{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "23a75407",
   "metadata": {
    "id": "23a75407"
   },
   "source": "# OceanML — geometriden bağımsız alıcı eğitimi (Google Colab)\n\n**BVK Technology · OceanChannel v1.1 tabanlı · 2026-10**\n\nBu defter, modemin \"eğitilmemiş bölgede BER yüksek\" sorununu üç adımda ele alır:\n\n1. **Teşhis.** Kanal, geometri 1–3 cm değişince kendine benzemez hale gelir (44 kHz'de dalga boyu 3,4 cm). 30 cm'lik bölgeler bu yüzden eğitilmiş noktaların arasını kapsamaz. Kutudaki farklı kanal sayısı 10⁸–10¹⁰ mertebesindedir; veriyle \"tüm geometrileri\" kaplamak mümkün değildir.\n2. **Çözüm: kanal-koşullu alıcı (KKA).** Geometri ezberlenmez. Her paketin kanalı preamble + 20 pilot sembolden ölçülür, ardından sıralı olarak karar verilen sembollerle iyileştirilir. Geçmiş sembollerin yankıları bu kanalla çıkarılır ve karar J+1 sembollük pencerede verilir. Bu yapıda, eğitimde görülmemiş bir geometri diye bir kavram kalmaz.\n3. **Öğrenme, genellenebilen yerde.** *Öğrenen kanal önseli* bir CNN'dir: her geometride ortak olan kanal yapısını (birkaç varış, her biri ön-uç biçimli) öğrenir ve düşük SNR'de kanal tahminini temizler. Ağ rastgele sürekli geometrilerle eğitilir.\n\nVeriler diske yazılmaz, her adımda anında üretilir. 20 GB'lık shard deposu gerekmez.\n\n> Bu paketteki sayılar, sözlük dosyası (`GlossaryQ_10P_200samplesX.mat`) ve `getTapChannel` / `getHarborChannel` elimizde olmadığı için **yer tutucu sözlük** ve OceanChannel dokümanı §4'teki görüntü yöntemiyle üretildi. §3'te kendi dosyalarınızı verdiğinizde aynı hücreler sizin sözlüğünüzle çalışır."
  },
  {
   "cell_type": "markdown",
   "id": "1cb5022d",
   "metadata": {
    "id": "1cb5022d"
   },
   "source": "## 0. Kurulum"
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "b175c08f",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-08T19:20:31.339135Z",
     "iopub.status.busy": "2026-10-08T19:20:31.338993Z",
     "iopub.status.idle": "2026-10-08T19:20:32.910250Z",
     "shell.execute_reply": "2026-10-08T19:20:32.909157Z"
    },
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 616
    },
    "id": "b175c08f",
    "executionInfo": {
     "status": "ok",
     "timestamp": 1791567052596,
     "user_tz": -180,
     "elapsed": 255598,
     "user": {
      "displayName": "Joseph Volkan ÖZCAN",
      "userId": "11389530967736397480"
     }
    },
    "outputId": "3e53fa33-82c4-4f4e-ce85-2f1d06a719ff"
   },
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Mounted at /content/drive\n",
      "Drive icinde oceanml_colab*.zip bulunamadi; dosyayi secin:\n"
     ]
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ],
      "text/html": [
       "\n",
       "     <input type=\"file\" id=\"files-4c12b9ea-7fb2-4505-ab61-e94692796af0\" name=\"files[]\" multiple disabled\n",
       "        style=\"border:none\" />\n",
       "     <output id=\"result-4c12b9ea-7fb2-4505-ab61-e94692796af0\">\n",
       "      Upload widget is only available when the cell has been executed in the\n",
       "      current browser session. Please rerun this cell to enable.\n",
       "      </output>\n",
       "      <script>// Copyright 2017 Google LLC\n",
       "//\n",
       "// Licensed under the Apache License, Version 2.0 (the \"License\");\n",
       "// you may not use this file except in compliance with the License.\n",
       "// You may obtain a copy of the License at\n",
       "//\n",
       "//      http://www.apache.org/licenses/LICENSE-2.0\n",
       "//\n",
       "// Unless required by applicable law or agreed to in writing, software\n",
       "// distributed under the License is distributed on an \"AS IS\" BASIS,\n",
       "// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n",
       "// See the License for the specific language governing permissions and\n",
       "// limitations under the License.\n",
       "\n",
       "/**\n",
       " * @fileoverview Helpers for google.colab Python module.\n",
       " */\n",
       "(function(scope) {\n",
       "function span(text, styleAttributes = {}) {\n",
       "  const element = document.createElement('span');\n",
       "  element.textContent = text;\n",
       "  for (const key of Object.keys(styleAttributes)) {\n",
       "    element.style[key] = styleAttributes[key];\n",
       "  }\n",
       "  return element;\n",
       "}\n",
       "\n",
       "// Max number of bytes which will be uploaded at a time.\n",
       "const MAX_PAYLOAD_SIZE = 100 * 1024;\n",
       "\n",
       "function _uploadFiles(inputId, outputId) {\n",
       "  const steps = uploadFilesStep(inputId, outputId);\n",
       "  const outputElement = document.getElementById(outputId);\n",
       "  // Cache steps on the outputElement to make it available for the next call\n",
       "  // to uploadFilesContinue from Python.\n",
       "  outputElement.steps = steps;\n",
       "\n",
       "  return _uploadFilesContinue(outputId);\n",
       "}\n",
       "\n",
       "// This is roughly an async generator (not supported in the browser yet),\n",
       "// where there are multiple asynchronous steps and the Python side is going\n",
       "// to poll for completion of each step.\n",
       "// This uses a Promise to block the python side on completion of each step,\n",
