);
for
(
i
=
0
;
i
<
len2
;
i
++
)
{
qam64_llr
((
short
*
)
rxF
,
(
short
*
)
ch_mag
,
(
short
*
)
ch_magb
,
llr2
,
len2
);
}
void
qam64_llr
(
int16_t
*
stream0_in
,
int16_t
*
chan_magn
,
int16_t
*
chan_magn_b
,
int16_t
*
llr
,
int
length
)
{
#if defined(__x86_64__) || defined(__i386__)
xmm1
=
_mm_abs_epi16
(
rxF
[
i
]);
xmm1
=
_mm_subs_epi16
(
ch_mag
[
i
],
xmm1
);
__m128i
*
rxF_128
=
(
__m128i
*
)
stream0_in
;
__m128i
*
ch_mag_128
=
(
__m128i
*
)
chan_magn
;
__m128i
*
ch_magb_128
=
(
__m128i
*
)
chan_magn_b
;
#elif defined(__arm__)
int16x8_t
*
rxF_128
=
(
int16x8_t
*
)
stream0_in
;
int16x8_t
*
ch_mag_128
=
(
int16x8_t
*
)
chan_magn
;
int16x8_t
*
ch_magb_128
=
(
int16x8_t
*
)
chan_magn_b
;
int16x8_t
xmm1
,
xmm2
;
#endif
int
i
;
//int16_t *llr2;
//llr2 = llr;
for
(
i
=
0
;
i
<
length
;
i
++
)
{
#if defined(__x86_64__) || defined(__i386__)
xmm1
=
_mm_abs_epi16
(
rxF_128
[
i
]);
xmm1
=
_mm_subs_epi16
(
ch_mag_128
[
i
],
xmm1
);
xmm2
=
_mm_abs_epi16
(
xmm1
);
xmm2
=
_mm_subs_epi16
(
ch_magb
[
i
],
xmm2
);
xmm2
=
_mm_subs_epi16
(
ch_magb
_128
[
i
],
xmm2
);
#elif defined(__arm__)
xmm1
=
vabsq_s16
(
rxF
[
i
]);
xmm1
=
vsubq_s16
(
ch_mag
[
i
],
xmm1
);
xmm1
=
vabsq_s16
(
rxF
_128
[
i
]);
xmm1
=
vsubq_s16
(
ch_mag
_128
[
i
],
xmm1
);
xmm2
=
vabsq_s16
(
xmm1
);
xmm2
=
vsubq_s16
(
ch_magb
[
i
],
xmm2
);
xmm2
=
vsubq_s16
(
ch_magb
_128
[
i
],
xmm2
);
#endif
// loop over all LLRs in quad word (24 coded bits)
/*
...
...
@@ -1141,64 +1177,64 @@ void dlsch_64qam_llr(LTE_DL_FRAME_PARMS *frame_parms,
llr2+=6;
}
*/
llr
2
[
0
]
=
((
short
*
)
&
rxF
[
i
])[
0
];
llr
2
[
1
]
=
((
short
*
)
&
rxF
[
i
])[
1
];
llr
[
0
]
=
((
short
*
)
&
rxF_128
[
i
])[
0
];
llr
[
1
]
=
((
short
*
)
&
rxF_128
[
i
])[
1
];
#if defined(__x86_64__) || defined(__i386__)
llr
2
[
2
]
=
_mm_extract_epi16
(
xmm1
,
0
);
llr
2
[
3
]
=
_mm_extract_epi16
(
xmm1
,
1
);
//((short *)&xmm1)[j+1];
llr
2
[
4
]
=
_mm_extract_epi16
(
xmm2
,
0
);
//((short *)&xmm2)[j];
llr
2
[
5
]
=
_mm_extract_epi16
(
xmm2
,
1
);
//((short *)&xmm2)[j+1];
llr
[
2
]
=
_mm_extract_epi16
(
xmm1
,
0
);
llr
[
3
]
=
_mm_extract_epi16
(
xmm1
,
1
);
//((short *)&xmm1)[j+1];
llr
[
4
]
=
_mm_extract_epi16
(
xmm2
,
0
);
//((short *)&xmm2)[j];
llr
[
5
]
=
_mm_extract_epi16
(
xmm2
,
1
);
//((short *)&xmm2)[j+1];
#elif defined(__arm__)
llr
2
[
2
]
=
vgetq_lane_s16
(
xmm1
,
0
);
llr
2
[
3
]
=
vgetq_lane_s16
(
xmm1
,
1
);
//((short *)&xmm1)[j+1];
llr
2
[
4
]
=
vgetq_lane_s16
(
xmm2
,
0
);
//((short *)&xmm2)[j];
llr
2
[
5
]
=
vgetq_lane_s16
(
xmm2
,
1
);
//((short *)&xmm2)[j+1];
llr
[
2
]
=
vgetq_lane_s16
(
xmm1
