diff --git a/lte/phy/include/liblte/phy/phch/pusch.h b/lte/phy/include/liblte/phy/phch/pusch.h
index ccbb1b386..dbdd12eb6 100644
--- a/lte/phy/include/liblte/phy/phch/pusch.h
+++ b/lte/phy/include/liblte/phy/phch/pusch.h
@@ -60,8 +60,6 @@ typedef struct LIBLTE_API {
   cf_t *pusch_d;
 
   void *pusch_q;
-  uint8_t *pusch_g_ri;
-  uint8_t *pusch_g_ack;
   void *pusch_g;
 
   /* tx & rx objects */
diff --git a/lte/phy/include/liblte/phy/phch/sch.h b/lte/phy/include/liblte/phy/phch/sch.h
index 519e3eaea..632c8be23 100644
--- a/lte/phy/include/liblte/phy/phch/sch.h
+++ b/lte/phy/include/liblte/phy/phch/sch.h
@@ -105,8 +105,6 @@ LIBLTE_API int ulsch_uci_encode(sch_t *q,
                                 uint8_t *data, 
                                 uci_data_t uci_data, 
                                 uint8_t *g_bits,
-                                uint8_t *g_bits_ack, 
-                                uint8_t *g_bits_ri,
                                 harq_t *harq_process, 
                                 uint32_t rv_idx, 
                                 uint8_t *q_bits);
diff --git a/lte/phy/include/liblte/phy/phch/uci.h b/lte/phy/include/liblte/phy/phch/uci.h
index b49663d9c..d4390903a 100644
--- a/lte/phy/include/liblte/phy/phch/uci.h
+++ b/lte/phy/include/liblte/phy/phch/uci.h
@@ -66,11 +66,19 @@ LIBLTE_API int uci_encode_cqi(uci_cqi_t *q,
                               harq_t *harq_process, 
                               uint8_t *q_bits);
 
-/* Encode UCI RI and HARQ ACK/NACK bits */
-LIBLTE_API int uci_encode_ri_ack(uint8_t data,
-                                 uint32_t O_cqi,
-                                 float beta, 
-                                 harq_t *harq_process,
-                                 uint8_t q_bits[6]); 
+LIBLTE_API int uci_encode_ack(uint8_t data,
+                              uint32_t O_cqi,
+                              float beta, 
+                              harq_t *harq_process,
+                              uint32_t H_prime_total, 
+                              uint8_t *q_bits); 
+
+LIBLTE_API int uci_encode_ri(uint8_t data,
+                             uint32_t O_cqi,
+                             float beta, 
+                             harq_t *harq_process,
+                             uint32_t H_prime_total, 
+                             uint8_t *q_bits); 
+
 
 #endif
\ No newline at end of file
diff --git a/lte/phy/lib/phch/src/pusch.c b/lte/phy/lib/phch/src/pusch.c
index a14d2bf1d..6caa342b4 100644
--- a/lte/phy/lib/phch/src/pusch.c
+++ b/lte/phy/lib/phch/src/pusch.c
@@ -131,17 +131,6 @@ int pusch_init(pusch_t *q, lte_cell_t cell) {
     if (!q->pusch_g) {
       goto clean;
     }
-
-    // Allocate buffers for q bits for coded RI and ACK bits 
-    q->pusch_g_ack = vec_malloc(sizeof(uint8_t) * 4 * q->cell.nof_prb * lte_mod_bits_x_symbol(LTE_QAM64));
-    if (!q->pusch_g_ack) {
-      goto clean;
-    }
-    q->pusch_g_ri = vec_malloc(sizeof(uint8_t) * 4 * q->cell.nof_prb * lte_mod_bits_x_symbol(LTE_QAM64));
-    if (!q->pusch_g_ri) {
-      goto clean;
-    }
-
     q->pusch_d = vec_malloc(sizeof(cf_t) * q->max_symbols);
     if (!q->pusch_d) {
       goto clean;
@@ -183,12 +172,6 @@ void pusch_free(pusch_t *q) {
   if (q->pusch_g) {
     free(q->pusch_g);
   }
-  if (q->pusch_g_ack) {
-    free(q->pusch_g_ack);
-  }
-  if (q->pusch_g_ri) {
-    free(q->pusch_g_ri);
-  }
   for (i = 0; i < q->cell.nof_ports; i++) {
     if (q->ce[i]) {
       free(q->ce[i]);
@@ -363,8 +346,7 @@ int pusch_uci_encode(pusch_t *q, uint8_t *data, uci_data_t uci_data,
       }
       memset(&x[q->cell.nof_ports], 0, sizeof(cf_t*) * (MAX_LAYERS - q->cell.nof_ports));
       