       "// then passes the result of the previous step as the input to the next step.\n",
       "function _uploadFilesContinue(outputId) {\n",
       "  const outputElement = document.getElementById(outputId);\n",
       "  const steps = outputElement.steps;\n",
       "\n",
       "  const next = steps.next(outputElement.lastPromiseValue);\n",
       "  return Promise.resolve(next.value.promise).then((value) => {\n",
       "    // Cache the last promise value to make it available to the next\n",
       "    // step of the generator.\n",
       "    outputElement.lastPromiseValue = value;\n",
       "    return next.value.response;\n",
       "  });\n",
       "}\n",
       "\n",
       "/**\n",
       " * Generator function which is called between each async step of the upload\n",
       " * process.\n",
       " * @param {string} inputId Element ID of the input file picker element.\n",
       " * @param {string} outputId Element ID of the output display.\n",
       " * @return {!Iterable<!Object>} Iterable of next steps.\n",
       " */\n",
       "function* uploadFilesStep(inputId, outputId) {\n",
       "  const inputElement = document.getElementById(inputId);\n",
       "  inputElement.disabled = false;\n",
       "\n",
       "  const outputElement = document.getElementById(outputId);\n",
       "  outputElement.innerHTML = '';\n",
       "\n",
       "  const pickedPromise = new Promise((resolve) => {\n",
       "    inputElement.addEventListener('change', (e) => {\n",
       "      resolve(e.target.files);\n",
       "    });\n",
       "  });\n",
       "\n",
       "  const cancel = document.createElement('button');\n",
       "  inputElement.parentElement.appendChild(cancel);\n",
       "  cancel.textContent = 'Cancel upload';\n",
       "  const cancelPromise = new Promise((resolve) => {\n",
       "    cancel.onclick = () => {\n",
       "      resolve(null);\n",
       "    };\n",
       "  });\n",
       "\n",
       "  // Wait for the user to pick the files.\n",
       "  const files = yield {\n",
       "    promise: Promise.race([pickedPromise, cancelPromise]),\n",
       "    response: {\n",
       "      action: 'starting',\n",
       "    }\n",
       "  };\n",
       "\n",
       "  cancel.remove();\n",
       "\n",
       "  // Disable the input element since further picks are not allowed.\n",
       "  inputElement.disabled = true;\n",
       "\n",
       "  if (!files) {\n",
       "    return {\n",
       "      response: {\n",
       "        action: 'complete',\n",
       "      }\n",
       "    };\n",
       "  }\n",
       "\n",
       "  for (const file of files) {\n",
       "    const li = document.createElement('li');\n",
       "    li.append(span(file.name, {fontWeight: 'bold'}));\n",
       "    li.append(span(\n",
       "        `(${file.type || 'n/a'}) - ${file.size} bytes, ` +\n",
       "        `last modified: ${\n",
       "            file.lastModifiedDate ? file.lastModifiedDate.toLocaleDateString() :\n",
       "                                    'n/a'} - `));\n",
       "    const percent = span('0% done');\n",
       "    li.appendChild(percent);\n",
       "\n",
       "    outputElement.appendChild(li);\n",
       "\n",
       "    const fileDataPromise = new Promise((resolve) => {\n",
       "      const reader = new FileReader();\n",
       "      reader.onload = (e) => {\n",
       "        resolve(e.target.result);\n",
       "      };\n",
       "      reader.readAsArrayBuffer(file);\n",
       "    });\n",
       "    // Wait for the data to be ready.\n",
       "    let fileData = yield {\n",
       "      promise: fileDataPromise,\n",
       "      response: {\n",
       "        action: 'continue',\n",
       "      }\n",
       "    };\n",
       "\n",
       "    // Use a chunked sending to avoid message size limits. See b/62115660.\n",
       "    let position = 0;\n",
       "    do {\n",
       "      const length = Math.min(fileData.byteLength - position, MAX_PAYLOAD_SIZE);\n",
       "      const chunk = new Uint8Array(fileData, position, length);\n",
       "      position += length;\n",
       "\n",
       "      const base64 = btoa(String.fromCharCode.apply(null, chunk));\n",
       "      yield {\n",
       "        response: {\n",
       "          action: 'append',\n",
       "          file: file.name,\n",
       "          data: base64,\n",
       "        },\n",
       "      };\n",
       "\n",
       "      let percentDone = fileData.byteLength === 0 ?\n",
       "          100 :\n",
       "          Math.round((position / fileData.byteLength) * 100);\n",
       "      percent.textContent = `${percentDone}% done`;\n",
       "\n",
       "    } while (position < fileData.byteLength);\n",
       "  }\n",
       "\n",
       "  // All done.\n",
       "  yield {\n",
       "    response: {\n",
       "      action: 'complete',\n",
       "    }\n",
       "  };\n",
       "}\n",
       "\n",
       "scope.google = scope.google || {};\n",
       "scope.google.colab = scope.google.colab || {};\n",
       "scope.google.colab._files = {\n",
       "  _uploadFiles,\n",
       "  _uploadFilesContinue,\n",
       "};\n",
       "})(self);\n",
