benchmark/code_benchmark.ipynb

{
 "metadata": {
  "name": "code benchmark"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Benchmark of A simple ECC certificate library\n",
      "=============================================\n",
      "\n",
      "_Author: Tony Cheneau_\n",
      "\n",
      "This page provides some benchmark of the simple ECC certificate library. The data has been obtained using a Redbee Econotag."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import pandas as pd\n",
      "import numpy as np\n",
      "import matplotlib.pyplot as plt\n",
      "from mpltools import style\n",
      "from mpltools import layout\n",
      "\n",
      "style.use('ggplot')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 1
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "These three files are a dump of the serial output of the Econotag. The dump was realized using [serialreadts](https://github.com/tcheneau/serialreadts), a tool that prepend a (local) timestamp in front of every lines of a serial output. The Econotag have been loaded with the images of the udp-client.c code with varying curve parameter (hence the file name). The image performs crypto operation in a loop (see the *test_crypto()* function)."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "log_192k1 = open(\"speed-secp192k1.txt\").read()\n",
      "log_160k1 = open(\"speed-secp160k1.txt\").read()\n",
      "log_128r1 = open(\"speed-secp128r1.txt\").read()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 2
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def parse_lines(data):\n",
      "    lines = data.splitlines()\n",
      "    lines = [line.split(\"]\") for line in lines]\n",
      "    return [(float(a[1:]), b) for [a,b] in lines]    \n",
      "\n",
      "def extract_timing_text(lines):\n",
      "    timing_text = [text for (time, text) in lines if \"done\" in text]\n",
      "    timing_text += [ text[7:] for text in timing_text]\n",
      "    return timing_text\n",
      "\n",
      "\n",
      "def parse_log(data, curve_name):\n",
      "    lines = parse_lines(data)\n",
      "    timing_text = extract_timing_text(lines)\n",
      "    start = []\n",
      "    stop = []\n",
      "    i = 0\n",
      "    for (time, text) in lines:\n",
      "        if text not in timing_text:\n",
      "            continue\n",
      "        if i%2 == 0 and 'done' in text:\n",
      "            continue\n",
      "\n",
      "        if i%2 == 0:\n",
      "            start.append((time, text))\n",
      "        else:\n",
      "            stop.append((time,text))\n",
      "        i += 1\n",
      "        \n",
      "    timings = []\n",
      "    length = min(len(start), len(stop))\n",
      "    for i in range(length):\n",
      "        (start_time, text) = start[i]\n",
      "        (stop_time, _) = stop[i]\n",
      "        timings.append((curve_name, text, stop_time - start_time))\n",
      "        \n",
      "    return timings\n",
      "    "
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 3
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "timings = parse_log(log_128r1, \"secp128r1\")\n",
      "timings += parse_log(log_160k1, \"secp160k1\")\n",
      "timings += parse_log(log_192k1, \"secp192k1\")"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 4
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Parse the data and make sure it looks OK."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "df = pd.DataFrame(timings, columns = [\"curve\", \"event\", \"duration\"])\n",
      "df = df.sort(\"curve\")\n",
      "df.head()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
        "<table border=\"1\" class=\"dataframe\">\n",
        "  <thead>\n",
        "    <tr style=\"text-align: right;\">\n",
        "      <th></th>\n",
        "      <th>curve</th>\n",
        "      <th>event</th>\n",
        "      <th>duration</th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>0</th>\n",
        "      <td> secp128r1</td>\n",
        "      <td> verifying ECDSA signature (bad 1/2)</td>\n",
        "      <td> 1.376329</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>1</th>\n",
        "      <td> secp128r1</td>\n",
        "      <td> verifying ECDSA signature (bad 2/2)</td>\n",
        "      <td> 0.282866</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>2</th>\n",
        "      <td> secp128r1</td>\n",
        "      <td>                       serialization</td>\n",
        "      <td> 0.001749</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>3</th>\n",
        "      <td> secp128r1</td>\n",
        "      <td>                     deserialization</td>\n",
        "      <td> 0.002000</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>4</th>\n",
        "      <td> secp128r1</td>\n",
        "      <td>                  ECDH first message</td>\n",
        "      <td> 0.115998</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
        "</div>"
       ],
       "output_type": "pyout",
       "prompt_number": 5,
       "text": [
        "       curve                                event  duration\n",
        "0  secp128r1  verifying ECDSA signature (bad 1/2)  1.376329\n",
        "1  secp128r1  verifying ECDSA signature (bad 2/2)  0.282866\n",
        "2  secp128r1                        serialization  0.001749\n",
