E2E_framework_WGAN_AWGN.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "a060dfa1-4be6-4214-a85d-ca8ae31e6631",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<torch._C.Generator at 0x7f0574713c70>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import torch\n",
    "from torch.nn import functional as F\n",
    "import torch.optim as optim\n",
    "from tqdm import tqdm\n",
    "\n",
    "import pickle\n",
    "import csv\n",
    "import time\n",
    "import sys\n",
    "\n",
    "sys.path.append('src/')\n",
    "\n",
    "# Own libraries\n",
    "import utils\n",
    "import trainer_WGAN as trainer\n",
    "import model_WGAN\n",
    "import WGAN\n",
    "import models\n",
    "\n",
    "torch.manual_seed(42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "c97ad605-5e1c-46f8-8f97-2f363762dcc7",
   "metadata": {},
   "outputs": [],
   "source": [
    "use_cuda = torch.cuda.is_available()\n",
    "device = torch.device(\"cuda\" if use_cuda else \"cpu\")\n",
    "\n",
    "# Encoding parameters\n",
    "k = 4\n",
    "n = 7\n",
    "M = 16\n",
    "TRAINING_SNR = 5\n",
    "rate = k/n\n",
    "noise_std = utils.EbNo_to_noise(TRAINING_SNR, rate)\n",
    "save_every = 1000\n",
    "\n",
    "#GAN parameters\n",
    "g_input_size = 2*n      # Random noise dimension coming into generator, per output vector\n",
    "g_hidden_size = 128 # Generator complexity\n",
    "g_output_size = n     # Size of generated output vector\n",
    "\n",
    "d_input_size = 2*n    # Minibatch size - cardinality of distributions\n",
    "d_hidden_size = 128    # Discriminator complexity\n",
    "d_output_size = 1     # Single dimension for 'real' vs. 'fake' classification\n",
    "\n",
    "batch_size = 5000\n",
    "num_epochs = 20000\n",
    "\n",
    "G = WGAN.Generator(input_size=g_input_size, hidden_size=g_hidden_size, output_size=g_output_size)\n",
    "D = WGAN.Discriminator(input_size=d_input_size, hidden_size=d_hidden_size, output_size=d_output_size)\n",
    "G = G.to(device)\n",
    "D = D.to(device)\n",
    "\n",
    "d_lr = 1e-4\n",
    "g_lr = 1e-4\n",
    "\n",
    "channel_awgn = model_WGAN.Channel_AWGN(noise_std)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "761135f8-b345-45c2-af6e-773d3d5d0d97",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<All keys matched successfully>"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "G.load_state_dict(torch.load('state_dict_WGAN/G_AWGN-40.pt', map_location=\"cpu\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "47d835bd",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "# of parameters (WGAN model Generator): 19335\n",
      "# of parameters (WGAN model Critic): 18561\n"
     ]
    }
   ],
   "source": [
    "total_params_g = sum(p.numel() for p in G.parameters())\n",
    "print('# of parameters (WGAN model Generator): ' + str(total_params_g))\n",
    "total_params_d = sum(p.numel() for p in D.parameters())\n",
    "print('# of parameters (WGAN model Critic): ' + str(total_params_d))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ba5f8a74-0c00-4804-89c3-e4dec24cd93e",
   "metadata": {},
   "source": [
    "# Training WGAN\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "5b31dfeb-169f-43c3-9d61-bd88442ad557",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Epoch 20000: KL Div 1.738e-03 G loss 1.000e-02: 100%|#| 20000/20000 [07:21<00:00\n"
     ]
    }
   ],
   "source": [
    "num_epochs = 20000\n",
    "epochs = tqdm(range(num_epochs), ascii=True)\n",
    "for epoch in epochs:\n",
    "    kl, g_loss = WGAN.train_WGAN(G, D, batch_size, n, channel_awgn, d_lr, g_lr)\n",
    "    if (epoch+1) % 50 == 0:\n",
    "        epochs.set_description(f\"Epoch {epoch+1}: KL Div {kl:.3e} G loss {-g_loss:.3e}\")\n",
    "    if (epoch+1) % save_every == 0:\n",
    "        torch.save(G.state_dict(), f'state_dict_WGAN/G_AWGN-{(epoch+1)//save_every}.pt')\n",
    "        torch.save(D.state_dict(), f'state_dict_WGAN/D_AWGN-{(epoch+1)//save_every}.pt')\n",
    "\n",