       "</script> "
      ]
     },
     "metadata": {}
    },
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Saving oceanml_colab.zip to oceanml_colab.zip\n",
      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m664.0/664.0 kB\u001b[0m \u001b[31m12.5 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[?25h  Installing build dependencies ... \u001b[?25l\u001b[?25hdone\n",
      "  Getting requirements to build wheel ... \u001b[?25l\u001b[?25hdone\n",
      "  Preparing metadata (pyproject.toml) ... \u001b[?25l\u001b[?25hdone\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m120.8/120.8 kB\u001b[0m \u001b[31m10.1 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m9.8/9.8 MB\u001b[0m \u001b[31m95.5 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m237.1/237.1 kB\u001b[0m \u001b[31m18.5 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m327.8/327.8 kB\u001b[0m \u001b[31m23.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m2.0/2.0 MB\u001b[0m \u001b[31m79.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m8.7/8.7 MB\u001b[0m \u001b[31m91.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m170.9/170.9 kB\u001b[0m \u001b[31m13.3 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m131.7/131.7 kB\u001b[0m \u001b[31m10.5 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m15.3/15.3 MB\u001b[0m \u001b[31m74.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m3.4/3.4 MB\u001b[0m \u001b[31m19.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m2.5/2.5 MB\u001b[0m \u001b[31m77.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m412.6/412.6 kB\u001b[0m \u001b[31m26.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m120.7/120.7 kB\u001b[0m \u001b[31m9.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m4.0/4.0 MB\u001b[0m \u001b[31m22.5 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m95.7/95.7 kB\u001b[0m \u001b[31m7.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m113.3/113.3 kB\u001b[0m \u001b[31m9.3 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m2.4/2.4 MB\u001b[0m \u001b[31m73.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m234.2/234.2 kB\u001b[0m \u001b[31m17.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m54.9/54.9 kB\u001b[0m \u001b[31m4.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m45.5/45.5 kB\u001b[0m \u001b[31m3.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m76.8/76.8 kB\u001b[0m \u001b[31m5.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m131.0/131.0 kB\u001b[0m \u001b[31m10.3 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m541.5/541.5 kB\u001b[0m \u001b[31m34.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
      "\u001b[?25h  Building wheel for docplex (pyproject.toml) ... \u001b[?25l\u001b[?25hdone\n",
      "device cpu | torch 2.11.0+cpu\n"
     ]
    }
   ],
   "source": "import os, sys, time, json, math, glob, zipfile\nIN_COLAB = 'google.colab' in sys.modules\nif IN_COLAB and not os.path.isdir('/content/oceanml/oceanml'):\n    from google.colab import drive, files\n    drive.mount('/content/drive')\n    # paketi Drive'da ara (MyDrive kökü ve bir alt klasör derinliği), yoksa doğrudan yükle\n    hits = glob.glob('/content/drive/MyDrive/oceanml_colab*.zip') + glob.glob('/content/drive/MyDrive/*/oceanml_colab*.zip')\n    if hits:\n        ZIP = sorted(hits, key=os.path.getmtime)[-1]; print('Drive paketi:', ZIP)\n    else:\n        print(\"Drive icinde oceanml_colab*.zip bulunamadi; dosyayi secin:\")\n        up = files.upload(); ZIP = '/content/' + next(iter(up));\n    with zipfile.ZipFile(ZIP) as z:\n        z.extractall('/content')\n    assert os.path.isdir('/content/oceanml/oceanml'), 'zip içinde oceanml/ klasörü yok: ' + str(os.listdir('/content'))\n    !pip -q install PyWavelets dwave-neal h5py dwave-system qiskit qiskit-optimization qiskit-algorithms qiskit-ibm-runtime\nif IN_COLAB:\n    os.chdir('/content/oceanml')\nelif not os.path.isdir('oceanml') and os.path.isdir('../oceanml'):\n    os.chdir('..')                                         # yerel: depo kökünden çalış\nsys.path.insert(0, os.getcwd())\nimport numpy as np, torch, matplotlib.pyplot as plt\nRESULTS = {}        # anlatımlı rapor (§12) için koşu çıktıları burada toplanır\nDEVICE = 'cuda' if torch.cuda.is_available() else 'cpu'\nQUICK = True        # False: tam boy koşu (Colab GPU'da ~1-2 saat)\nprint('device', DEVICE, '| torch', torch.__version__)"
  },
  {
   "cell_type": "markdown",
   "id": "fdd9f1ad",
   "metadata": {
    "id": "fdd9f1ad"
   },
   "source": "## 1. Kanal derleyici — doküman §4 ile birebir\n\n`oceanml/channel.py`, `och_compile.py`'nin görüntü yöntemini Python'a taşır: dört görüntü ailesi, Rayleigh dip yansıması, yüzey kaybı, eşik, nicemleme ve ortak ölçek. İki modu vardır. *Fizik* modu dokümandaki derleyicidir. *MATLAB-uyumlu* mod ise GEO_TABLE / `build_geometry_table.m` ile aynı formülü kullanır.\n\nDokümanın iki örneği doğrulama testi olarak kullanılıyor. Not: §5.1'deki gerçek çıktı (1 km, kum: 29 yankı, son gecikme 15943 µs) yalnızca aynı gecikmeye düşen yolların **eşikten önce** toplanmasıyla tutuyor. §4.12'deki metin ise sırayı \"önce eşik, sonra birleştir\" diye veriyor. Derleyici kodu ile metin arasındaki bu farkın dokümanda düzeltilmesi önerilir."