-      if (ulsch_uci_encode(&q->dl_sch, data, uci_data, q->pusch_g, 
-        q->pusch_g_ack, q->pusch_g_ri, harq_process, rv_idx, q->pusch_q)) 
+      if (ulsch_uci_encode(&q->dl_sch, data, uci_data, q->pusch_g, harq_process, rv_idx, q->pusch_q)) 
       {
         fprintf(stderr, "Error encoding TB\n");
         return LIBLTE_ERROR;
diff --git a/lte/phy/lib/phch/src/sch.c b/lte/phy/lib/phch/src/sch.c
index ca81fc5e3..0544d68c9 100644
--- a/lte/phy/lib/phch/src/sch.c
+++ b/lte/phy/lib/phch/src/sch.c
@@ -403,192 +403,27 @@ int ulsch_decode(sch_t *q, float *e_bits, uint8_t *data, uint32_t tbs, uint32_t
 }
 
 
-uint8_t ulsch_y_idx[10000];
-uint8_t ulsch_y_mat[10000];
-
 /* UL-SCH channel interleaver according to 5.5.2.8 of 36.212 */
-void ulsch_interleave2(uint8_t *q_bits, uint32_t nb_q, 
-                      uint8_t q_bits_ack[6], uint32_t Q_prime_ack, 
-                      uint8_t q_bits_ri[6], uint32_t Q_prime_ri,
-                      uint32_t Q_m, uint8_t *g_bits) 
+void ulsch_interleave(uint8_t *g_bits, uint32_t Q_m, uint32_t H_prime_total, uint32_t N_pusch_symbs, uint8_t *q_bits) 
 {
-  uint32_t C_mux;
-  uint32_t H_prime;
-  uint32_t H_prime_total;
-  uint32_t R_mux;
-  uint32_t R_prime_mux;
-  uint32_t i;
-  uint32_t j;
-  uint32_t k;
-  uint32_t r;
-  uint32_t idx;
-  uint32_t ri_column_set[4]  = {1, 4, 7, 10};
-  uint32_t ack_column_set[4] = {2, 3, 8, 9};
-  uint32_t C_ri;
-  uint32_t C_ack;
-  uint32_t N_pusch_symbs = 12; 
   
-  // Step 1: Define C_mux
-  C_mux = N_pusch_symbs;
-
-  // Step 2: Define R_mux and R_prime_mux
-  H_prime       = nb_q;
-  H_prime_total = H_prime + Q_prime_ri;
-  R_mux         = (H_prime_total*Q_m)/C_mux;
-  R_prime_mux   = R_mux/Q_m;
- 
-
-  // Initialize the matricies
-  printf("Cmux*R_prime=%d*%d=%d, H_prime=%d, H_prime_total=%d\n",C_mux, R_prime_mux, C_mux*R_prime_mux, H_prime, H_prime_total);
-  for(i=0; i<C_mux*R_prime_mux; i++) {
-    ulsch_y_idx[i] = 100;
-  }
-  for(i=0; i<C_mux*R_mux; i++) {
-    ulsch_y_mat[i] = 0;
-  }
-
-  // Step 3: Interleave the RI control bits
-  i = 0;
-  j = 0;
-  r = R_prime_mux-1;
-  while(i < Q_prime_ri) {
-    C_ri = ri_column_set[j];
-    ulsch_y_idx[r*C_mux + C_ri] = 1;
-    for(k=0; k<Q_m; k++) {
-      ulsch_y_mat[(C_mux*r*Q_m) + C_ri*Q_m + k] = q_bits_ri[Q_m*i+k];
-    }
-    i++;
-    r = R_prime_mux - 1 - i/4;
-    j = (j + 3) % 4;
-  }
-  printf("H_prime: %d, C_mux: %d, R_prime: %d\n", H_prime, C_mux, R_prime_mux);
+  uint32_t rows = H_prime_total/N_pusch_symbs;
+  uint32_t cols = N_pusch_symbs;
   