    "    #    WGAN.plot_GAN(G, 10000,M, n, channel_awgn, epoch)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "c1b98c07-460a-4373-8701-db2fa4c868e9",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Epoch 10000: KL Div 1.906e-03 G loss 1.000e-02: 100%|#| 10000/10000 [04:19<00:00\n"
     ]
    }
   ],
   "source": [
    "num_epochs_prev = num_epochs\n",
    "num_epochs = 10000\n",
    "d_lr = 1e-5; g_lr = 1e-5\n",
    "epochs = tqdm(range(num_epochs), ascii=True)\n",
    "for epoch in epochs:\n",
    "    kl, g_loss = WGAN.train_WGAN(G, D, batch_size, n, channel_awgn, d_lr, g_lr)\n",
    "    if (epoch+1) % 50 == 0:\n",
    "        epochs.set_description(f\"Epoch {epoch+1}: KL Div {kl:.3e} G loss {-g_loss:.3e}\")\n",
    "    if (epoch+1) % save_every == 0:\n",
    "        torch.save(G.state_dict(), f'state_dict_WGAN/G_AWGN-{(num_epochs_prev+epoch+1)//save_every}.pt')\n",
    "        torch.save(D.state_dict(), f'state_dict_WGAN/D_AWGN-{(num_epochs_prev+epoch+1)//save_every}.pt')\n",
    "\n",
    "    #    WGAN.plot_GAN(G, 10000,M, n, channel_awgn, epoch)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "6b96b9c9-5cf7-4beb-b98c-6b936bc1d884",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1:  \n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x216 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "WGAN.plot_GAN(G, 10000, M, n, channel_awgn, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "f89ff8b1-c765-4929-b1be-d02c5a424e1c",
   "metadata": {},
   "outputs": [],
   "source": [
    "## Measuring the sampling time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "05f51b61-b874-4d54-b182-8409edc9e179",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.4778146743774414\n"
     ]
    }
   ],
   "source": [
    "# Set the message samples to test out\n",
    "num_samples = 2 ** 20 # per message\n",
    "#m_sample = torch.eye(M)\n",
    "#m = m_sample.repeat_interleave(num_samples, dim=0)\n",
    "batch_size = 1024 \n",
    "\n",
    "#encoder = models.Encoder_1h( M, n, n).to(device)\n",
    "x = torch.ones(batch_size, n).to(device) \n",
    "t = time.time() # tic\n",
    "\n",
    "#m_batched = torch.split(m, batch_size)\n",
    "#for m_b in m_batched:\n",
    "for i in range(num_samples // batch_size):\n",
    "    #m_b = m_b.to(device)\n",
    "    #x = encoder(m_b)\n",
    "    x_seq = G(x)\n",
    "    #samples_DDPM = x_seq[-1]\n",
    "sampling_time_WGAN = time.time() - t # toc\n",
    "print(sampling_time_WGAN)\n",
    "np.save('results/t_WGAN_AWGN.npy', sampling_time_WGAN)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "43096bd1-8797-44d0-aa02-bdebfb25fe63",
   "metadata": {},
   "source": [
    "# Training GAN Encoder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "23194890-8bc3-4913-bb24-de6235227d5e",
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset = np.random.randint(0, M, size=10000000) # size = 10000000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "d58c10c6-1688-48e0-9ef7-393d403f23ff",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "epoch 1: loss 4.51e-01 SER 1.49e-01: 100%|#| 10000/10000 [00:32<00:00, 309.42it/\n",
      "epoch 2: loss 1.94e-01 SER 6.01e-02: 100%|#| 10000/10000 [00:47<00:00, 212.07it/\n",
      "epoch 3: loss 1.71e-01 SER 5.35e-02: 100%|#| 10000/10000 [00:57<00:00, 172.87it/\n",
      "epoch 4: loss 1.60e-01 SER 4.98e-02: 100%|#| 10000/10000 [00:56<00:00, 177.15it/\n",
      "epoch 5: loss 1.53e-01 SER 4.78e-02: 100%|#| 10000/10000 [00:53<00:00, 187.09it/\n",
      "epoch 6: loss 1.48e-01 SER 4.63e-02: 100%|#| 10000/10000 [00:53<00:00, 188.20it/\n",
      "epoch 7: loss 1.44e-01 SER 4.51e-02: 100%|#| 10000/10000 [00:54<00:00, 184.06it/\n",
      "epoch 8: loss 1.41e-01 SER 4.43e-02: 100%|#| 10000/10000 [00:58<00:00, 172.40it/\n",
      "epoch 9: loss 1.39e-01 SER 4.36e-02: 100%|#| 10000/10000 [00:58<00:00, 172.09it/\n",
      "epoch 10: loss 1.38e-01 SER 4.32e-02: 100%|#| 10000/10000 [00:58<00:00, 170.66it\n",
      "epoch 11: loss 1.36e-01 SER 4.27e-02: 100%|#| 10000/10000 [00:59<00:00, 168.39it\n",