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f84922db",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-08T19:20:32.912398Z",
     "iopub.status.busy": "2026-10-08T19:20:32.912175Z",
     "iopub.status.idle": "2026-10-08T19:20:32.923868Z",
     "shell.execute_reply": "2026-10-08T19:20:32.922587Z"
    },
    "id": "f84922db",
    "outputId": "77866be5-ab5f-47a4-9bb6-09112d3a094a"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "§4.12  (d, a): [(0, 1.0), (961, -0.832), (2149, 0.453), (4538, -0.277), (6422, -0.237), (9540, 0.175)]\n",
      "§5.1   tap 29+doğrudan | ölçek 0.568 | rms 2.326 | tepe 10.82 | ilk 12 µs | son 15943 µs | TL 74.8 dB\n"
     ]
    }
   ],
   "source": "from oceanml.channel import Scenario, compile_scenario\nc = compile_scenario(Scenario(r=20, zs=3, zr=5, D=10, bottom='kum', surf_loss_db=1.0))\nprint('§4.12  (d, a):', [(int(d), round(float(a), 3)) for d, a in zip(c.d, c.a)])\nc = compile_scenario(Scenario(r=1000, zs=3, zr=3, D=10, bottom='kum'))\nprint(f'§5.1   tap {len(c.d)-1}+doğrudan | ölçek {c.scale:.3f} | rms {c.rms_gain:.3f} | tepe {c.peak_gain:.2f} | '\n      f'ilk {c.d[1]} µs | son {c.d[-1]} µs | TL {c.tl_db:.1f} dB')"
  },
  {
   "cell_type": "markdown",
   "id": "9f99fdcc",
   "metadata": {
    "id": "9f99fdcc"
   },
   "source": "## 2. Teşhis — 30 cm neden yetmiyor\n\nGeometri δ kadar kaydırıldığında yankı kısmının bant içi (38–50 kHz) dalga biçimi korelasyonu ölçülür. Doğrudan yol çıkarılmıştır."
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "a5c9979c",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-08T19:20:32.925709Z",
     "iopub.status.busy": "2026-10-08T19:20:32.925531Z",
     "iopub.status.idle": "2026-10-08T19:20:41.591437Z",
     "shell.execute_reply": "2026-10-08T19:20:41.590216Z"
    },
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 366
    },
    "id": "a5c9979c",
    "executionInfo": {
     "status": "ok",
     "timestamp": 1791567111435,
     "user_tz": -180,
     "elapsed": 23672,
     "user": {
      "displayName": "Joseph Volkan ÖZCAN",
      "userId": "11389530967736397480"
     }
    },
    "outputId": "1b353523-5751-4dc8-fddc-141c8a6191f0"
   },
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "<Figure size 1100x360 with 2 Axes>"
      ],
      "image/png": "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\n"
     },
     "metadata": {}
    }
   ],
   "source": "from oceanml.coverage import axis_coherence\nD = [0, 0.0025, 0.005, 0.0075, 0.01, 0.015, 0.02, 0.03, 0.05, 0.1]\nspan = dict(r=16, zs=3, zr=3, D=10)\nfig, axs = plt.subplots(1, 2, figsize=(11, 3.6))\nfor ax, mode in zip(axs, ['matlab', 'physics']):\n    cells = 1\n    for a in ['r', 'zs', 'zr', 'D']:\n        med, lc = axis_coherence(a, D, n=120 if QUICK else 400, mode=mode)\n        ax.plot([d * 100 for d in D], [med[d] for d in D], 'o-', label=f'{a}  (L_c≈{lc*100:.1f} cm)'); cells *= span[a] / lc\n        RESULTS.setdefault('coherence', {}).setdefault(mode, {})[a] = lc * 100\n    ax.axvline(30, color='k', ls='--'); ax.set_xscale('symlog', linthresh=1); ax.set_xlabel('kayma δ [cm]')\n    ax.set_title(f'{mode}: ~{cells:.0e} ayrı kanal (30 cm ızgara: 1.8e5 hücre)'); ax.set_ylabel('yankı korelasyonu'); ax.legend(fontsize=8); ax.grid(alpha=.3)\nplt.tight_layout(); plt.show()"
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "b3667d47",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-08T19:20:41.593541Z",
     "iopub.status.busy": "2026-10-08T19:20:41.593217Z",
     "iopub.status.idle": "2026-10-08T19:20:42.797812Z",
     "shell.execute_reply": "2026-10-08T19:20:42.797043Z"
    },
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "b3667d47",
    "executionInfo": {
     "status": "ok",
     "timestamp": 1791567118632,
     "user_tz": -180,
     "elapsed": 2472,
     "user": {
      "displayName": "Joseph Volkan ÖZCAN",
      "userId": "11389530967736397480"
     }
    },
    "outputId": "dabae5df-8e97-43e3-9dda-95f06c48786d"
   },
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "matlab  : yankı yayılımı medyan 14.7 ms (p90 23.1); enerjinin 36%'i (medyan) MLP'nin 2-sembol penceresinin DIŞINDAKİ sembollerden geliyor\n",
      "physics : yankı yayılımı medyan 4.2 ms (p90 7.6); enerjinin 8%'i (medyan) MLP'nin 2-sembol penceresinin DIŞINDAKİ sembollerden geliyor\n"
     ]
    }
   ],
   "source": "from oceanml.coherence import isi_stats\nfor mode in ['matlab', 'physics']:\n    st = isi_stats(n=800, mode=mode)\n    RESULTS.setdefault('isi', {})[mode] = (float(np.median(st['spread_ms'])), float(np.median(st['unseen'])))\n    print(f\"{mode:8s}: yankı yayılımı medyan {np.median(st['spread_ms']):.1f} ms (p90 {np.percentile(st['spread_ms'],90):.1f}); \"\n          f\"enerjinin {100*np.median(st['unseen']):.0f}%'i (medyan) MLP'nin 2-sembol penceresinin DIŞINDAKİ sembollerden geliyor\")"
  },
  {
   "cell_type": "markdown",
   "id": "4317a2fc",
   "metadata": {
    "id": "4317a2fc"
   },
   "source": "## 3. Kendi dosyalarınız (isteğe bağlı)\n\nYollar doluysa yer tutucu yerine sizin sözlüğünüz kullanılır. v012 ağınız da aynı test kurulumlarında değerlendirilir. GEO_TABLE verilirse derleyicinin MATLAB modu, tablodaki tap listeleriyle birebir karşılaştırılır (doküman T-1)."