        "3  secp128r1                      deserialization  0.002000\n",
        "4  secp128r1                   ECDH first message  0.115998"
       ]
      }
     ],
     "prompt_number": 5
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# here we remove uninteresting \"event\"\n",
      "mask = (df['event']!=\"ECDH derive key\") & \\\n",
      "   (df['event']!=\"verifying ECDSA signature (bad 1/2)\") & \\\n",
      "   (df['event']!=\"verifying ECDSA signature (bad 2/2)\") & \\\n",
      "   (df['event']!=\"verify a client certificate\") & \\\n",
      "   (df['event']!=\"deserialization\") & \\\n",
      "   (df['event']!=\"serialization\")\n",
      "df = df[mask]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 6
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def rename_event(event_name):\n",
      "    if event_name == \"sign random data\":\n",
      "        return \"ECDSA sign 200 bytes\"\n",
      "    elif event_name == \"verify own's certificate\":\n",
      "        return \"ECDSA verify 200 bytes\"\n",
      "    elif event_name == \"generate a random certificate\":\n",
      "        return \"Generate certificate\"\n",
      "    elif event_name == \"verifying ECDSA signature (good)\":\n",
      "        return \"Verify a certificate\"\n",
      "    elif event_name == \"ECDH first message\":\n",
      "        return \"ECDH first event\"\n",
      "    elif event_name == \"ECDH second message\":\n",
      "        return \"ECDH second event\"\n",
      "    else:\n",
      "        return event_name\n",
      "    \n",
      "\n",
      "df['event'] = df['event'].apply(rename_event)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 7
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# count the number of events we have in our dataset\n",
      "# this is to make sure we have enough \"raw\" material\n",
      "df.groupby(['curve','event']).apply(lambda x: len(x))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "pyout",
       "prompt_number": 8,
       "text": [
        "curve      event                 \n",
        "secp128r1  ECDH first event          589\n",
        "           ECDH second event         589\n",
        "           ECDSA sign 200 bytes      589\n",
        "           ECDSA verify 200 bytes    589\n",
        "           Generate certificate      589\n",
        "           Verify a certificate      589\n",
        "secp160k1  ECDH first event          611\n",
        "           ECDH second event         611\n",
        "           ECDSA sign 200 bytes      612\n",
        "           ECDSA verify 200 bytes    611\n",
        "           Generate certificate      611\n",
        "           Verify a certificate      612\n",
        "secp192k1  ECDH first event          666\n",
        "           ECDH second event         666\n",
        "           ECDSA sign 200 bytes      667\n",
        "           ECDSA verify 200 bytes    666\n",
        "           Generate certificate      666\n",
        "           Verify a certificate      667\n",
        "dtype: int64"
       ]
      }
     ],
     "prompt_number": 8
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "The results below are in seconds."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "res = df.groupby([\"curve\", \"event\"]).agg(np.mean)\n",
      "res"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
        "<table border=\"1\" class=\"dataframe\">\n",
        "  <thead>\n",
        "    <tr style=\"text-align: right;\">\n",
        "      <th></th>\n",
        "      <th></th>\n",
        "      <th>duration</th>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>curve</th>\n",
        "      <th>event</th>\n",
        "      <th></th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
        "      <th rowspan=\"6\" valign=\"top\">secp128r1</th>\n",
        "      <th>ECDH first event</th>\n",
        "      <td> 0.111218</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>ECDH second event</th>\n",
        "      <td> 0.171697</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>ECDSA sign 200 bytes</th>\n",
        "      <td> 0.788145</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>ECDSA verify 200 bytes</th>\n",
        "      <td> 1.382880</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>Generate certificate</th>\n",
        "      <td> 0.124332</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>Verify a certificate</th>\n",
        "      <td> 1.390121</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th rowspan=\"6\" valign=\"top\">secp160k1</th>\n",
        "      <th>ECDH first event</th>\n",
        "      <td> 0.136917</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>ECDH second event</th>\n",
        "      <td> 0.188761</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>ECDSA sign 200 bytes</th>\n",
        "      <td> 0.842946</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>ECDSA verify 200 bytes</th>\n",
        "      <td> 1.409523</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>Generate certificate</th>\n",
        "      <td> 0.149816</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>Verify a certificate</th>\n",
        "      <td> 1.417879</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th rowspan=\"6\" valign=\"top\">secp192k1</th>\n",
        "      <th>ECDH first event</th>\n",