      "epoch 12: loss 1.35e-01 SER 4.24e-02: 100%|#| 10000/10000 [00:59<00:00, 168.48it\n",
      "epoch 13: loss 1.34e-01 SER 4.19e-02: 100%|#| 10000/10000 [00:58<00:00, 170.41it\n",
      "epoch 14: loss 1.32e-01 SER 4.15e-02: 100%|#| 10000/10000 [00:56<00:00, 177.17it\n",
      "epoch 15: loss 1.32e-01 SER 4.14e-02: 100%|#| 10000/10000 [00:46<00:00, 215.99it\n",
      "epoch 16: loss 1.31e-01 SER 4.10e-02: 100%|#| 10000/10000 [00:30<00:00, 333.05it\n",
      "epoch 17: loss 1.30e-01 SER 4.07e-02: 100%|#| 10000/10000 [00:29<00:00, 333.94it\n",
      "epoch 18: loss 1.29e-01 SER 4.06e-02: 100%|#| 10000/10000 [00:29<00:00, 336.83it\n",
      "epoch 19: loss 1.28e-01 SER 4.03e-02: 100%|#| 10000/10000 [00:30<00:00, 324.67it\n",
      "epoch 20: loss 1.27e-01 SER 3.99e-02: 100%|#| 10000/10000 [00:29<00:00, 337.50it\n"
     ]
    }
   ],
   "source": [
    "encoder_GAN = models.Encoder_1h( M, n, n)\n",
    "decoder_GAN = models.Decoder_1h( M, n)\n",
    "channel = model_WGAN.Channel_GAN(encoder_GAN, decoder_GAN, G).to(device)\n",
    "tconf = trainer.TrainerConfig(max_epochs=20, noise_std = noise_std, batch_size=1000, M=M, rate=rate) # epoch = 10, batch_size = 1000\n",
    "weights = filter(lambda p: p.requires_grad, list(encoder_GAN.parameters())+list(decoder_GAN.parameters()))\n",
    "t_gan = trainer.Trainer(channel, dataset, tconf, one_hot=True)\n",
    "t_gan.train(weights=weights, GAN=True, global_epoch=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "12e35c96-035c-4c24-b3ac-8141e7aeaa06",
   "metadata": {},
   "outputs": [],
   "source": [
    "torch.save(encoder_GAN.state_dict(), 'state_dict_WGAN/enc_GAN_AWGN.pt')\n",
    "torch.save(decoder_GAN.state_dict(), 'state_dict_WGAN/dec_GAN_AWGN.pt')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4dc83fc0-eb93-4eef-a2bf-a80e9f726fed",
   "metadata": {},
   "source": [
    "# Training AE Encoder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "f7db9a9d-916e-4934-980b-70420f64ce7a",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "epoch 1: loss 6.14e-02 SER 1.65e-02: 100%|#| 10000/10000 [00:30<00:00, 329.01it/\n",
      "epoch 2: loss 1.36e-02 SER 4.57e-03: 100%|#| 10000/10000 [00:36<00:00, 272.82it/\n",
      "epoch 3: loss 1.14e-02 SER 3.83e-03: 100%|#| 10000/10000 [00:36<00:00, 271.41it/\n",
      "epoch 4: loss 1.10e-02 SER 3.71e-03: 100%|#| 10000/10000 [00:37<00:00, 269.70it/\n",
      "epoch 5: loss 1.07e-02 SER 3.57e-03: 100%|#| 10000/10000 [00:38<00:00, 260.27it/\n",
      "epoch 6: loss 1.06e-02 SER 3.55e-03: 100%|#| 10000/10000 [00:40<00:00, 248.66it/\n",
      "epoch 7: loss 1.06e-02 SER 3.54e-03: 100%|#| 10000/10000 [00:36<00:00, 277.32it/\n",
      "epoch 8: loss 1.05e-02 SER 3.53e-03: 100%|#| 10000/10000 [00:38<00:00, 258.88it/\n",
      "epoch 9: loss 1.05e-02 SER 3.51e-03: 100%|#| 10000/10000 [00:40<00:00, 247.93it/\n",
      "epoch 10: loss 1.05e-02 SER 3.52e-03: 100%|#| 10000/10000 [00:40<00:00, 248.69it\n"
     ]
    }
   ],
   "source": [
    "encoder_AE = models.Encoder_1h(M, n, n)# models.Encoder_1h( M, n)\n",
    "decoder_AE = models.Decoder_1h(M, n)\n",
    "model_2 = model_WGAN.Channel_GAN(encoder_AE, decoder_AE, channel_awgn).to(device)\n",
    "#model_2 = model_WGAN.Channel(encoder_AE, decoder_AE).to(device)\n",
    "\n",
    "tconf = trainer.TrainerConfig(max_epochs=10, noise_std = noise_std, batch_size=1000, M=M, rate=rate)\n",
    "weights = filter(lambda p: p.requires_grad, model_2.parameters())\n",
    "\n",
    "t_ae = trainer.Trainer(model_2, dataset, tconf, one_hot=True)\n",
    "t_ae.train(weights=weights, GAN=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b447c298-6d74-42f5-9ce0-17e6fbf33088",
   "metadata": {},
   "source": [
    "# Testing AE & GAN Encoder under real channel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "18846fc9-872d-4803-857b-e0bd77ab0014",
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset = np.random.randint(0,M, size=10000000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8d14d414-cd0a-45b9-89dc-98685ef74105",
   "metadata": {},
   "outputs": [],
   "source": [