,
0
);
llr
[
3
]
=
vgetq_lane_s16
(
xmm1
,
1
);
//((short *)&xmm1)[j+1];
llr
[
4
]
=
vgetq_lane_s16
(
xmm2
,
0
);
//((short *)&xmm2)[j];
llr
[
5
]
=
vgetq_lane_s16
(
xmm2
,
1
);
//((short *)&xmm2)[j+1];
#endif
llr
2
+=
6
;
llr
2
[
0
]
=
((
short
*
)
&
rxF
[
i
])[
2
];
llr
2
[
1
]
=
((
short
*
)
&
rxF
[
i
])[
3
];
llr
+=
6
;
llr
[
0
]
=
((
short
*
)
&
rxF_128
[
i
])[
2
];
llr
[
1
]
=
((
short
*
)
&
rxF_128
[
i
])[
3
];
#if defined(__x86_64__) || defined(__i386__)
llr
2
[
2
]
=
_mm_extract_epi16
(
xmm1
,
2
);
llr
2
[
3
]
=
_mm_extract_epi16
(
xmm1
,
3
);
//((short *)&xmm1)[j+1];
llr
2
[
4
]
=
_mm_extract_epi16
(
xmm2
,
2
);
//((short *)&xmm2)[j];
llr
2
[
5
]
=
_mm_extract_epi16
(
xmm2
,
3
);
//((short *)&xmm2)[j+1];
llr
[
2
]
=
_mm_extract_epi16
(
xmm1
,
2
);
llr
[
3
]
=
_mm_extract_epi16
(
xmm1
,
3
);
//((short *)&xmm1)[j+1];
llr
[
4
]
=
_mm_extract_epi16
(
xmm2
,
2
);
//((short *)&xmm2)[j];
llr
[
5
]
=
_mm_extract_epi16
(
xmm2
,
3
);
//((short *)&xmm2)[j+1];
#elif defined(__arm__)
llr
2
[
2
]
=
vgetq_lane_s16
(
xmm1
,
2
);
llr
2
[
3
]
=
vgetq_lane_s16
(
xmm1
,
3
);
//((short *)&xmm1)[j+1];
llr
2
[
4
]
=
vgetq_lane_s16
(
xmm2
,
2
);
//((short *)&xmm2)[j];
llr
2
[
5
]
=
vgetq_lane_s16
(
xmm2
,
3
);
//((short *)&xmm2)[j+1];
llr
[
2
]
=
vgetq_lane_s16
(
xmm1
,
2
);
llr
[
3
]
=
vgetq_lane_s16
(
xmm1
,
3
);
//((short *)&xmm1)[j+1];
llr
[
4
]
=
vgetq_lane_s16
(
xmm2
,
2
);
//((short *)&xmm2)[j];
llr
[
5
]
=
vgetq_lane_s16
(
xmm2
,
3
);
//((short *)&xmm2)[j+1];
#endif
llr
2
+=
6
;
llr
2
[
0
]
=
((
short
*
)
&
rxF
[
i
])[
4
];
llr
2
[
1
]
=
((
short
*
)
&
rxF
[
i
])[
5
];
llr
+=
6
;
llr
[
0
]
=
((
short
*
)
&
rxF_128
[
i
])[
4
];
llr
[
1
]
=
((
short
*
)
&
rxF_128
[
i
])[
5
];
#if defined(__x86_64__) || defined(__i386__)
llr
2
[
2
]
=
_mm_extract_epi16
(
xmm1
,
4
);
llr
2
[
3
]
=
_mm_extract_epi16
(
xmm1
,
5
);
//((short *)&xmm1)[j+1];
llr
2
[
4
]
=
_mm_extract_epi16
(
xmm2
,
4
);
//((short *)&xmm2)[j];
llr
2
[
5
]
=
_mm_extract_epi16
(
xmm2
,
5
);
//((short *)&xmm2)[j+1];
llr
[
2
]
=
_mm_extract_epi16
(
xmm1
,
4
);
llr
[
3
]
=
_mm_extract_epi16
(
xmm1
,
5
);
//((short *)&xmm1)[j+1];
llr
[
4
]
=
_mm_extract_epi16
(
xmm2
,
4
);
//((short *)&xmm2)[j];
llr
[
5
]
=
_mm_extract_epi16
(
xmm2
,
5
);
//((short *)&xmm2)[j+1];
#elif defined(__arm__)
llr
2
[
2
]
=
vgetq_lane_s16
(
xmm1
,
4
);
llr
2
[
3
]
=
vgetq_lane_s16
(
xmm1
,
5
);
//((short *)&xmm1)[j+1];
llr
2
[
4
]
=
vgetq_lane_s16
(
xmm2
,
4
);
//((short *)&xmm2)[j];
llr
2
[
5
]
=
vgetq_lane_s16
(
xmm2
,
5
);
//((short *)&xmm2)[j+1];
llr
[
2
]
=
vgetq_lane_s16
(
xmm1
,
4
);
llr
[
3