-  // Step 4: Interleave the data bits
-  i = 0;
-  k = 0;
-  while(k < H_prime) {
-    if(ulsch_y_idx[i] == 100) {
-      ulsch_y_idx[i] = 1;
-      for(j=0; j<Q_m; j++) {
-        ulsch_y_mat[i*Q_m + j] = q_bits[Q_m*k+j];
-      }
-      k++;
-    }
-    i++;
-  }
-
-  // Step 5: Interleave the ACK control bits
-  i = 0;
-  j = 0;
-  r = R_prime_mux-1;
-  while(i < Q_prime_ack) {
-    C_ack = ack_column_set[j];
-    ulsch_y_idx[r*C_mux + C_ack] = 2;
-    for(k=0; k<Q_m; k++) {
-      ulsch_y_mat[(C_mux*r*Q_m) + C_ack*Q_m + k] = q_bits_ack[Q_m*i+k];
-    }
-    i++;
-    r = R_prime_mux - 1 - i/4;
-    j = (j + 3) % 4;
-  }
-
-  // Step 6: Read out the bits
-  idx = 0;
-  printf("go for C_mux: %d, R_prime: %d, Q_m: %d\n", C_mux, R_prime_mux, Q_m);
-  for(i=0; i<C_mux; i++) {
-    for(j=0; j<R_prime_mux; j++) {
-      for(k=0; k<Q_m; k++) {
-        g_bits[idx++] = ulsch_y_mat[j*C_mux*Q_m + i*Q_m + k];          
-      }
-    }
-  }
-  
-}
-
-/* UL-SCH channel interleaver according to 5.5.2.8 of 36.212 */
-void ulsch_interleave(uint8_t *g_bits, uint32_t nb_q, 
-                      uint8_t g_bits_ack[6], uint32_t Q_prime_ack, 
-                      uint8_t g_bits_ri[6], uint32_t Q_prime_ri,
-                      uint32_t Q_m,
-                     uint8_t *q_bits) 
-{
-  uint32_t C_mux;
-  uint32_t H_prime;
-  uint32_t H_prime_total;
-  uint32_t R_mux;
-  uint32_t R_prime_mux;
-  uint32_t i;
-  uint32_t j;
-  uint32_t k;
-  uint32_t r;
-  uint32_t idx;
-  uint32_t ri_column_set[4]  = {1, 4, 7, 10};
-  uint32_t ack_column_set[4] = {2, 3, 8, 9};
-  uint32_t C_ri;
-  uint32_t C_ack;
-  uint32_t N_pusch_symbs = 12; 
-  
-  // Step 1: Define C_mux
-  C_mux = N_pusch_symbs;
-
-  // Step 2: Define R_mux and R_prime_mux
-  H_prime       = nb_q;
-  H_prime_total = H_prime + Q_prime_ri;
-  R_mux         = (H_prime_total*Q_m)/C_mux;
-  R_prime_mux   = R_mux/Q_m;
-
-  
-  
-  // ACK insertion can be done at uci.c 
-  
-  // Step 5: Interleave the ACK control bits
-  i = 0;
-  j = 0;
-  r = R_prime_mux-1;
-  while(i < Q_prime_ack) {
-    C_ack = ack_column_set[j];
-    for(k=0; k<Q_m; k++) {
-      g_bits[(C_mux*r*Q_m) + C_ack*Q_m + k] = g_bits_ack[Q_m*i+k];
-    }
-    i++;
-    r = R_prime_mux - 1 - i/4;
-    j = (j + 3) % 4;
-  }
-
-  // Step 3: Interleave the RI control bits
-  i = 0;
-  j = 0;
-  r = R_prime_mux-1;
-  while(i < Q_prime_ri) {
-    C_ri = ri_column_set[j];
-    ulsch_y_idx[r*C_mux + C_ri] = 1;
-    for(k=0; k<Q_m; k++) {
-      q_bits[(r*Q_m) + C_mux*C_ri*Q_m + k] = 10+g_bits_ri[Q_m*i+k];
-    }
-    i++;
-    r = R_prime_mux - 1 - i/4;
-    j = (j + 3) % 4;
-  }
-  
-  // Step 6: Read out the bits
-  idx = 0;
-  printf("go for C_mux: %d, R_prime: %d, Q_m: %d\n", C_mux, R_prime_mux, Q_m);
-  for(i=0; i<C_mux; i++) {
-    for(j=0; j<R_prime_mux; j++) {
-      for(k=0; k<Q_m; k++) {
-        if (q_bits[idx] >= 10) {
-          printf("10 at %d is %d\n",idx, q_bits[idx]);
-          //q_bits[idx] -= 10;          
-        } else {          
-          printf("%d->%d\n", j*C_mux*Q_m + i*Q_m + k,idx);
-          q_bits[idx] = g_bits[j*C_mux*Q_m + i*Q_m + k];                       
+  uint32_t idx = 0;
+  for(uint32_t j=0; j<rows; j++) {        
+    for(uint32_t i=0; i<cols; i++) {
+      for(uint32_t k=0; k<Q_m; k++) {
+        if (q_bits[j*Q_m + i*rows*Q_m + k] >= 10) {
+          q_bits[j*Q_m + i*rows*Q_m + k] -= 10;
+        } else {
+          q_bits[j*Q_m + i*rows*Q_m + k] = g_bits[idx];                                
+          idx++;                  
         }
-        idx++;
       }
     }
   }
   