    "bler_testGAN =[]\n",
    "ebno_range = [2, 3, 4, 5, 6, 7, 8]\n",
    "\n",
    "model_testGAN = model_WGAN.Channel(encoder_GAN, decoder_GAN).to(device)\n",
    "tconf = trainer.TrainerConfig(max_epochs=5, noise_std = noise_std, batch_size=1000, M=M, rate=rate)\n",
    "\n",
    "t_testGAN = trainer.Trainer(model_testAE, dataset, tconf, one_hot=True)\n",
    "\n",
    "\n",
    "bler_testGAN = t_testAE.test(snr_range=ebno_range, one_hot=True, erasure_bound=0.7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "c9e54194-cf6f-4953-879f-eda61fc58639",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 2  iter 9900: SER 6.419e-02: 100%|###| 10000/10000 [00:19<00:00, 518.79it/s]\n",
      "SNR 3  iter 9900: SER 3.031e-02: 100%|###| 10000/10000 [00:18<00:00, 529.24it/s]\n",
      "SNR 4  iter 9900: SER 1.164e-02: 100%|###| 10000/10000 [00:19<00:00, 514.06it/s]\n",
      "SNR 5  iter 9900: SER 3.487e-03: 100%|###| 10000/10000 [00:18<00:00, 527.25it/s]\n",
      "SNR 6  iter 9900: SER 7.456e-04: 100%|###| 10000/10000 [00:19<00:00, 518.20it/s]\n",
      "SNR 7  iter 9900: SER 1.107e-04: 100%|###| 10000/10000 [00:19<00:00, 525.61it/s]\n",
      "SNR 7  iter 9900: SER 1.055e-04: 100%|###| 10000/10000 [00:19<00:00, 512.60it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  20000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 7  iter 9900: SER 1.085e-04: 100%|###| 10000/10000 [00:18<00:00, 531.16it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  30000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 7  iter 9900: SER 1.124e-04: 100%|###| 10000/10000 [00:20<00:00, 488.70it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  40000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 7  iter 9900: SER 1.051e-04: 100%|###| 10000/10000 [00:23<00:00, 434.02it/s]\n",
      "SNR 8  iter 9900: SER 1.071e-05: 100%|###| 10000/10000 [00:23<00:00, 423.60it/s]\n",
      "SNR 8  iter 9900: SER 9.595e-06: 100%|###| 10000/10000 [00:18<00:00, 533.04it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  20000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.121e-05: 100%|###| 10000/10000 [00:18<00:00, 542.52it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  30000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.050e-05: 100%|###| 10000/10000 [00:16<00:00, 593.58it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  40000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.141e-05: 100%|###| 10000/10000 [00:17<00:00, 581.04it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  50000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 9.999e-06: 100%|###| 10000/10000 [00:18<00:00, 546.12it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  60000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.101e-05: 100%|###| 10000/10000 [00:18<00:00, 536.56it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  70000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.101e-05: 100%|###| 10000/10000 [00:18<00:00, 538.30it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  80000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.192e-05: 100%|###| 10000/10000 [00:16<00:00, 594.93it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  90000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.081e-05: 100%|###| 10000/10000 [00:16<00:00, 601.16it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  100000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 7.979e-06: 100%|###| 10000/10000 [00:17<00:00, 578.70it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  110000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.161e-05: 100%|###| 10000/10000 [00:20<00:00, 491.23it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  120000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 9.999e-06: 100%|###| 10000/10000 [00:21<00:00, 474.71it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  130000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.060e-05: 100%|###| 10000/10000 [00:20<00:00, 493.18it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  