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "41b3cd71",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-08T19:20:42.800437Z",
     "iopub.status.busy": "2026-10-08T19:20:42.800197Z",
     "iopub.status.idle": "2026-10-08T19:20:43.196207Z",
     "shell.execute_reply": "2026-10-08T19:20:43.194682Z"
    },
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "41b3cd71",
    "executionInfo": {
     "status": "ok",
     "timestamp": 1791567125383,
     "user_tz": -180,
     "elapsed": 1117,
     "user": {
      "displayName": "Joseph Volkan ÖZCAN",
      "userId": "11389530967736397480"
     }
    },
    "outputId": "417d926f-df84-4763-b3a9-627649a0ffa9"
   },
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "sözlük: GERÇEK\n",
      "GEO_TABLE: 9746 geometri; tap kanalı olarak tablonun kendisi kullanılacak (örtük doğrudan yol d=0, a=1 eklenir)\n"
     ]
    }
   ],
   "source": "# Gerçek dosyalar pakette (data_real/). Başka bir sürüm kullanmak için yolları değiştirin.\nGLOSSARY = 'data_real/GlossaryQ_10P_200samplesX.mat'\nNET_V012 = 'data_real/net_v012_table.mat'\nGEO_TABLE = 'data_real/GEO_TABLE.mat'\nfrom oceanml.sim import Modem, load_glossary\nfrom oceanml import matlab_io\nwords = load_glossary(GLOSSARY) if os.path.exists(GLOSSARY) else None\nprint('sözlük:', 'GERÇEK' if words is not None else 'yer tutucu (M=2, 200 örnek)')\nnet_user = matlab_io.mlp_from_matlab(NET_V012) if os.path.exists(NET_V012) else None\nUSE_GEO = os.path.exists(GEO_TABLE)\nif USE_GEO:\n    from oceanml.geometry import load_real_geo\n    G_real = load_real_geo(GEO_TABLE)\n    print('GEO_TABLE:', len(G_real), 'geometri; tap kanalı olarak tablonun kendisi kullanılacak (örtük doğrudan yol d=0, a=1 eklenir)')"
  },
  {
   "cell_type": "markdown",
   "id": "f774eb6d",
   "metadata": {
    "id": "f774eb6d"
   },
   "source": "## 4. Anında veri üretimi (20 GB yerine)\n\nPaket: önek + chirp preamble + sözcükler + koruma aralığı. Ardından TX ön-ucu, çokyol, RX ön-ucu, \"ölçülen\" SNR'de AWGN ve rastgele polarite gelir. Ön-uç ailesi v012 üretim kuralına göre çekilir. Özellikler `dwt_test_v012.m` ile birebirdir: db2 seviye-1 yaklaşık katsayılar ('sym' uzatma), maks-mutlak normalizasyon ve [önceki; şimdiki] birleşimi."
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "e5475f97",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-08T19:20:43.198656Z",
     "iopub.status.busy": "2026-10-08T19:20:43.198346Z",
     "iopub.status.idle": "2026-10-08T19:20:44.220714Z",
     "shell.execute_reply": "2026-10-08T19:20:44.220139Z"
    },
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "e5475f97",
    "executionInfo": {
     "status": "ok",
     "timestamp": 1791567134230,
     "user_tz": -180,
     "elapsed": 2788,
     "user": {
      "displayName": "Joseph Volkan ÖZCAN",
      "userId": "11389530967736397480"
     }
    },
    "outputId": "71a28099-7a01-4563-c32c-e3565ed387c3"
   },
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "24 paket (4800 sembol) 2.66 s | özellik (24, 200, 400) | kilit ofseti 78 örnek\n"
     ]
    }
   ],
   "source": "from oceanml.data import make_packets, SNR_LIST\nfrom oceanml.geometry import continuous, src_of\nm = Modem(words=words, nsym=200, seed=1, device=DEVICE)\nrng = np.random.default_rng(0)\nt0 = time.time(); rx, bits, lock, info = make_packets(m, [continuous(rng) for _ in range(24)], ['tap']*24, [8]*24)\nF = m.features_v012(rx, lock)\nprint(f'24 paket ({24*m.nsym} sembol) {time.time()-t0:.2f} s | özellik {tuple(F.shape)} | kilit ofseti {np.median(lock-m.pad):.0f} örnek')"
  },
  {
   "cell_type": "markdown",
   "id": "6bc84690",
   "metadata": {
    "id": "6bc84690"
   },
   "source": "## 5. Referans: v012 MLP (sürekli geometri ile eğitilmiş)\n\nAğ `ann_train_v012.m` ile aynıdır: 400 → 32 sigmoid → 1, BCE, Adam 2e-3, minibatch 2048, 8 epoch. Tek fark, sabit shard'lar yerine her paketin yeni ve sürekli bir geometriden üretilmesidir."