-  
 }
 
 int ulsch_encode(sch_t *q, uint8_t *data, uint8_t *g_bits, 
@@ -596,12 +431,11 @@ int ulsch_encode(sch_t *q, uint8_t *data, uint8_t *g_bits,
 {
   uci_data_t uci_data; 
   bzero(&uci_data, sizeof(uci_data_t));
-  return ulsch_uci_encode(q, data, uci_data, g_bits, NULL, NULL, harq_process, rv_idx, q_bits);
+  return ulsch_uci_encode(q, data, uci_data, g_bits, harq_process, rv_idx, q_bits);
 }
 
 
 int ulsch_uci_encode(sch_t *q, uint8_t *data, uci_data_t uci_data, uint8_t *g_bits, 
-                 uint8_t *g_bits_ack, uint8_t *g_bits_ri, 
                  harq_t *harq_process, uint32_t rv_idx, uint8_t *q_bits) 
 {
   int ret; 
@@ -615,26 +449,15 @@ int ulsch_uci_encode(sch_t *q, uint8_t *data, uci_data_t uci_data, uint8_t *g_bi
   uint32_t nof_symbols = 12*harq_process->prb_alloc.slot[0].nof_prb*RE_X_RB;
   uint32_t nb_q = nof_symbols * Q_m;
 
- // Encode ACK
-  if (uci_data.uci_ack_len > 0) {
-    float beta = uci_data.beta_ack; 
-    if (harq_process->mcs.tbs == 0) {
-        beta /= uci_data.beta_cqi;
-    }
-    ret = uci_encode_ri_ack(uci_data.uci_ack, uci_data.uci_cqi_len, beta, harq_process, g_bits_ack);
-    if (ret < 0) {
-      return ret; 
-    }
-    Q_prime_ack = (uint32_t) ret; 
-  }
-    
+  bzero(q_bits, sizeof(uint8_t) * nb_q);
+
   // Encode RI
   if (uci_data.uci_ri_len > 0) {
     float beta = uci_data.beta_ri; 
     if (harq_process->mcs.tbs == 0) {
         beta /= uci_data.beta_cqi;
     }
-    ret = uci_encode_ri_ack(uci_data.uci_ri, uci_data.uci_cqi_len, beta, harq_process, g_bits_ri);
+    ret = uci_encode_ri(uci_data.uci_ri, uci_data.uci_cqi_len, beta, harq_process, nb_q/Q_m, q_bits);
     if (ret < 0) {
       return ret; 
     }
@@ -643,6 +466,7 @@ int ulsch_uci_encode(sch_t *q, uint8_t *data, uci_data_t uci_data, uint8_t *g_bi
 
   // Encode CQI
   if (uci_data.uci_cqi_len > 0) {
+
     ret = uci_encode_cqi(&q->uci_cqi, uci_data.uci_cqi, uci_data.uci_cqi_len, uci_data.beta_cqi, 
                                  Q_prime_ri, harq_process, g_bits);
     if (ret < 0) {
@@ -652,8 +476,7 @@ int ulsch_uci_encode(sch_t *q, uint8_t *data, uci_data_t uci_data, uint8_t *g_bi
   }
   
   e_offset += Q_prime_cqi*Q_m;
-
-  printf("Q_prime_ack=%d, Q_prime_cqi=%d, Q_prime_ri=%d\n",Q_prime_ack, Q_prime_cqi, Q_prime_ri);  
+  
   // Encode UL-SCH
   if (harq_process->mcs.tbs > 0) {
     uint32_t G = nb_q/Q_m - Q_prime_ri - Q_prime_cqi;     
@@ -663,12 +486,24 @@ int ulsch_uci_encode(sch_t *q, uint8_t *data, uci_data_t uci_data, uint8_t *g_bi
       return ret; 
     }    
   } 
-
-  // Multiplexing and Interleaving 
-  ulsch_interleave(g_bits, nb_q/Q_m-Q_prime_ri, 
-                   g_bits_ack, Q_prime_ack,
-                    g_bits_ri, Q_prime_ri, 
-                   Q_m, q_bits);
+    
+  // Interleave UL-SCH (and RI and CQI)
+  ulsch_interleave(g_bits, Q_m, nb_q/Q_m, harq_process->N_symb_ul, q_bits);
+  
+   // Encode (and interleave) ACK
+  if (uci_data.uci_ack_len > 0) {
+    float beta = uci_data.beta_ack; 
+    if (harq_process->mcs.tbs == 0) {
+        beta /= uci_data.beta_cqi;
+    }
+    ret = uci_encode_ack(uci_data.uci_ack, uci_data.uci_cqi_len, beta, harq_process, nb_q/Q_m, q_bits);
+    if (ret < 0) {
+      return ret; 
+    }
+    Q_prime_ack = (uint32_t) ret; 
+  }
+    
+  INFO("Q_prime_ack=%d, Q_prime_cqi=%d, Q_prime_ri=%d\n",Q_prime_ack, Q_prime_cqi, Q_prime_ri);  
 
 
 
diff --git a/lte/phy/lib/phch/src/uci.c b/lte/phy/lib/phch/src/uci.c
index 34778e127..750cd9543 100644
--- a/lte/phy/lib/phch/src/uci.c
+++ b/lte/phy/lib/phch/src/uci.c
@@ -195,6 +195,55 @@ int uci_encode_cqi(uci_cqi_t *q, uint8_t *cqi_data, uint32_t cqi_len, float beta
   }
 }
 