140000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.060e-05: 100%|###| 10000/10000 [00:20<00:00, 492.29it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  150000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 9.595e-06: 100%|###| 10000/10000 [00:20<00:00, 490.58it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  160000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 8.080e-06: 100%|###| 10000/10000 [00:20<00:00, 489.29it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  170000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.202e-05: 100%|###| 10000/10000 [00:22<00:00, 443.19it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  180000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.091e-05: 100%|###| 10000/10000 [00:23<00:00, 425.01it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  190000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 9.191e-06: 100%|###| 10000/10000 [00:22<00:00, 436.44it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  200000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 9.393e-06: 100%|###| 10000/10000 [00:23<00:00, 430.05it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  210000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.060e-05: 100%|###| 10000/10000 [00:23<00:00, 434.05it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  220000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.030e-05: 100%|###| 10000/10000 [00:22<00:00, 436.90it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  230000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.040e-05: 100%|###| 10000/10000 [00:22<00:00, 441.24it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  240000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.141e-05: 100%|###| 10000/10000 [00:26<00:00, 382.73it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  250000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 9.292e-06: 100%|###| 10000/10000 [00:25<00:00, 388.21it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  260000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 9.494e-06: 100%|###| 10000/10000 [00:24<00:00, 408.12it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  270000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 8.282e-06: 100%|###| 10000/10000 [00:22<00:00, 443.25it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  280000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.081e-05: 100%|###| 10000/10000 [00:23<00:00, 432.69it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  290000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.020e-05: 100%|###| 10000/10000 [00:25<00:00, 387.27it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  300000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 9.696e-06: 100%|###| 10000/10000 [00:25<00:00, 398.79it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  310000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 9.393e-06: 100%|###| 10000/10000 [00:24<00:00, 402.93it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  320000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.111e-05: 100%|###| 10000/10000 [00:24<00:00, 401.16it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  330000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 9.494e-06: 100%|###| 10000/10000 [00:24<00:00, 400.13it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  340000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.020e-05: 100%|###| 10000/10000 [00:24<00:00, 400.66it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  350000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 9.696e-06: 100%|###| 10000/10000 [00:25<00:00, 394.97it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  360000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.091e-05: 100%|###| 10000/10000 [00:24<00:00, 401.18it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  370000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.030e-05: 100%|###| 10000/10000 [00:25<00:00, 397.63it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  