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "1de4017f",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-08T19:20:44.224318Z",
     "iopub.status.busy": "2026-10-08T19:20:44.224059Z",
     "iopub.status.idle": "2026-10-08T19:21:36.955199Z",
     "shell.execute_reply": "2026-10-08T19:21:36.954479Z"
    },
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "1de4017f",
    "executionInfo": {
     "status": "ok",
     "timestamp": 1791567225843,
     "user_tz": -180,
     "elapsed": 84736,
     "user": {
      "displayName": "Joseph Volkan ÖZCAN",
      "userId": "11389530967736397480"
     }
    },
    "outputId": "40359f94-e73c-492d-80a3-db0774a38501"
   },
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "eğitim örneği 240000 | eğitim hatası 0.07834000140428543\n"
     ]
    }
   ],
   "source": "from oceanml.nets import mlp_v012, train, predict\nn_pk = 1200 if QUICK else 5200\nXs, Ys = [], []\nr = np.random.default_rng(11)\nfor j in range(0, n_pk, 24):\n    rx, bits, lock, _ = make_packets(m, [continuous(r) for _ in range(24)], [src_of(j+i) for i in range(24)], r.choice(SNR_LIST, 24))\n    F = m.features_v012(rx, lock); Xs.append(F.reshape(-1, F.shape[-1]).half().cpu()); Ys.append(torch.tensor(bits.reshape(-1), dtype=torch.float32))\nX, Y = torch.cat(Xs), torch.cat(Ys); del Xs, Ys\nnet = train(mlp_v012(X.shape[1]).to(DEVICE), X, Y)\nRESULTS['train_err'] = float((predict(net, X[:200000]) != (Y[:200000] > .5)).float().mean())\nprint('eğitim örneği', len(Y), '| eğitim hatası', RESULTS['train_err'])"
  },
  {
   "cell_type": "markdown",
   "id": "696aa4e9",
   "metadata": {
    "id": "696aa4e9"
   },
   "source": "## 6. Kanal-koşullu alıcı (KKA) — görülmemiş geometrilerde\n\nTest kurulumları sürekli ve rastgele geometrilerdir; hiçbiri eğitimde yer almaz. KKA'nın zaten eğitilen bir parçası yoktur. Aynı paketler dört alıcıdan geçirilir:\n\n- v012 MLP\n- KKA (20 pilot)\n- KKA (pilotsuz, yinelemeli)\n- KKA (kanal bilinseydi, yani üst sınır)"
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "35386c30",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-08T19:21:36.957162Z",
     "iopub.status.busy": "2026-10-08T19:21:36.956876Z",
     "iopub.status.idle": "2026-10-08T19:23:20.476023Z",
     "shell.execute_reply": "2026-10-08T19:23:20.474530Z"
    },
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 543
    },
    "id": "35386c30",
    "executionInfo": {
     "status": "ok",
     "timestamp": 1791567634548,
     "user_tz": -180,
     "elapsed": 356732,
     "user": {
      "displayName": "Joseph Volkan ÖZCAN",
      "userId": "11389530967736397480"
     }
    },
    "outputId": "fc309a00-f065-4809-9c17-31a6d2c57d04"
   },
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "harbor -10 {'v012 MLP': 0.1005, 'KKA pilot': 0.1171, 'KKA pilot + önsel': 0.0148, 'KKA pilotsuz': 0.1472, 'KKA oracle': 0.0106, 'v012 (sizin ağ)': 0.0921}\n",
      "harbor -4 {'v012 MLP': 0.0347, 'KKA pilot': 0.0069, 'KKA pilot + önsel': 0.0, 'KKA pilotsuz': 0.0782, 'KKA oracle': 0.0, 'v012 (sizin ağ)': 0.0347}\n",
      "harbor 2 {'v012 MLP': 0.0208, 'KKA pilot': 0.0005, 'KKA pilot + önsel': 0.0, 'KKA pilotsuz': 0.05, 'KKA oracle': 0.0, 'v012 (sizin ağ)': 0.0074}\n",
      "harbor 8 {'v012 MLP': 0.0019, 'KKA pilot': 0.0, 'KKA pilot + önsel': 0.0, 'KKA pilotsuz': 0.0389, 'KKA oracle': 0.0, 'v012 (sizin ağ)': 0.0009}\n",
      "harbor 14 {'v012 MLP': 0.0, 'KKA pilot': 0.0, 'KKA pilot + önsel': 0.0, 'KKA pilotsuz': 0.0338, 'KKA oracle': 0.0, 'v012 (sizin ağ)': 0.0009}\n",
      "tap -10 {'v012 MLP': 0.1181, 'KKA pilot': 0.1366, 'KKA pilot + önsel': 0.0315, 'KKA pilotsuz': 0.1389, 'KKA oracle': 0.0069, 'v012 (sizin ağ)': 0.1167}\n",
      "tap -4 {'v012 MLP': 0.0505, 'KKA pilot': 0.0125, 'KKA pilot + önsel': 0.0, 'KKA pilotsuz': 0.0843, 'KKA oracle': 0.0, 'v012 (sizin ağ)': 0.0579}\n",
      "tap 2 {'v012 MLP': 0.0162, 'KKA pilot': 0.0, 'KKA pilot + önsel': 0.0, 'KKA pilotsuz': 0.0611, 'KKA oracle': 0.0, 'v012 (sizin ağ)': 0.0139}\n",