+/* Inserts UCI-ACK bits into the correct positions in the g buffer before interleaving */
+static int uci_ulsch_interleave_ack(uint8_t ack_coded_bits[6], uint32_t ack_q_bit_idx, 
+                          uint32_t Q_m, uint32_t H_prime_total, uint32_t N_pusch_symbs, lte_cp_t cp,
+                          uint8_t *q_bits) {
+
+  const uint32_t ack_column_set_norm[4] = {2, 3, 8, 9};
+  const uint32_t ack_column_set_ext[4] = {1, 2, 6, 7};
+  
+  if (H_prime_total/N_pusch_symbs > 1+ack_q_bit_idx/4) {
+    uint32_t row = H_prime_total/N_pusch_symbs-1-ack_q_bit_idx/4;
+    uint32_t colidx = (3*ack_q_bit_idx)%4;
+    uint32_t col = CP_ISNORM(cp)?ack_column_set_norm[colidx]:ack_column_set_ext[colidx];
+    for(uint32_t k=0; k<Q_m; k++) {
+      q_bits[row *Q_m + 
+             (H_prime_total/N_pusch_symbs)*col*Q_m + k] = ack_coded_bits[k];
+    }    
+    return LIBLTE_SUCCESS;
+  } else {
+    fprintf(stderr, "Error interleaving UCI-ACK bit idx %d for H_prime_total=%d and N_pusch_symbs=%d\n",
+            ack_q_bit_idx, H_prime_total, N_pusch_symbs);
+    return LIBLTE_ERROR;
+  }
+}
+
+/* Inserts UCI-RI bits into the correct positions in the g buffer before interleaving */
+static int uci_ulsch_interleave_ri(uint8_t ri_coded_bits[6], uint32_t ri_q_bit_idx, 
+                          uint32_t Q_m, uint32_t H_prime_total, uint32_t N_pusch_symbs, lte_cp_t cp,
+                          uint8_t *q_bits) {
+  
+  static uint32_t ri_column_set_norm[4]  = {1, 4, 7, 10};
+  static uint32_t ri_column_set_ext[4]  = {0, 3, 5, 8};
+
+  if (H_prime_total/N_pusch_symbs > 1+ri_q_bit_idx/4) {
+    uint32_t row = H_prime_total/N_pusch_symbs-1-ri_q_bit_idx/4;
+    uint32_t colidx = (3*ri_q_bit_idx)%4;
+    uint32_t col = CP_ISNORM(cp)?ri_column_set_norm[colidx]:ri_column_set_ext[colidx];
+    printf("r=%d-%d\n",H_prime_total/N_pusch_symbs,1+ri_q_bit_idx/4);
+    for(uint32_t k=0; k<Q_m; k++) {
+      q_bits[row *Q_m + (H_prime_total/N_pusch_symbs)*col*Q_m + k] = 10+ri_coded_bits[k];
+    }    
+    return LIBLTE_SUCCESS;
+  } else {
+    fprintf(stderr, "Error interleaving UCI-RI bit idx %d for H_prime_total=%d and N_pusch_symbs=%d\n",
+            ri_q_bit_idx, H_prime_total, N_pusch_symbs);
+    return LIBLTE_ERROR;
+  }
+
+}
+
 static uint32_t Q_prime_ri_ack(uint32_t O, uint32_t O_cqi, float beta, harq_t *harq_process) {
   uint32_t M_sc = harq_process->prb_alloc.slot[0].nof_prb * RE_X_RB;
   
@@ -219,23 +268,46 @@ static uint32_t Q_prime_ri_ack(uint32_t O, uint32_t O_cqi, float beta, harq_t *h
   return Q_prime; 
 }
 
-/* Encode UCI RI and HARQ bits as described in 5.2.2.6 of 36.212 
- *  Currently only supporting 1-bit RI or 1-bit HARQ
- */
-int uci_encode_ri_ack(uint8_t data, uint32_t O_cqi, float beta, harq_t *harq_process, uint8_t *q_bits)
-{
-  uint32_t Q_m = lte_mod_bits_x_symbol(harq_process->mcs.mod);
-  
-  q_bits[0] = data;
-  q_bits[1] = 2; 
+static void encode_ri_ack(uint8_t data, uint8_t q_encoded_bits[6], uint8_t Q_m) {
+  q_encoded_bits[0] = data;
+  q_encoded_bits[1] = 2; 
   for (uint32_t i=2;i<Q_m;i++) {
-    q_bits[i] = 3;
+    q_encoded_bits[i] = 3;
   }
-  
-  uint32_t Qprime = Q_prime_ri_ack(1, O_cqi, beta, harq_process);
+}
 