]
=
vgetq_lane_s16
(
xmm1
,
5
);
//((short *)&xmm1)[j+1];
llr
[
4
]
=
vgetq_lane_s16
(
xmm2
,
4
);
//((short *)&xmm2)[j];
llr
[
5
]
=
vgetq_lane_s16
(
xmm2
,
5
);
//((short *)&xmm2)[j+1];
#endif
llr
2
+=
6
;
llr
2
[
0
]
=
((
short
*
)
&
rxF
[
i
])[
6
];
llr
2
[
1
]
=
((
short
*
)
&
rxF
[
i
])[
7
];
llr
+=
6
;
llr
[
0
]
=
((
short
*
)
&
rxF_128
[
i
])[
6
];
llr
[
1
]
=
((
short
*
)
&
rxF_128
[
i
])[
7
];
#if defined(__x86_64__) || defined(__i386__)
llr
2
[
2
]
=
_mm_extract_epi16
(
xmm1
,
6
);
llr
2
[
3
]
=
_mm_extract_epi16
(
xmm1
,
7
);
//((short *)&xmm1)[j+1];
llr
2
[
4
]
=
_mm_extract_epi16
(
xmm2
,
6
);
//((short *)&xmm2)[j];
llr
2
[
5
]
=
_mm_extract_epi16
(
xmm2
,
7
);
//((short *)&xmm2)[j+1];
llr
[
2
]
=
_mm_extract_epi16
(
xmm1
,
6
);
llr
[
3
]
=
_mm_extract_epi16
(
xmm1
,
7
);
//((short *)&xmm1)[j+1];
llr
[
4
]
=
_mm_extract_epi16
(
xmm2
,
6
);
//((short *)&xmm2)[j];
llr
[
5
]
=
_mm_extract_epi16
(
xmm2
,
7
);
//((short *)&xmm2)[j+1];
#elif defined(__arm__)
llr
2
[
2
]
=
vgetq_lane_s16
(
xmm1
,
6
);
llr
2
[
3
]
=
vgetq_lane_s16
(
xmm1
,
7
);
//((short *)&xmm1)[j+1];
llr
2
[
4
]
=
vgetq_lane_s16
(
xmm2
,
6
);
//((short *)&xmm2)[j];
llr
2
[
5
]
=
vgetq_lane_s16
(
xmm2
,
7
);
//((short *)&xmm2)[j+1];
llr
[
2
]
=
vgetq_lane_s16
(
xmm1
,
6
);
llr
[
3
]
=
vgetq_lane_s16
(
xmm1
,
7
);
//((short *)&xmm1)[j+1];
llr
[
4
]
=
vgetq_lane_s16
(
xmm2
,
6
);
//((short *)&xmm2)[j];
llr
[
5
]
=
vgetq_lane_s16
(
xmm2
,
7
);
//((short *)&xmm2)[j+1];
#endif
llr
2
+=
6
;
llr
+=
6
;
}
...
...
@@ -8865,7 +8901,7 @@ int dlsch_64qam_64qam_llr(LTE_DL_FRAME_PARMS *frame_parms,
(
int16_t
*
)
llr16
,
(
int32_t
*
)
rho_256i
,
len
);
free16
(
rxF_256i
,
sizeof
(
rxF_256i
));
free16
(
rxF_i_256i
,
sizeof
(
rxF_i_256i
));
free16
(
ch_mag_256i
,
sizeof
(
ch_mag_256i
));
...
...
openair1/PHY/LTE_UE_TRANSPORT/transport_proto_ue.h
View file @
1d765027
...
...
@@ -106,6 +106,22 @@ void qpsk_qpsk(int16_t *stream0_in,
@param nb_rb number of RBs for this allocation
@param pbch_pss_sss_adj Number of channel bits taken by PBCH/PSS/SSS
@param llr128p pointer to pointer to symbol in dlsch_llr*/
void
qpsk_llr
(
int16_t
*
stream0_in
,
int16_t
*
stream0_out
,
int
length
);
void
qam16_llr
(
int16_t
*
stream0_in
,
int16_t
*
chan_magn
,
int16_t
*
llr
,
int
length
);
void
qam64_llr
(
int16_t
*
stream0_in
,
int16_t
*
chan_magn
,
int16_t
*
chan_magn_b
,
int16_t
*
llr
,
int
length
);
int32_t
dlsch_qpsk_qpsk_llr
(
LTE_DL_FRAME_PARMS
*
frame_parms
,
int32_t
**
rxdataF_comp
,
int32_t
**
rxdataF_comp_i
,
...
...
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment