From aacc37613b3299bf122774d36148039ec65014ed Mon Sep 17 00:00:00 2001 From: sfn Date: Thu, 22 Apr 2021 16:40:37 +0200 Subject: [PATCH] Fix zero-forcing for 2 layers --- .../NR_UE_TRANSPORT/nr_dlsch_demodulation.c | 491 +++++++++--------- .../NR_UE_TRANSPORT/nr_transport_proto_ue.h | 18 +- 2 files changed, 266 insertions(+), 243 deletions(-) diff --git a/openair1/PHY/NR_UE_TRANSPORT/nr_dlsch_demodulation.c b/openair1/PHY/NR_UE_TRANSPORT/nr_dlsch_demodulation.c index 4838cd5f11..92b2fbfca7 100644 --- a/openair1/PHY/NR_UE_TRANSPORT/nr_dlsch_demodulation.c +++ b/openair1/PHY/NR_UE_TRANSPORT/nr_dlsch_demodulation.c @@ -84,8 +84,8 @@ unsigned char offset_mumimo_llr_drange[29][3]={{8,8,8},{7,7,7},{7,7,7},{7,7,7},{ */ //unsigned char offset_mumimo_llr_drange[29][3]= {{0, 6, 5},{0, 4, 5},{0, 4, 5},{0, 5, 4},{0, 5, 6},{0, 5, 3},{0, 4, 4},{0, 4, 4},{0, 3, 3},{0, 1, 2},{1, 1, 0},{1, 3, 2},{3, 4, 1},{2, 0, 0},{2, 2, 2},{1, 1, 1},{2, 1, 0},{2, 1, 1},{1, 0, 1},{1, 0, 1},{0, 0, 0},{1, 0, 0},{0, 0, 0},{0, 1, 0},{1, 0, 0},{0, 0, 0},{0, 0, 0},{0, 0, 0},{0, 0, 0}}; -#define print_ints(s,x) printf("%s %d %d %d %d\n",s,(x)[0],(x)[1],(x)[2],(x)[3]) -#define print_shorts(s,x) printf("%s %d,%d,%d,%d,%d,%d,%d,%d\n",s,(x)[0],(x)[1],(x)[2],(x)[3],(x)[4],(x)[5],(x)[6],(x)[7]) +#define print_ints(s,x) printf("%s = %d %d %d %d\n",s,(x)[0],(x)[1],(x)[2],(x)[3]) +#define print_shorts(s,x) printf("%s = [%d+j*%d, %d+j*%d, %d+j*%d, %d+j*%d]\n",s,(x)[0],(x)[1],(x)[2],(x)[3],(x)[4],(x)[5],(x)[6],(x)[7]) static void nr_dlsch_dual_stream_correlation_core(int **dl_ch_estimates_ext, int **dl_ch_estimates_ext_i, @@ -99,12 +99,14 @@ static void nr_dlsch_dual_stream_correlation_core(int **dl_ch_estimates_ext, uint8_t nr_zero_forcing_rx_2layers(int **rxdataF_comp, int **dl_ch_mag, int **dl_ch_magb, + int **dl_ch_magr, int **dl_ch_estimates_ext, unsigned short nb_rb, unsigned char n_rx, unsigned char mod_order, int shift, - unsigned char symbol); + unsigned char symbol, + int length); static void nr_dlsch_layer_demapping(int16_t **llr_cw, uint8_t Nl, @@ -414,7 +416,7 @@ int nr_rx_pdsch(PHY_VARS_NR_UE *ue, symbol, pilots, len, - nb_rb); + nb_rb_pdsch); #if UE_TIMING_TRACE stop_meas(&ue->generic_stat_bis[proc->thread_id][slot]); @@ -436,7 +438,7 @@ int nr_rx_pdsch(PHY_VARS_NR_UE *ue, avg, symbol, len, - nb_rb); + nb_rb_pdsch); avgs = 0; for (aatx=0;aatxQm, - nb_rb, + nb_rb_pdsch, pdsch_vars[gNB_id]->log2_maxh, measurements); // log2_maxh+I0_shift } @@ -558,20 +560,24 @@ int nr_rx_pdsch(PHY_VARS_NR_UE *ue, (n_tx>1)? pdsch_vars[gNB_id]->rho : NULL, pdsch_vars[gNB_id]->dl_ch_mag0, pdsch_vars[gNB_id]->dl_ch_magb0, + pdsch_vars[gNB_id]->dl_ch_magr0, n_tx, n_rx, symbol, - nb_rb); + nb_rb_pdsch, + len); if (n_tx == 2)//Apply zero forcing for 2 Tx layers nr_zero_forcing_rx_2layers(pdsch_vars[gNB_id]->rxdataF_comp0, pdsch_vars[gNB_id]->dl_ch_mag0, pdsch_vars[gNB_id]->dl_ch_magb0, + pdsch_vars[gNB_id]->dl_ch_magr0, pdsch_vars[gNB_id]->dl_ch_estimates_ext, - nb_rb, + nb_rb_pdsch, n_rx, dlsch0_harq->Qm, pdsch_vars[gNB_id]->log2_maxh, - symbol); + symbol, + len); } else if (dlsch0_harq->mimo_mode == NR_DUALSTREAM) { nr_dlsch_detection_mrc_core(pdsch_vars[gNB_id]->rxdataF_comp0, @@ -795,7 +801,7 @@ void nr_dlsch_channel_compensation(int **rxdataF_ext, NR_DL_FRAME_PARMS *frame_parms, uint8_t nb_aatx, unsigned char symbol, - uint8_t pilots, + int length, uint8_t first_symbol_flag, unsigned char mod_order, unsigned short nb_rb, @@ -811,6 +817,7 @@ void nr_dlsch_channel_compensation(int **rxdataF_ext, __m128i mmtmpD0,mmtmpD1,mmtmpD2,mmtmpD3,QAM_amp128,QAM_amp128b,QAM_amp128r; QAM_amp128b = _mm_setzero_si128(); + uint32_t nb_rb_0 = length/12 + ((length%12)?1:0); for (aatx=0; aatxnb_antennas_rx)+aarx][symbol*nb_rb*12]; - for (rb=0; rb2) { // get channel amplitude if not QPSK @@ -865,18 +872,16 @@ void nr_dlsch_channel_compensation(int **rxdataF_ext, dl_ch_mag128[1] = _mm_mulhi_epi16(dl_ch_mag128[1],QAM_amp128); dl_ch_mag128[1] = _mm_slli_epi16(dl_ch_mag128[1],1); - if (pilots==0) { - mmtmpD0 = _mm_madd_epi16(dl_ch128[2],dl_ch128[2]); - mmtmpD0 = _mm_srai_epi32(mmtmpD0,output_shift); - mmtmpD1 = _mm_packs_epi32(mmtmpD0,mmtmpD0); + mmtmpD0 = _mm_madd_epi16(dl_ch128[2],dl_ch128[2]);//[H_I(0)^2+H_Q(0)^2 H_I(1)^2+H_Q(1)^2 H_I(2)^2+H_Q(2)^2 H_I(3)^2+H_Q(3)^2] + mmtmpD0 = _mm_srai_epi32(mmtmpD0,output_shift); + mmtmpD1 = _mm_packs_epi32(mmtmpD0,mmtmpD0);//[|H(0)|^2 |H(1)|^2 |H(2)|^2 |H(3)|^2 |H(0)|^2 |H(1)|^2 |H(2)|^2 |H(3)|^2] - dl_ch_mag128[2] = _mm_unpacklo_epi16(mmtmpD1,mmtmpD1); - dl_ch_mag128b[2] = dl_ch_mag128[2]; - dl_ch_mag128r[2] = dl_ch_mag128[2]; + dl_ch_mag128[2] = _mm_unpacklo_epi16(mmtmpD1,mmtmpD1);//[|H(0)|^2 |H(0)|^2 |H(1)|^2 |H(1)|^2 |H(2)|^2 |H(2)|^2 |H(3)|^2 |H(3)|^2] + dl_ch_mag128b[2] = dl_ch_mag128[2]; + dl_ch_mag128r[2] = dl_ch_mag128[2]; - dl_ch_mag128[2] = _mm_mulhi_epi16(dl_ch_mag128[2],QAM_amp128); - dl_ch_mag128[2] = _mm_slli_epi16(dl_ch_mag128[2],1); - } + dl_ch_mag128[2] = _mm_mulhi_epi16(dl_ch_mag128[2],QAM_amp128); + dl_ch_mag128[2] = _mm_slli_epi16(dl_ch_mag128[2],1); dl_ch_mag128b[0] = _mm_mulhi_epi16(dl_ch_mag128b[0],QAM_amp128b); dl_ch_mag128b[0] = _mm_slli_epi16(dl_ch_mag128b[0],1); @@ -891,13 +896,11 @@ void nr_dlsch_channel_compensation(int **rxdataF_ext, dl_ch_mag128r[1] = _mm_mulhi_epi16(dl_ch_mag128r[1],QAM_amp128r); dl_ch_mag128r[1] = _mm_slli_epi16(dl_ch_mag128r[1],1); - if (pilots==0) { - dl_ch_mag128b[2] = _mm_mulhi_epi16(dl_ch_mag128b[2],QAM_amp128b); - dl_ch_mag128b[2] = _mm_slli_epi16(dl_ch_mag128b[2],1); + dl_ch_mag128b[2] = _mm_mulhi_epi16(dl_ch_mag128b[2],QAM_amp128b); + dl_ch_mag128b[2] = _mm_slli_epi16(dl_ch_mag128b[2],1); - dl_ch_mag128r[2] = _mm_mulhi_epi16(dl_ch_mag128r[2],QAM_amp128r); - dl_ch_mag128r[2] = _mm_slli_epi16(dl_ch_mag128r[2],1); - } + dl_ch_mag128r[2] = _mm_mulhi_epi16(dl_ch_mag128r[2],QAM_amp128r); + dl_ch_mag128r[2] = _mm_slli_epi16(dl_ch_mag128r[2],1); } // multiply by conjugated channel @@ -920,15 +923,17 @@ void nr_dlsch_channel_compensation(int **rxdataF_ext, // print_ints("c0",&mmtmpD2); // print_ints("c1",&mmtmpD3); rxdataF_comp128[0] = _mm_packs_epi32(mmtmpD2,mmtmpD3); - //printf("arx%d atx%d rb_index %d symbol%d\n",aarx,aatx,rb,symbol); - // print_shorts("rx:",rxdataF128); - // print_shorts("ch:",dl_ch128); - // print_shorts("pack:",rxdataF_comp128); - //printf("arx%d atx%d rb_index %d symbol%d\n",aarx,aatx,rb,symbol); - //print_shorts("rx:",(int16_t*)&rxdataF128[0]); - //print_shorts("ch:",(int16_t*)&dl_ch128[0]); - //print_shorts("pack:",(int16_t*)&rxdataF_comp128[0]); +#ifdef DEBUG_DLSCH_DEMOD + printf("%%arx%d atx%d rb_index %d symbol %d shift %d\n",aarx,aatx,rb,symbol,output_shift); + printf("rx_%d(%d,:)",aarx+1,rb+1); + print_shorts(" ",(int16_t *)&rxdataF128[0]); + printf("ch_%d%d(%d,:)",aarx+1,aatx+1,rb+1); + print_shorts(" ",(int16_t *)&dl_ch128[0]); + printf("rx_comp_%d%d(%d,:)",aarx+1,aatx+1,rb+1); + print_shorts(" ",(int16_t *)&rxdataF_comp128[0]); +#endif + // multiply by conjugated channel mmtmpD0 = _mm_madd_epi16(dl_ch128[1],rxdataF128[1]); // mmtmpD0 contains real part of 4 consecutive outputs (32-bit) @@ -947,40 +952,30 @@ void nr_dlsch_channel_compensation(int **rxdataF_ext, //print_shorts("ch:",(int16_t*)&dl_ch128[1]); //print_shorts("pack:",(int16_t*)&rxdataF_comp128[1]); - if (pilots==0) { - // multiply by conjugated channel - mmtmpD0 = _mm_madd_epi16(dl_ch128[2],rxdataF128[2]); - // mmtmpD0 contains real part of 4 consecutive outputs (32-bit) - mmtmpD1 = _mm_shufflelo_epi16(dl_ch128[2],_MM_SHUFFLE(2,3,0,1)); - mmtmpD1 = _mm_shufflehi_epi16(mmtmpD1,_MM_SHUFFLE(2,3,0,1)); - mmtmpD1 = _mm_sign_epi16(mmtmpD1,*(__m128i*)conjugate); - mmtmpD1 = _mm_madd_epi16(mmtmpD1,rxdataF128[2]); - // mmtmpD1 contains imag part of 4 consecutive outputs (32-bit) - mmtmpD0 = _mm_srai_epi32(mmtmpD0,output_shift); - mmtmpD1 = _mm_srai_epi32(mmtmpD1,output_shift); - mmtmpD2 = _mm_unpacklo_epi32(mmtmpD0,mmtmpD1); - mmtmpD3 = _mm_unpackhi_epi32(mmtmpD0,mmtmpD1); + // multiply by conjugated channel + mmtmpD0 = _mm_madd_epi16(dl_ch128[2],rxdataF128[2]); + // mmtmpD0 contains real part of 4 consecutive outputs (32-bit) + mmtmpD1 = _mm_shufflelo_epi16(dl_ch128[2],_MM_SHUFFLE(2,3,0,1)); + mmtmpD1 = _mm_shufflehi_epi16(mmtmpD1,_MM_SHUFFLE(2,3,0,1)); + mmtmpD1 = _mm_sign_epi16(mmtmpD1,*(__m128i*)conjugate); + mmtmpD1 = _mm_madd_epi16(mmtmpD1,rxdataF128[2]); + // mmtmpD1 contains imag part of 4 consecutive outputs (32-bit) + mmtmpD0 = _mm_srai_epi32(mmtmpD0,output_shift); + mmtmpD1 = _mm_srai_epi32(mmtmpD1,output_shift); + mmtmpD2 = _mm_unpacklo_epi32(mmtmpD0,mmtmpD1); + mmtmpD3 = _mm_unpackhi_epi32(mmtmpD0,mmtmpD1); - rxdataF_comp128[2] = _mm_packs_epi32(mmtmpD2,mmtmpD3); - //print_shorts("rx:",(int16_t*)&rxdataF128[2]); - //print_shorts("ch:",(int16_t*)&dl_ch128[2]); - //print_shorts("pack:",(int16_t*)&rxdataF_comp128[2]); + rxdataF_comp128[2] = _mm_packs_epi32(mmtmpD2,mmtmpD3); + //print_shorts("rx:",(int16_t*)&rxdataF128[2]); + //print_shorts("ch:",(int16_t*)&dl_ch128[2]); + //print_shorts("pack:",(int16_t*)&rxdataF_comp128[2]); - dl_ch128+=3; - dl_ch_mag128+=3; - dl_ch_mag128b+=3; - dl_ch_mag128r+=3; - rxdataF128+=3; - rxdataF_comp128+=3; - } - else { // we have a smaller PDSCH in symbols with pilots so skip last group of 4 REs and increment less - dl_ch128+=2; - dl_ch_mag128+=2; - dl_ch_mag128b+=2; - dl_ch_mag128r+=2; - rxdataF128+=2; - rxdataF_comp128+=2; - } + dl_ch128+=3; + dl_ch_mag128+=3; + dl_ch_mag128b+=3; + dl_ch_mag128r+=3; + rxdataF128+=3; + rxdataF_comp128+=3; } } } @@ -1004,7 +999,7 @@ void nr_dlsch_channel_compensation(int **rxdataF_ext, rho128 = (__m128i *)&rho[aarx][aatx*nb_aatx+atx][symbol*nb_rb*12]; dl_ch128_2 = (__m128i *)&dl_ch_estimates_ext[atx*frame_parms->nb_antennas_rx+aarx][symbol*nb_rb*12]; - for (rb=0; rbrx_correlation[0][0][aarx] = signal_energy(&rho[aarx][aatx*nb_aatx+atx][symbol*nb_rb*12],rb*12); avg_rho_re[aarx][aatx*nb_aatx+atx] = 16*avg_rho_re[aarx][aatx*nb_aatx+atx]/(nb_rb*12); avg_rho_im[aarx][aatx*nb_aatx+atx] = 16*avg_rho_im[aarx][aatx*nb_aatx+atx]/(nb_rb*12); - printf("rho[rx]%d tx%d tx%d = Re: %d Im: %d\n",aarx, aatx,atx, avg_rho_re[aarx][aatx*nb_aatx+atx], avg_rho_im[aarx][aatx*nb_aatx+atx]); + //printf("rho[rx]%d tx%d tx%d = Re: %d Im: %d\n",aarx, aatx,atx, avg_rho_re[aarx][aatx*nb_aatx+atx], avg_rho_im[aarx][aatx*nb_aatx+atx]); } } } @@ -1112,7 +1107,7 @@ void nr_dlsch_channel_compensation(int **rxdataF_ext, #elif defined(__arm__) unsigned short rb; - unsigned char aatx,aarx,symbol_mod,pilots=0; + unsigned char aatx,aarx,symbol_mod; int16x4_t *dl_ch128,*dl_ch128_2,*rxdataF128; int32x4_t mmtmpD0,mmtmpD1,mmtmpD0b,mmtmpD1b; @@ -1125,14 +1120,7 @@ void nr_dlsch_channel_compensation(int **rxdataF_ext, symbol_mod = (symbol>=(7-frame_parms->Ncp)) ? symbol-(7-frame_parms->Ncp) : symbol; - if ((symbol_mod == 0) || (symbol_mod == (4-frame_parms->Ncp))) { - if (frame_parms->nb_antenna_ports_gNB==1) { // 10 out of 12 so don't reduce size - nb_rb=1+(5*nb_rb/6); - } - else { - pilots=1; - } - } + uint32_t nb_rb_0 = length/12 + ((length%12)?1:0); for (aatx=0; aatxnb_antenna_ports_gNB; aatx++) { if (mod_order == 4) { @@ -1151,7 +1139,7 @@ void nr_dlsch_channel_compensation(int **rxdataF_ext, rxdataF128 = (int16x4_t*)&rxdataF_ext[aarx][symbol*frame_parms->N_RB_DL*12]; rxdataF_comp128 = (int16x4x2_t*)&rxdataF_comp[(aatx<<1)+aarx][symbol*frame_parms->N_RB_DL*12]; - for (rb=0; rb2) { // get channel amplitude if not QPSK mmtmpD0 = vmull_s16(dl_ch128[0], dl_ch128[0]); @@ -1167,23 +1155,20 @@ void nr_dlsch_channel_compensation(int **rxdataF_ext, mmtmpD1 = vmull_s16(dl_ch128[3], dl_ch128[3]); mmtmpD1 = vqshlq_s32(vqaddq_s32(mmtmpD1,vrev64q_s32(mmtmpD1)),output_shift128); mmtmpD3 = vcombine_s16(vmovn_s32(mmtmpD0),vmovn_s32(mmtmpD1)); - if (pilots==0) { - mmtmpD0 = vmull_s16(dl_ch128[4], dl_ch128[4]); - mmtmpD0 = vqshlq_s32(vqaddq_s32(mmtmpD0,vrev64q_s32(mmtmpD0)),output_shift128); - mmtmpD1 = vmull_s16(dl_ch128[5], dl_ch128[5]); - mmtmpD1 = vqshlq_s32(vqaddq_s32(mmtmpD1,vrev64q_s32(mmtmpD1)),output_shift128); - mmtmpD4 = vcombine_s16(vmovn_s32(mmtmpD0),vmovn_s32(mmtmpD1)); - } + + mmtmpD0 = vmull_s16(dl_ch128[4], dl_ch128[4]); + mmtmpD0 = vqshlq_s32(vqaddq_s32(mmtmpD0,vrev64q_s32(mmtmpD0)),output_shift128); + mmtmpD1 = vmull_s16(dl_ch128[5], dl_ch128[5]); + mmtmpD1 = vqshlq_s32(vqaddq_s32(mmtmpD1,vrev64q_s32(mmtmpD1)),output_shift128); + mmtmpD4 = vcombine_s16(vmovn_s32(mmtmpD0),vmovn_s32(mmtmpD1)); dl_ch_mag128b[0] = vqdmulhq_s16(mmtmpD2,QAM_amp128b); dl_ch_mag128b[1] = vqdmulhq_s16(mmtmpD3,QAM_amp128b); dl_ch_mag128[0] = vqdmulhq_s16(mmtmpD2,QAM_amp128); dl_ch_mag128[1] = vqdmulhq_s16(mmtmpD3,QAM_amp128); - if (pilots==0) { - dl_ch_mag128b[2] = vqdmulhq_s16(mmtmpD4,QAM_amp128b); - dl_ch_mag128[2] = vqdmulhq_s16(mmtmpD4,QAM_amp128); - } + dl_ch_mag128b[2] = vqdmulhq_s16(mmtmpD4,QAM_amp128b); + dl_ch_mag128[2] = vqdmulhq_s16(mmtmpD4,QAM_amp128); } mmtmpD0 = vmull_s16(dl_ch128[0], rxdataF128[0]); @@ -1217,36 +1202,29 @@ void nr_dlsch_channel_compensation(int **rxdataF_ext, mmtmpD1 = vqshlq_s32(mmtmpD1,output_shift128); rxdataF_comp128[1] = vzip_s16(vmovn_s32(mmtmpD0),vmovn_s32(mmtmpD1)); - if (pilots==0) { - mmtmpD0 = vmull_s16(dl_ch128[4], rxdataF128[4]); - mmtmpD1 = vmull_s16(dl_ch128[5], rxdataF128[5]); - mmtmpD0 = vcombine_s32(vpadd_s32(vget_low_s32(mmtmpD0),vget_high_s32(mmtmpD0)), - vpadd_s32(vget_low_s32(mmtmpD1),vget_high_s32(mmtmpD1))); - mmtmpD0b = vmull_s16(vrev32_s16(vmul_s16(dl_ch128[4],*(int16x4_t*)conj)), rxdataF128[4]); - mmtmpD1b = vmull_s16(vrev32_s16(vmul_s16(dl_ch128[5],*(int16x4_t*)conj)), rxdataF128[5]); - mmtmpD1 = vcombine_s32(vpadd_s32(vget_low_s32(mmtmpD0b),vget_high_s32(mmtmpD0b)), - vpadd_s32(vget_low_s32(mmtmpD1b),vget_high_s32(mmtmpD1b))); + mmtmpD0 = vmull_s16(dl_ch128[4], rxdataF128[4]); + mmtmpD1 = vmull_s16(dl_ch128[5], rxdataF128[5]); + mmtmpD0 = vcombine_s32(vpadd_s32(vget_low_s32(mmtmpD0),vget_high_s32(mmtmpD0)), + vpadd_s32(vget_low_s32(mmtmpD1),vget_high_s32(mmtmpD1))); + + mmtmpD0b = vmull_s16(vrev32_s16(vmul_s16(dl_ch128[4],*(int16x4_t*)conj)), rxdataF128[4]); + mmtmpD1b = vmull_s16(vrev32_s16(vmul_s16(dl_ch128[5],*(int16x4_t*)conj)), rxdataF128[5]); + mmtmpD1 = vcombine_s32(vpadd_s32(vget_low_s32(mmtmpD0b),vget_high_s32(mmtmpD0b)), + vpadd_s32(vget_low_s32(mmtmpD1b),vget_high_s32(mmtmpD1b))); - mmtmpD0 = vqshlq_s32(mmtmpD0,output_shift128); - mmtmpD1 = vqshlq_s32(mmtmpD1,output_shift128); - rxdataF_comp128[2] = vzip_s16(vmovn_s32(mmtmpD0),vmovn_s32(mmtmpD1)); + mmtmpD0 = vqshlq_s32(mmtmpD0,output_shift128); + mmtmpD1 = vqshlq_s32(mmtmpD1,output_shift128); + rxdataF_comp128[2] = vzip_s16(vmovn_s32(mmtmpD0),vmovn_s32(mmtmpD1)); - dl_ch128+=6; - dl_ch_mag128+=3; - dl_ch_mag128b+=3; - rxdataF128+=6; - rxdataF_comp128+=3; + dl_ch128+=6; + dl_ch_mag128+=3; + dl_ch_mag128b+=3; + rxdataF128+=6; + rxdataF_comp128+=3; - } else { // we have a smaller PDSCH in symbols with pilots so skip last group of 4 REs and increment less - dl_ch128+=4; - dl_ch_mag128+=2; - dl_ch_mag128b+=2; - rxdataF128+=4; - rxdataF_comp128+=2; - } } } } @@ -1256,7 +1234,7 @@ void nr_dlsch_channel_compensation(int **rxdataF_ext, rho128 = (int16x4x2_t*)&rho[aarx][symbol*frame_parms->N_RB_DL*12]; dl_ch128 = (int16x4_t*)&dl_ch_estimates_ext[aarx][symbol*frame_parms->N_RB_DL*12]; dl_ch128_2 = (int16x4_t*)&dl_ch_estimates_ext[2+aarx][symbol*frame_parms->N_RB_DL*12]; - for (rb=0; rb4) ? 4 : x; int16_t y = (len)>>x; //printf("len = %d = %d * 2^(%d)\n",len,y,x); + uint32_t nb_rb_0 = len/12 + ((len%12)?1:0); AssertFatal(y!=0,"Cannot divide by zero: in function %s of file %s\n", __func__, __FILE__); @@ -1604,7 +1579,7 @@ void nr_dlsch_channel_level(int **dl_ch_estimates_ext, dl_ch128=(__m128i *)&dl_ch_estimates_ext[(aatx*frame_parms->nb_antennas_rx)+aarx][symbol*nb_rb*12]; - for (rb=0;rb=(7-frame_parms->Ncp)) ? symbol-(7-frame_parms->Ncp) : symbol; - + uint32_t nb_rb_0 = len/12 + ((len%12)?1:0); for (aatx=0; aatxnb_antenna_ports_gNB; aatx++) for (aarx=0; aarxnb_antennas_rx; aarx++) { //clear average level @@ -1638,7 +1613,7 @@ void nr_dlsch_channel_level(int **dl_ch_estimates_ext, dl_ch128=(int16x4_t *)&dl_ch_estimates_ext[(aatx<<1)+aarx][symbol*frame_parms->N_RB_DL*12]; - for (rb=0; rbofdm_symbol_size)))]; for (rb = 0; rb < nb_rb_pdsch; rb++) { - if (k>frame_parms->ofdm_symbol_size) { + if (k>=frame_parms->ofdm_symbol_size) { k = k-frame_parms->ofdm_symbol_size; rxF = &rxdataF[aarx][(k+(symbol*(frame_parms->ofdm_symbol_size)))]; } @@ -2196,7 +2171,7 @@ unsigned short nr_dlsch_extract_rbs_multiple(int **rxdataF, for (aarx=0; aarxnb_antennas_rx; aarx++) { k = frame_parms->first_carrier_offset + NR_NB_SC_PER_RB*start_rb; - if (k>frame_parms->ofdm_symbol_size) + if (k>=frame_parms->ofdm_symbol_size) k = k-frame_parms->ofdm_symbol_size; rxF_ext = &rxdataF_ext[aarx][symbol*(nb_rb_pdsch*12)]; @@ -2265,43 +2240,55 @@ void nr_dlsch_detection_mrc(int **rxdataF_comp, int ***rho, int **dl_ch_mag, int **dl_ch_magb, + int **dl_ch_magr, short n_tx, short n_rx, unsigned char symbol, - unsigned short nb_rb) { + unsigned short nb_rb, + int length) { #if defined(__x86_64__)||defined(__i386__) unsigned char aatx, aarx; int i; - __m128i *rxdataF_comp128_0,*rxdataF_comp128_1,*dl_ch_mag128_0,*dl_ch_mag128_1,*dl_ch_mag128_0b,*dl_ch_mag128_1b; + __m128i *rxdataF_comp128_0,*rxdataF_comp128_1,*dl_ch_mag128_0,*dl_ch_mag128_1,*dl_ch_mag128_0b,*dl_ch_mag128_1b,*dl_ch_mag128_0r,*dl_ch_mag128_1r; + + uint32_t nb_rb_0 = length/12 + ((length%12)?1:0); if (n_rx>1) { for (aatx=0; aatxN_RB_DL*12]; rho128_1 = (__m128i *) &rho[1][symbol*frame_parms->N_RB_DL*12]; - for (i=0; i>2); + //printf("det_%d = %d log2 =%d \n",k,(((int *)&det[0])[k]),log2_approx(((int *)&det[0])[k])); + } + + xy_re_128 = _mm_slli_epi32(xy_re_128,5); + xy_re_128 = _mm_srai_epi32(xy_re_128,log2_approx(sum_det)); + xy_re_128 = _mm_slli_epi32(xy_re_128,5); + + xy_im_128 = _mm_slli_epi32(xy_im_128,5); + xy_im_128 = _mm_srai_epi32(xy_im_128,log2_approx(sum_det)); + xy_im_128 = _mm_slli_epi32(xy_im_128,5); + + tmp_z0 = _mm_unpacklo_epi32(xy_re_128,xy_im_128); + //print_ints("unpack lo:",&tmp_z0[0]); + tmp_z1 = _mm_unpackhi_epi32(xy_re_128,xy_im_128); + //print_ints("unpack hi:",&tmp_z1[0]); + output = _mm_packs_epi32(tmp_z0,tmp_z1); + + _mm_empty(); + _m_empty(); + return(output); +} /* Zero Forcing Rx function: nr_construct_HhH_elements() * * @@ -2549,15 +2589,16 @@ void nr_construct_HhH_elements(int *conjch00_ch00, if (conjch21_ch20 != NULL) after_mf_10_128[0] =_mm_adds_epi16(after_mf_10_128[0],conjch21_ch20_128[0]); if (conjch31_ch30 != NULL) after_mf_10_128[0] =_mm_adds_epi16(after_mf_10_128[0],conjch31_ch30_128[0]); - /*if ((rb==0)&&(symbol==1)) +#ifdef DEBUG_DLSCH_DEMOD + if ((rb<=30)) { printf(" \n construct_HhH_elements \n"); print_shorts("after_mf_00_128:",(int16_t*)&after_mf_00_128[0]); print_shorts("after_mf_01_128:",(int16_t*)&after_mf_01_128[0]); print_shorts("after_mf_10_128:",(int16_t*)&after_mf_10_128[0]); print_shorts("after_mf_11_128:",(int16_t*)&after_mf_11_128[0]); - }*/ - + } +#endif conjch00_ch00_128+=1; conjch10_ch10_128+=1; conjch01_ch01_128+=1; @@ -2594,16 +2635,18 @@ void nr_construct_HhH_elements(int *conjch00_ch00, uint8_t nr_zero_forcing_rx_2layers(int **rxdataF_comp, int **dl_ch_mag, int **dl_ch_magb, + int **dl_ch_magr, int **dl_ch_estimates_ext, unsigned short nb_rb, unsigned char n_rx, unsigned char mod_order, int shift, - unsigned char symbol) + unsigned char symbol, + int length) { int *ch00, *ch01, *ch10, *ch11; int *ch20, *ch30, *ch21, *ch31; - + uint32_t nb_rb_0 = length/12 + ((length%12)?1:0); /* we need at least alignment to 16 bytes, let's put 32 to be sure * (maybe not necessary but doesn't hurt) */ @@ -2671,49 +2714,49 @@ uint8_t nr_zero_forcing_rx_2layers(int **rxdataF_comp, nr_conjch0_mult_ch1(ch00, ch00, conjch00_ch00, - nb_rb, + nb_rb_0, shift); // (1/2^log2_maxh)*conj_H_10xH_10: (1/(64*2))conjH_10*H_10*2^15 nr_conjch0_mult_ch1(ch10, ch10, conjch10_ch10, - nb_rb, + nb_rb_0, shift); // conj_H_00xH_01 nr_conjch0_mult_ch1(ch00, ch01, conjch00_ch01, - nb_rb, + nb_rb_0, shift); // this shift is equal to the channel level log2_maxh // conj_H_10xH_11 nr_conjch0_mult_ch1(ch10, ch11, conjch10_ch11, - nb_rb, + nb_rb_0, shift); // conj_H_01xH_01 nr_conjch0_mult_ch1(ch01, ch01, conjch01_ch01, - nb_rb, + nb_rb_0, shift); // conj_H_11xH_11 nr_conjch0_mult_ch1(ch11, ch11, conjch11_ch11, - nb_rb, + nb_rb_0, shift); // conj_H_01xH_00 nr_conjch0_mult_ch1(ch01, ch00, conjch01_ch00, - nb_rb, + nb_rb_0, shift); // conj_H_11xH_10 nr_conjch0_mult_ch1(ch11, ch10, conjch11_ch10, - nb_rb, + nb_rb_0, shift); } if (n_rx==4){ @@ -2721,52 +2764,52 @@ uint8_t nr_zero_forcing_rx_2layers(int **rxdataF_comp, nr_conjch0_mult_ch1(ch20, ch20, conjch20_ch20, - nb_rb, + nb_rb_0, shift); // (1/2^log2_maxh)*conj_H_30xH_30: (1/(64*2*4))conjH_30*H_30*2^15 nr_conjch0_mult_ch1(ch30, ch30, conjch30_ch30, - nb_rb, + nb_rb_0, shift); // (1/2^log2_maxh)*conj_H_20xH_20: (1/(64*2))conjH_20*H_20*2^15 