-  for (uint32_t i=1;i<Qprime;i++) {
-    memcpy(&q_bits[i*Q_m], q_bits, Q_m*sizeof(uint8_t));  
+/* Encode UCI HARQ/ACK bits as described in 5.2.2.6 of 36.212 
+ *  Currently only supporting 1-bit HARQ
+ */
+int uci_encode_ack(uint8_t data, uint32_t O_cqi, float beta, harq_t *harq_process, uint32_t H_prime_total, uint8_t *q_bits)
+{
+  uint32_t Q_m = lte_mod_bits_x_symbol(harq_process->mcs.mod);  
+  uint32_t Qprime = Q_prime_ri_ack(1, O_cqi, beta, harq_process);
+  uint8_t q_encoded_bits[6];
+
+  encode_ri_ack(data, q_encoded_bits, Q_m);
+  
+  for (uint32_t i=0;i<Qprime;i++) {
+    uci_ulsch_interleave_ack(q_encoded_bits, i, Q_m, H_prime_total, harq_process->N_symb_ul, harq_process->cell.cp, q_bits);
+  }
+  
+  return (int) Qprime;
+}
+
+
+/* Encode UCI RI bits as described in 5.2.2.6 of 36.212 
+ *  Currently only supporting 1-bit RI
+ */
+int uci_encode_ri(uint8_t data, uint32_t O_cqi, float beta, harq_t *harq_process, uint32_t H_prime_total, uint8_t *q_bits)
+{
+  uint32_t Q_m = lte_mod_bits_x_symbol(harq_process->mcs.mod);  
+  uint32_t Qprime = Q_prime_ri_ack(1, O_cqi, beta, harq_process);
+  uint8_t q_encoded_bits[6];
+
+  encode_ri_ack(data, q_encoded_bits, Q_m);
+  
+  for (uint32_t i=0;i<Qprime;i++) {
+    uci_ulsch_interleave_ri(q_encoded_bits, i, Q_m, H_prime_total, harq_process->N_symb_ul, harq_process->cell.cp, q_bits);
   }
   
   return (int) Qprime;
diff --git a/lte/phy/lib/phch/test/pusch_test.c b/lte/phy/lib/phch/test/pusch_test.c
index 06e4ffe53..87231ab6c 100644
--- a/lte/phy/lib/phch/test/pusch_test.c
+++ b/lte/phy/lib/phch/test/pusch_test.c
@@ -201,8 +201,9 @@ int main(int argc, char **argv) {
     uci_data.uci_cqi = tmp;
     uci_data.uci_cqi_len = 0; 
     uci_data.uci_ri_len = 1; 
-    uci_data.uci_ack_len = 0; 
     uci_data.uci_ri = 1; 
+    uci_data.uci_ack_len = 1; 
+    uci_data.uci_ack = 1; 
     
     uint32_t nof_symbols = 12*harq_process.prb_alloc.slot[0].nof_prb*RE_X_RB;
     uint32_t nof_bits_e = nof_symbols * lte_mod_bits_x_symbol(harq_process.mcs.mod);
@@ -210,8 +211,7 @@ int main(int argc, char **argv) {
     bzero(pusch.pusch_q, nof_bits_e*sizeof(uint8_t));
     
 
-    if (ulsch_uci_encode(&pusch.dl_sch, data, uci_data, pusch.pusch_g, 
-      pusch.pusch_g_ack, pusch.pusch_g_ri, &harq_process, rv, pusch.pusch_q)) 
+    if (ulsch_uci_encode(&pusch.dl_sch, data, uci_data, pusch.pusch_g, &harq_process, rv, pusch.pusch_q)) 
     {
       fprintf(stderr, "Error encoding TB\n");
       exit(-1);
diff --git a/lte/phy/lib/phch/test/ulsch_encode_test_mex.c b/lte/phy/lib/phch/test/ulsch_encode_test_mex.c
index 472472ead..8507f672d 100644
--- a/lte/phy/lib/phch/test/ulsch_encode_test_mex.c
+++ b/lte/phy/lib/phch/test/ulsch_encode_test_mex.c
@@ -153,8 +153,6 @@ void mexFunction(int nlhs, mxArray *plhs[], int nrhs, const mxArray *prhs[])
     
   uint32_t nof_symbols = 12*harq_process.prb_alloc.slot[0].nof_prb*RE_X_RB;
   uint32_t nof_q_bits = nof_symbols * lte_mod_bits_x_symbol(harq_process.mcs.mod);
-  
-  printf("alloc for %d bits\n", nof_q_bits);
 
   uint8_t *q_bits = vec_malloc(nof_q_bits * sizeof(uint8_t));
   if (!q_bits) {
@@ -164,17 +162,8 @@ void mexFunction(int nlhs, mxArray *plhs[], int nrhs, const mxArray *prhs[])
   if (!g_bits) {
     return;
   }
-  uint8_t *g_bits_ack = vec_malloc(nof_q_bits * sizeof(uint8_t));
-  if (!g_bits_ack) {
-    return;
-  }
-  uint8_t *g_bits_ri = vec_malloc(nof_q_bits * sizeof(uint8_t));
-  if (!g_bits_ri) {
-    return;
-  }
 
-  if (ulsch_uci_encode(&ulsch, trblkin, uci_data, g_bits,  
-                       g_bits_ack, g_bits_ri, &harq_process, rv, q_bits)) 
+  if (ulsch_uci_encode(&ulsch, trblkin, uci_data, g_bits, &harq_process, rv, q_bits)) 
   {
     mexErrMsgTxt("Error encoding TB\n");
     return;
@@ -188,8 +177,6 @@ void mexFunction(int nlhs, mxArray *plhs[], int nrhs, const mxArray *prhs[])
   
   free(trblkin);
   free(g_bits);  
-  free(g_bits_ack);
-  free(g_bits_ri);
   
   if (uci_data.uci_cqi_len > 0) {
     free(uci_data.uci_cqi);
diff --git a/matlab/tests/ulsch_test.m b/matlab/tests/ulsch_test.m
index 527bd68f7..e1aa28c83 100644
--- a/matlab/tests/ulsch_test.m
+++ b/matlab/tests/ulsch_test.m
@@ -1,23 +1,23 @@
 ueConfig=struct('NCellID',1,'CyclicPrefixUL','Normal','NTxAnts',1);
-puschConfig=struct('NLayers',1,'OrthCover','Off','PRBSet',0,'Modulation','QPSK','RV',0,'Shortened',0);
+puschConfig=struct('NLayers',1,'OrthCover','Off','PRBSet',(0:1)','Modulation','16QAM','RV',0,'Shortened',0);
 
 addpath('../../debug/lte/phy/lib/phch/test')
 
-TBs=100;
+TBs=200;
 error=zeros(size(TBs));
 for i=1:length(error)
-    %trblkin=randi(2,TBs(i),1)-1;
-    trblkin=ones(TBs(i),1);
+    trblkin=randi(2,TBs(i),1)-1;
+    %trblkin=ones(TBs(i),1);
     %trblkin=[];
     
     puschConfig.BetaCQI = 2.0; 
     puschConfig.BetaRI = 2.0;
     puschConfig.BetaACK = 2.0;
 
-    [mat, info]=lteULSCH(ueConfig,puschConfig,trblkin,[],[1],[],[]);
+    [mat, info]=lteULSCH(ueConfig,puschConfig,trblkin,[ones(1,20)],[1],[1],[]);
     mat(mat==-2)=2;
     mat(mat==-1)=3;
-    [lib]=liblte_ulsch_encode(ueConfig,puschConfig,trblkin,[],[1],[]);
+    [lib]=liblte_ulsch_encode(ueConfig,puschConfig,trblkin,[ones(1,20)],[1],[1]);
     error(i)=sum(abs(double(mat)-double(lib)));
     if (length(TBs) == 1)
         disp(error(i))    
@@ -27,7 +27,8 @@ end
 if (length(TBs) == 1)
     %disp(info)
     n=1:length(mat);
-    plot(abs(double(mat)-double(lib)))
+    %plot(abs(double(mat)-double(lib)))
+    plot(n,lib(n),n,mat(n))
 else
     plot(error)
     disp(sum(error))