      "tap 8 {'v012 MLP': 0.0079, 'KKA pilot': 0.0019, 'KKA pilot + önsel': 0.0, 'KKA pilotsuz': 0.0847, 'KKA oracle': 0.0, 'v012 (sizin ağ)': 0.0181}\n",
      "tap 14 {'v012 MLP': 0.0125, 'KKA pilot': 0.0005, 'KKA pilot + önsel': 0.0, 'KKA pilotsuz': 0.0694, 'KKA oracle': 0.0, 'v012 (sizin ağ)': 0.0407}\n"
     ]
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "<Figure size 1100x360 with 2 Axes>"
      ],
      "image/png": "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\n"
     },
     "metadata": {}
    }
   ],
   "source": "from oceanml.receiver import KKA, KKASeq, KKAPrior\nfrom oceanml.chprior import ChannelPrior\nprior = ChannelPrior(c=32); prior.load_state_dict(torch.load('weights/prior_c32.pt', map_location=DEVICE))\nNI = 12 if QUICK else 60\nsnrs = [-10, -4, 2, 8, 14] if QUICK else list(SNR_LIST)\nPIL = 20\nkseq, kit, kpri = KKASeq(m, pilots=PIL), KKA(m), KKAPrior(m, prior, pilots=PIL)\nrt = np.random.default_rng(20260912); inst = [continuous(rt) for _ in range(NI)]\nif USE_GEO:                                   # eğitim seçimi DIŞINDAKİ gerçek geometriler\n    d1 = np.array([g['d'][0] for g in G_real]); ph = np.mod(d1, 1e6/44e3); co = np.floor(d1/50)\n    sel = set(np.lexsort((co, np.floor(ph/(1e6/44e3)*8)))[np.round(np.linspace(0, len(G_real)-1, 3000)).astype(int)].tolist())\n    geo_idx = rt.choice([i for i in range(len(G_real)) if i not in sel], NI, replace=False).tolist()\nR = {}\nfor src in ['harbor', 'tap']:\n    E = {a: np.zeros((NI, len(snrs))) for a in ['v012 MLP', 'KKA pilot', 'KKA pilot + önsel', 'KKA pilotsuz', 'KKA oracle'] + (['v012 (sizin ağ)'] if net_user else [])}\n    for si, snr in enumerate(snrs):\n        for j in range(0, NI, 12):\n            if src == 'tap' and USE_GEO:\n                rx, bits, lock, info = make_packets(m, [{'idx': i} for i in geo_idx[j:j+12]], ['geo']*12, [snr]*12)\n            else:\n                rx, bits, lock, info = make_packets(m, inst[j:j+12], [src]*12, [snr]*12)\n            F = m.features_v012(rx, lock).reshape(-1, 2*m.NDWT)\n            E['v012 MLP'][j:j+12, si] = (predict(net, F).numpy().reshape(bits.shape)[:, PIL:] != bits[:, PIL:]).mean(1)\n            if net_user:\n                E['v012 (sizin ağ)'][j:j+12, si] = (predict(net_user.to(DEVICE), F).numpy().reshape(bits.shape)[:, PIL:] != bits[:, PIL:]).mean(1)\n            b, _ = kseq.run(rx, lock, true_bits=bits); E['KKA pilot'][j:j+12, si] = (b[-1][:, PIL:] != bits[:, PIL:]).mean(1)\n            b, _ = kpri.run(rx, lock, true_bits=bits); E['KKA pilot + önsel'][j:j+12, si] = (b[-1][:, PIL:] != bits[:, PIL:]).mean(1)\n            b, _ = kit.run(rx, lock, iters=3);           E['KKA pilotsuz'][j:j+12, si] = (b[-1][:, PIL:] != bits[:, PIL:]).mean(1)\n            b, _ = kit.run(rx, lock, oracle_h=kit.oracle(info, lock)); E['KKA oracle'][j:j+12, si] = (b[0][:, PIL:] != bits[:, PIL:]).mean(1)\n        print(src, snr, {k: round(float(v[:, si].mean()), 4) for k, v in E.items()})\n    R[src] = E\nRESULTS['exp2'] = dict(snrs=list(snrs), NI=NI, R=R)\nfig, axs = plt.subplots(1, 2, figsize=(11, 3.6))\nfor ax, src in zip(axs, R):\n    for k, v in R[src].items(): ax.semilogy(snrs, np.maximum(v.mean(0), 1e-5), 'o-', label=k)\n    ax.set_title(f\"{'GEO_TABLE (gerçek)' if src=='tap' and USE_GEO else src} kanalı, görülmemiş {NI} kurulum\"); ax.set_xlabel('ölçülen SNR [dB]'); ax.set_ylabel('BER'); ax.grid(alpha=.3); ax.legend(fontsize=8)\nplt.tight_layout(); plt.show()"
  },
  {
   "cell_type": "markdown",
   "id": "9d98d055",
   "metadata": {
    "id": "9d98d055"
   },
   "source": "## 7. Öğrenen kanal önseli (düşük SNR için)\n\nKKA'nın kalan hataları düşük SNR'de, kanal tahmini gürültülüyken ortaya çıkar. Önsel, tahmin edilen kanalı karmaşık temel banda (40 kHz) indirir. Ardından bir CNN, \"birkaç varış + ön-uç biçimi\" yapısını kullanarak gürültüyü temizler. Ağ rastgele sürekli geometrilerle eğitilir ve geometri ezberlemez. Paketle gelen ağırlıklar (`weights/prior_c32.pt`) bu ortamda üretildi; `TRAIN_PRIOR=True` ile yeniden eğitebilirsiniz."