nr_conjch0_mult_ch1(ch20, ch21, conjch20_ch21, - nb_rb, + nb_rb_0, shift); nr_conjch0_mult_ch1(ch30, ch31, conjch30_ch31, - nb_rb, + nb_rb_0, shift); nr_conjch0_mult_ch1(ch21, ch21, conjch21_ch21, - nb_rb, + nb_rb_0, shift); nr_conjch0_mult_ch1(ch31, ch31, conjch31_ch31, - nb_rb, + nb_rb_0, shift); // (1/2^log2_maxh)*conj_H_20xH_20: (1/(64*2))conjH_20*H_20*2^15 nr_conjch0_mult_ch1(ch21, ch20, conjch21_ch20, - nb_rb, + nb_rb_0, shift); nr_conjch0_mult_ch1(ch31, ch30, conjch31_ch30, - nb_rb, + nb_rb_0, shift); nr_construct_HhH_elements(conjch00_ch00, @@ -2789,7 +2832,7 @@ uint8_t nr_zero_forcing_rx_2layers(int **rxdataF_comp, af_mf_01, af_mf_10, af_mf_11, - nb_rb, + nb_rb_0, symbol); } if (n_rx==2){ @@ -2813,7 +2856,7 @@ uint8_t nr_zero_forcing_rx_2layers(int **rxdataF_comp, af_mf_01, af_mf_10, af_mf_11, - nb_rb, + nb_rb_0, symbol); } //det_HhH = ad -bc @@ -2822,8 +2865,9 @@ uint8_t nr_zero_forcing_rx_2layers(int **rxdataF_comp, af_mf_10,//c af_mf_11,//d determ_fin, - nb_rb, - symbol); + nb_rb_0, + symbol, + shift); /* 2- Compute the channel matrix inversion ********************************** * * |(conj_H_00xH_00+conj_H_10xH_10) (conj_H_00xH_01+conj_H_10xH_11)| @@ -2837,8 +2881,8 @@ uint8_t nr_zero_forcing_rx_2layers(int **rxdataF_comp, * * **************************************************************************/ - __m128i *rxdataF_comp128_0,*rxdataF_comp128_1,*dl_ch_mag128_0,*dl_ch_mag128b_0,*dl_ch_mag128_1,*dl_ch_mag128b_1,*determ_fin_128; - __m128i mmtmpD2,mmtmpD3,mmtmpD0,mmtmpD1,QAM_amp128,QAM_amp128b; + __m128i *rxdataF_comp128_0,*rxdataF_comp128_1,*dl_ch_mag128_0,*dl_ch_mag128b_0,*dl_ch_mag128r_0,*dl_ch_mag128_1,*dl_ch_mag128b_1,*dl_ch_mag128r_1,*determ_fin_128; + __m128i mmtmpD0,mmtmpD1;//mmtmpD2,mmtmpD3, __m128i *after_mf_a_128,*after_mf_b_128, *after_mf_c_128, *after_mf_d_128; determ_fin_128 = (__m128i *)&determ_fin[0]; @@ -2850,79 +2894,56 @@ uint8_t nr_zero_forcing_rx_2layers(int **rxdataF_comp, dl_ch_mag128_1 = (__m128i *)&dl_ch_mag[n_rx][symbol*nb_rb*12]; dl_ch_mag128b_0 = (__m128i *)&dl_ch_magb[0][symbol*nb_rb*12]; dl_ch_mag128b_1 = (__m128i *)&dl_ch_magb[n_rx][symbol*nb_rb*12]; + dl_ch_mag128r_0 = (__m128i *)&dl_ch_magr[0][symbol*nb_rb*12]; + dl_ch_mag128r_1 = (__m128i *)&dl_ch_magr[n_rx][symbol*nb_rb*12]; after_mf_a_128 = (__m128i *)af_mf_00; after_mf_b_128 = (__m128i *)af_mf_01; after_mf_c_128 = (__m128i *)af_mf_10; after_mf_d_128 = (__m128i *)af_mf_11; - QAM_amp128b = _mm_setzero_si128(); - - if (mod_order == 4) { - QAM_amp128 = _mm_set1_epi16(QAM16_n1); // 2/sqrt(10) - QAM_amp128b = _mm_setzero_si128(); - } else if (mod_order == 6) { - QAM_amp128 = _mm_set1_epi16(QAM64_n1); // - QAM_amp128b = _mm_set1_epi16(QAM64_n2); - } - - for (int rb=0; rb<3*nb_rb; rb++) { + for (int rb=0; rb<3*nb_rb_0; rb++) { if (mod_order>2) { - // get channel determ (da -bc) if not QPSK - mmtmpD0 = _mm_packs_epi32(determ_fin_128[0],determ_fin_128[0]);//convert 32 bits to 16 bits - - dl_ch_mag128_0[0] = _mm_unpacklo_epi16(mmtmpD0,mmtmpD0); - dl_ch_mag128b_0[0] = dl_ch_mag128_0[0]; - - dl_ch_mag128_0[0] = _mm_mulhi_epi16(dl_ch_mag128_0[0],QAM_amp128); - dl_ch_mag128_0[0] = _mm_slli_epi16(dl_ch_mag128_0[0],1);//aatx=0 @ aarx =0 - dl_ch_mag128_1[0] = dl_ch_mag128_0[0];//aatx=1 @ aarx =0 - - dl_ch_mag128b_0[0] = _mm_mulhi_epi16(dl_ch_mag128b_0[0],QAM_amp128b); - dl_ch_mag128b_0[0] = _mm_slli_epi16(dl_ch_mag128b_0[0],1);//aatx=0 @ aarx = - dl_ch_mag128b_1[0] = dl_ch_mag128b_0[0];//aatx=1 @ aarx =0 - - if ((rb==0)&&(symbol==1)) { - printf("\n Signal mag after ZF \n"); - print_shorts("mag