WMMSE_eval.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "0c128502-d09a-4a98-9f63-d469d3368b64",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\VQ72TN\\Anaconda3\\envs\\pyg_env3\\lib\\site-packages\\tqdm\\auto.py:22: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
      "  from .autonotebook import tqdm as notebook_tqdm\n"
     ]
    }
   ],
   "source": [
    "import scipy.io as sio    # import scipy.io for .mat file I/O\n",
    "import scipy\n",
    "from scipy import stats  \n",
    "import numpy as np                         # import numpy\n",
    "import matplotlib.pyplot as plt            # import matplotlib.pyplot for figure plotting\n",
    "import torch\n",
    "import function_wmmse_powercontrol as wmmse_pc\n",
    "import subnetwork_generator\n",
    "import time\n",
    "import math"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "5e4163ce-f991-47eb-98ec-ebf711aa3f12",
   "metadata": {
    "scrolled": true,
    "tags": []
   },
   "outputs": [],
   "source": [
    "class init_parameters:\n",
    "    def __init__(self,S_sd,rng, num_of_subn):\n",
    "        self.num_of_subnetworks = num_of_subn\n",
    "        self.deploy_length = 20    # the length and breadth of the factory area (m)  \n",
    "        self.subnet_radius = 2      # the radius of the subnetwork cell (m)\n",
    "        self.minD = 0.5               #minimum distance from device to controller(access point) (m)\n",
    "        self.minDistance = self.subnet_radius    #minimum controller to controller distance (m)\n",
    "        self.sigmaS = S_sd        #shadowing standard deviation\n",
    "        self.transmit_power = 1      #normalized transmit power mW\n",
    "        self.rng_value = np.random.RandomState(rng)\n",
    "        self.bandwidth = 5e6         #bandwidth (Hz)\n",
    "        self.frequency = 6e9          #Operating frequency (Hz)\n",
    "        self.lambdA = 3e8/6e9        \n",
    "        self.plExponent = 2.7        #path loss exponent\n",
    "\n",
    "\n",
    "testsh_sd = 7\n",
    "test_num_subn = 20\n",
    "config = init_parameters(testsh_sd ,0,test_num_subn)\n",
    "testing_snapshots = 50000\n",
    "\n",
    "test_powers, dist = subnetwork_generator.generate_samples(config, testing_snapshots)\n",
    "bandwidth = config.bandwidth\n",
    "Noise_power = np.power(10,((-174+10+10*np.log10(bandwidth))/10))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "e63f828f-dd7b-4e02-a78a-8021e9f9437f",
   "metadata": {},
   "outputs": [],
   "source": [
    "def mycapacity(weights, data, num_subnetworks):\n",
    "\n",
    "    weights = weights.reshape([-1,num_subnetworks,1,1])\n",
    "    \n",
    "    power_mat = data.reshape([-1,num_subnetworks,num_subnetworks,1])\n",
    "\n",
    "    weighted_powers = torch.mul(weights,power_mat) #received powers\n",
    "    \n",
    "    eye = torch.eye(num_subnetworks)\n",
    "    \n",
    "    desired_rcv_power = torch.sum(torch.mul(weighted_powers.squeeze(-1),eye), dim=1)\n",
    "   \n",
    "    Interference_power = torch.sum(torch.mul(weighted_powers.squeeze(-1),1-eye), dim=1)\n",
    "\n",
    "    signal_interference_ratio = torch.divide(desired_rcv_power,Interference_power+Noise_power)\n",
    "    \n",
    "    capacity = torch.log2(1+signal_interference_ratio)\n",
    "    \n",
    "    return capacity, weighted_powers   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "82587ae2-b987-49f6-9afe-27148412cb65",
   "metadata": {},
   "outputs": [],
   "source": [
    "def generate_cdf(values, bins_):\n",
    "    data = np.array(values)\n",
    "    count, bins_count = np.histogram(data, bins=bins_)\n",
    "    pdf = count / sum(count)\n",
    "    cdf = np.cumsum(pdf)\n",
    "    return bins_count[1:], cdf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "369d4197-84a1-41ab-aa4b-3088d52f247b",
   "metadata": {},
   "outputs": [],
   "source": [
    "def WMMSE_powers_sum_rate(K=config.num_of_subnetworks, num_H=testing_snapshots, Pmax=1, Pmin=0):\n",
    "    Pini = Pmax*np.ones(K)\n",
    "    var_noise = Noise_power\n",
    "    X=np.zeros((K**2,num_H))\n",
    "    Y=np.zeros((K,num_H))\n",
    "    iter_ = np.zeros(num_H)\n",
    "    total_time = 0.0\n",
    "    for loop in range(num_H):\n",
    "        H=test_powers[loop,:,:]\n",
    "        X[:,loop] = np.reshape(H, (K**2,), order=\"F\")\n",
    "        H=np.reshape(X[:,loop], (K,K), order=\"F\")\n",
    "        mid_time = time.time()\n",
    "        Y[:,loop], iter_[loop] = wmmse_pc.WMMSE_sum_rate(Pini, H, Pmax, var_noise)\n",
    "        total_time = total_time + time.time() - mid_time\n",
    "    # print(\"wmmse time: %0.2f s\" % total_time)\n",
    "    return X, Y, total_time, iter_\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "560411e9-bb70-4853-b004-c0357b1f9158",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_sr,Y_sr,t_sr, iter_ = WMMSE_powers_sum_rate()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "e021c0f4-e1f6-48e3-a7ba-fc6e65c7bccc",
   "metadata": {},
   "outputs": [],
   "source": [
    "WMMSE_capacities_sr, powers_sr = mycapacity(torch.tensor(np.transpose(Y_sr)), torch.tensor(test_powers), config.num_of_subnetworks)\n",
    "weights_ones = torch.ones_like(torch.tensor(np.transpose(Y_sr)))\n",
    "capacities_ones, Uniform_pow = mycapacity(weights_ones, torch.tensor(test_powers),config.num_of_subnetworks)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "5ec45a30-6ad5-4bbf-a6d5-e7400372c8f4",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#WMMSE Sum rate powers\n",
    "x,y = generate_cdf(10*np.log10(np.sum(Y_sr,0)/20), 1000)\n",
    "plt.plot(x,y)\n",
    "plt.ylabel('cdf')\n",
    "plt.xlabel('Average deployment transmit power (dBm)')\n",
    "plt.grid()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9bed898a-935e-4580-a933-08ac172c74a3",
   "metadata": {},
   "outputs": [],
   "source": [
    "## WMMSE Sum rate sum capacities\n",
    "x,y = generate_cdf(torch.sum(WMMSE_capacities_sr,1)/20, 1000)\n",
    "plt.plot(x,y,label=\"WMMSE\")\n",
    "x,y = generate_cdf(torch.sum(capacities_ones,1)/20, 1000)\n",
    "plt.plot(x,y,label=\"Uniform\")\n",
    "plt.legend()\n",
    "plt.ylabel('CDF')\n",
    "plt.xlabel('Spectral efficiency (bits/s/Hz)')\n",
    "tikzplotlib.save(\"tikz files/csWMMSE_sr_sumCap.tex\")\n",
    "plt.grid()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "94f3e4c5-9f26-48cb-a770-de4ea7df8868",
   "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": "3.10.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}