380000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 9.292e-06: 100%|###| 10000/10000 [00:24<00:00, 401.65it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  390000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.071e-05: 100%|###| 10000/10000 [00:25<00:00, 398.98it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  400000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 8.282e-06: 100%|###| 10000/10000 [00:24<00:00, 405.01it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  410000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.121e-05: 100%|###| 10000/10000 [00:25<00:00, 396.43it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  420000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.202e-05: 100%|###| 10000/10000 [00:25<00:00, 399.62it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  430000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.030e-05: 100%|###| 10000/10000 [00:25<00:00, 399.82it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  440000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.182e-05: 100%|###| 10000/10000 [00:24<00:00, 404.74it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  450000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 9.393e-06: 100%|###| 10000/10000 [00:25<00:00, 399.75it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  460000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 8.787e-06: 100%|###| 10000/10000 [00:24<00:00, 404.38it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  470000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.182e-05: 100%|###| 10000/10000 [00:25<00:00, 399.88it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of samples is not enough. One more epoch is running. Total # of samples used:  480000000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "SNR 8  iter 9900: SER 1.010e-05: 100%|###| 10000/10000 [00:24<00:00, 403.99it/s]\n"
     ]
    }
   ],
   "source": [
    "bler_testAE =[]\n",
    "ebno_range = [2, 3, 4, 5, 6, 7, 8]\n",
    "\n",
    "model_testAE = model_WGAN.Channel(encoder_AE, decoder_AE).to(device)\n",
    "tconf = trainer.TrainerConfig(max_epochs=5, noise_std = noise_std, batch_size=1000, M=M, rate=rate)\n",
    "\n",
    "t_testAE = trainer.Trainer(model_testAE, dataset, tconf, one_hot=True)\n",
    "\n",
    "\n",
    "bler_testAE = t_testAE.test(snr_range=ebno_range, one_hot=True, erasure_bound=0.7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "92420f2f-3f1d-45f8-bdb3-39b057b5ba8c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1600x1000 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "fig = plt.figure(figsize=(16, 10))\n",
    "# Plot Autoencoder\n",
    "plt.semilogy(ebno_range, bler_testAE, 'o-')\n",
    "#plt.semilogy(bber_data[0], bber_data[2], 'o-')\n",
    "#plt.semilogy(bber_data[0], vec, 'o-')\n",
    "\n",
    "plt.semilogy(ebno_range, bler_testGAN, '^-')\n",
    "\n",
    "# Plot approx 16 QAM\n",
    "#ebnodbs = np.linspace(5,14,10)\n",
    "#a, _ = utils.MQAM_rayleigh_approx(16,ebnodbs)\n",
    "#plt.plot(ebnodbs, a, '^-');\n",
    "\n",
    "#b = [utils.rayleigh_error(t) for t in ebnodbs]\n",
    "#plt.plot(ebnodbs, b, '^-');\n",
    "\n",
    "#ebnodbs = np.linspace(0,11,22)\n",
    "#uncoded_16QAM_Rayleigh_SER = [0.5501, 0.5040, 0.4564, 0.4082, 0.3606, 0.3147, 0.2713, 0.2313, 0.1951, 0.1631, 0.1352, 0.1112, 0.091, 0.07396, 0.05989, 0.04832, 0.03886, 0.03119, 0.02497, 0.01996, 0.01594, 0.01271]\n",
    "#c = [(1-(1-t)**4) for t in uncoded_16QAM_Rayleigh_SER]\n",
    "#plt.plot(ebnodbs, c, '^-');\n",
    "\n",
    "plt.gca().set_ylim(1e-5, 1)\n",
    "plt.gca().set_xlim(1, 9)\n",
    "plt.ylabel(\"Block Error Rate\", fontsize=18, rotation=90)\n",
    "plt.xlabel(\"Eb/No [dB]\", fontsize=18)\n",
    "plt.legend(['AE', 'WGAN'], prop={'size': 16}, loc='upper right');\n",
    "plt.grid(True, which=\"both\")\n",
    "#plt.savefig('ae-ray-testing.pdf')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fa74c701-f06c-4c3c-a93a-13fb062a7aea",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}