layer 1:",(int16_t*)&dl_ch_mag128_0[0]); - print_shorts("mag layer 2:",(int16_t*)&dl_ch_mag128_1[0]); - print_shorts("magb layer 1:",(int16_t*)&dl_ch_mag128b_0[0]); - print_shorts("magb layer 2:",(int16_t*)&dl_ch_mag128b_1[0]); - } + //The update value of dl_ch_mag128, dl_ch_mag128b, and dl_ch_mag128r shall go here + //printf("\n Signal mag after ZF \n"); + //print_shorts("mag layer 1:",(int16_t*)&dl_ch_mag128_0[0]); + //print_shorts("mag layer 2:",(int16_t*)&dl_ch_mag128_1[0]); + //print_shorts("magb layer 1:",(int16_t*)&dl_ch_mag128b_0[0]); + //print_shorts("magb layer 2:",(int16_t*)&dl_ch_mag128b_1[0]); + //print_shorts("magr layer 1:",(int16_t*)&dl_ch_mag128r_0[0]); + //print_shorts("magr layer 2:",(int16_t*)&dl_ch_mag128r_1[0]); } // multiply by channel Inv //rxdataF_zf128_0 = rxdataF_comp128_0*d - b*rxdataF_comp128_1 //rxdataF_zf128_1 = rxdataF_comp128_1*a - c*rxdataF_comp128_0 - mmtmpD2 = nr_inv_comp_muli(rxdataF_comp128_0[0],//x - after_mf_d_128[0]);//y - mmtmpD3 = nr_inv_comp_muli(rxdataF_comp128_1[0],//x - after_mf_b_128[0]);//y - //mmtmpD2[0] - mmtmpD3[0] - //rxdataF_zf128_0 - mmtmpD0 = _mm_sub_epi16(mmtmpD2, mmtmpD3); + //printf("layer_1 \n"); + mmtmpD0 = nr_comp_muli_sum(rxdataF_comp128_0[0], + after_mf_d_128[0], + rxdataF_comp128_1[0], + after_mf_b_128[0], + determ_fin_128[0]); - mmtmpD2 = nr_inv_comp_muli(rxdataF_comp128_1[0],//x - after_mf_a_128[0]);//y - mmtmpD3 = nr_inv_comp_muli(rxdataF_comp128_0[0],//x - after_mf_c_128[0]);//y - //mmtmpD2[0] - mmtmpD3[0] - //rxdataF_zf128_1 - mmtmpD1 = _mm_sub_epi16(mmtmpD2, mmtmpD3); + //printf("layer_2 \n"); + mmtmpD1 = nr_comp_muli_sum(rxdataF_comp128_1[0], + after_mf_a_128[0], + rxdataF_comp128_0[0], + after_mf_c_128[0], + determ_fin_128[0]); rxdataF_comp128_0[0] = mmtmpD0; rxdataF_comp128_1[0] = mmtmpD1; - - /*if ((rb==0)&&(symbol==1)) { - printf("\n Rx signal after ZF \n"); - print_shorts("Rx layer 1:",(int16_t*)&rxdataF_comp128_0[0]); - print_shorts("Rx layer 2:",(int16_t*)&rxdataF_comp128_1[0]); - }*/ - +#ifdef DEBUG_DLSCH_DEMOD + printf("\n Rx signal after ZF l%d rb%d\n",symbol,rb); + print_shorts(" Rx layer 1:",(int16_t*)&rxdataF_comp128_0[0]); + print_shorts(" Rx layer 2:",(int16_t*)&rxdataF_comp128_1[0]); +#endif determ_fin_128 += 1; dl_ch_mag128_0 += 1; dl_ch_mag128b_0 += 1; + dl_ch_mag128r_0 += 1; dl_ch_mag128_1 += 1; dl_ch_mag128b_1 += 1; + dl_ch_mag128r_1 += 1; rxdataF_comp128_0 += 1; rxdataF_comp128_1 += 1; after_mf_a_128 += 1; diff --git a/openair1/PHY/NR_UE_TRANSPORT/nr_transport_proto_ue.h b/openair1/PHY/NR_UE_TRANSPORT/nr_transport_proto_ue.h index 6b76645fef..eb56c76366 100644 --- a/openair1/PHY/NR_UE_TRANSPORT/nr_transport_proto_ue.h +++ b/openair1/PHY/NR_UE_TRANSPORT/nr_transport_proto_ue.h @@ -804,7 +804,7 @@ void nr_dlsch_channel_compensation(int32_t **rxdataF_ext, NR_DL_FRAME_PARMS *frame_parms, uint8_t nb_aatx, uint8_t symbol, - uint8_t start_symbol, + int length, uint8_t first_symbol_flag, uint8_t mod_order, uint16_t nb_rb, @@ -888,13 +888,15 @@ void nr_dlsch_channel_level_median(int **dl_ch_estimates_ext, int start_point); void nr_dlsch_detection_mrc(int **rxdataF_comp, - int ***rho, - int **dl_ch_mag, - int **dl_ch_magb, - short n_tx, - short n_rx, - unsigned char symbol, - unsigned short nb_rb); + int ***rho, + int **dl_ch_mag, + int **dl_ch_magb, + int **dl_ch_magr, + short n_tx, + short n_rx, + unsigned char symbol, + unsigned short nb_rb, + int length); void nr_dlsch_detection_mrc_core(int **rxdataF_comp, int **rxdataF_comp_i,