        "      <td> 0.173947</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>ECDH second event</th>\n",
        "      <td> 0.233143</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>ECDSA sign 200 bytes</th>\n",
        "      <td> 1.178730</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>ECDSA verify 200 bytes</th>\n",
        "      <td> 2.033945</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>Generate certificate</th>\n",
        "      <td> 0.187457</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>Verify a certificate</th>\n",
        "      <td> 2.013429</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
        "</div>"
       ],
       "output_type": "pyout",
       "prompt_number": 9,
       "text": [
        "                                  duration\n",
        "curve     event                           \n",
        "secp128r1 ECDH first event        0.111218\n",
        "          ECDH second event       0.171697\n",
        "          ECDSA sign 200 bytes    0.788145\n",
        "          ECDSA verify 200 bytes  1.382880\n",
        "          Generate certificate    0.124332\n",
        "          Verify a certificate    1.390121\n",
        "secp160k1 ECDH first event        0.136917\n",
        "          ECDH second event       0.188761\n",
        "          ECDSA sign 200 bytes    0.842946\n",
        "          ECDSA verify 200 bytes  1.409523\n",
        "          Generate certificate    0.149816\n",
        "          Verify a certificate    1.417879\n",
        "secp192k1 ECDH first event        0.173947\n",
        "          ECDH second event       0.233143\n",
        "          ECDSA sign 200 bytes    1.178730\n",
        "          ECDSA verify 200 bytes  2.033945\n",
        "          Generate certificate    0.187457\n",
        "          Verify a certificate    2.013429"
       ]
      }
     ],
     "prompt_number": 9
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# same data, presented differently\n",
      "res.unstack(0)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
        "<table border=\"1\" class=\"dataframe\">\n",
        "  <thead>\n",
        "    <tr>\n",
        "      <th></th>\n",
        "      <th colspan=\"3\" halign=\"left\">duration</th>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>curve</th>\n",
        "      <th>secp128r1</th>\n",
        "      <th>secp160k1</th>\n",
        "      <th>secp192k1</th>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>event</th>\n",
        "      <th></th>\n",
        "      <th></th>\n",
        "      <th></th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>ECDH first event</th>\n",
        "      <td> 0.111218</td>\n",
        "      <td> 0.136917</td>\n",
        "      <td> 0.173947</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>ECDH second event</th>\n",
        "      <td> 0.171697</td>\n",
        "      <td> 0.188761</td>\n",
        "      <td> 0.233143</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>ECDSA sign 200 bytes</th>\n",
        "      <td> 0.788145</td>\n",
        "      <td> 0.842946</td>\n",
        "      <td> 1.178730</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>ECDSA verify 200 bytes</th>\n",
        "      <td> 1.382880</td>\n",
        "      <td> 1.409523</td>\n",
        "      <td> 2.033945</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>Generate certificate</th>\n",
        "      <td> 0.124332</td>\n",
        "      <td> 0.149816</td>\n",
        "      <td> 0.187457</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>Verify a certificate</th>\n",
        "      <td> 1.390121</td>\n",
        "      <td> 1.417879</td>\n",
        "      <td> 2.013429</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
        "</div>"
       ],
       "output_type": "pyout",
       "prompt_number": 10,
       "text": [
        "                         duration                      \n",
        "curve                   secp128r1  secp160k1  secp192k1\n",
        "event                                                  \n",
        "ECDH first event         0.111218   0.136917   0.173947\n",
        "ECDH second event        0.171697   0.188761   0.233143\n",
        "ECDSA sign 200 bytes     0.788145   0.842946   1.178730\n",
        "ECDSA verify 200 bytes   1.382880   1.409523   2.033945\n",
        "Generate certificate     0.124332   0.149816   0.187457\n",
        "Verify a certificate     1.390121   1.417879   2.013429"
       ]
      }
     ],
     "prompt_number": 10
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "unstacked_res =  res.unstack(0)\n",
      "unstacked_res.columns = [curve for (name, curve) in  unstacked_res.columns]\n",
      "\n",
      "del unstacked_res.index.name\n",
      "#unstacked_res.sort_index(axis=1)\n",
      "#unstacked_res.index.name = \"Cryptographic primitive\"\n",
      "unstacked_res.plot(kind=\"bar\")\n",
      "plt.xticks(rotation=65)\n",
      "plt.ylabel(\"duration (in seconds)\")\n",
      "plt.legend(title=\"Curves\", loc=2)\n",
      "plt.show()\n",
      "# plt.savefig(\"crypto-speed.pdf\", bbox_inches=\"tight\", dpi=300)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAYMAAAFoCAYAAAC155dTAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n"
      }
     ],
     "prompt_number": 11
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 11
    }
   ],
   "metadata": {}
  }
 ]
}