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "0775a3dd",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-08T19:23:20.478261Z",
     "iopub.status.busy": "2026-10-08T19:23:20.478068Z",
     "iopub.status.idle": "2026-10-08T19:23:49.005403Z",
     "shell.execute_reply": "2026-10-08T19:23:49.004320Z"
    },
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "0775a3dd",
    "executionInfo": {
     "status": "ok",
     "timestamp": 1791567846172,
     "user_tz": -180,
     "elapsed": 103848,
     "user": {
      "displayName": "Joseph Volkan ÖZCAN",
      "userId": "11389530967736397480"
     }
    },
    "outputId": "573937bf-0751-4cdb-af30-81dca6059ed2"
   },
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "harbor -10 dB  KKA 0.1009  ->  KKA+önsel 0.0157\n",
      "harbor -6 dB  KKA 0.0204  ->  KKA+önsel 0.0005\n",
      "harbor -2 dB  KKA 0.0005  ->  KKA+önsel 0.0000\n",
      "tap    -10 dB  KKA 0.1880  ->  KKA+önsel 0.0431\n",
      "tap    -6 dB  KKA 0.0657  ->  KKA+önsel 0.0037\n",
      "tap    -2 dB  KKA 0.0088  ->  KKA+önsel 0.0000\n"
     ]
    }
   ],
   "source": "from oceanml.chprior import ChannelPrior\nfrom oceanml.receiver import KKAPrior\nTRAIN_PRIOR = False\nif TRAIN_PRIOR:\n    !python gen_prior_data.py 960 1000 runs/prior_train.pt && python gen_prior_data.py 120 2000 runs/prior_val.pt\n    !python train_prior.py runs/prior_train.pt runs/prior_val.pt weights/prior_c32.pt 2 1500\n    prior.load_state_dict(torch.load('weights/prior_c32.pt', map_location=DEVICE))\nkp = kpri\nfor src in ['harbor', 'tap']:\n    for snr in [-10, -6, -2]:\n        rx, bits, lock, info = make_packets(m, inst[:12], [src]*12, [snr]*12)\n        outs, _ = kp.run(rx, lock, true_bits=bits)\n        b0, b1 = float(np.mean(outs[-2][:, PIL:] != bits[:, PIL:])), float(np.mean(outs[-1][:, PIL:] != bits[:, PIL:]))\n        RESULTS.setdefault('prior', []).append((src, snr, b0, b1))\n        print(f'{src:6s} {snr:+d} dB  KKA {b0:.4f}  ->  KKA+önsel {b1:.4f}')"
  },
  {
   "cell_type": "markdown",
   "id": "1c80e9de",
   "metadata": {
    "id": "1c80e9de"
   },
   "source": "## 8. Kabul ölçütleri — koşudan ÖNCE sabitlendi\n\nv012 disiplini aynen korunuyor: ölçütler koşudan önce yazılır, koşudan sonra yalnızca PASS/FAIL okunur. Değerler v012'nin K1–K3'üyle aynıdır, ancak **tamamen görülmemiş geometrilerde** ölçülür. K3'te 5/9 hariç tutma kuralı kaldırılmıştır. Bunlara yeni bir K4 eklenmiştir.\n\n| Kod | Ölçüt |\n|---|---|\n| K1' | harbor + tap, tüm SNR'ler: ortalama BER ≤ 0,025 |\n| K2' | SNR ≥ +8: kurulum-BER medyanı ≤ 0,005 |\n| K3' | SNR ≥ +8: BER > 1e-2 olan kurulum ≤ %15 (20'de 3) |\n| K4 | görülmemiş/görülmüş farkı: KKA'da eğitilen parça olmadığı için yapısal olarak sağlanır; öğrenen önsel için ayrık doğrulama seti ile NMSE farkı ≤ 1 dB |"
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "e2760d4f",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-08T19:23:49.008054Z",
     "iopub.status.busy": "2026-10-08T19:23:49.007863Z",
     "iopub.status.idle": "2026-10-08T19:23:49.015409Z",
     "shell.execute_reply": "2026-10-08T19:23:49.014546Z"
    },
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "e2760d4f",
    "executionInfo": {
     "status": "ok",
     "timestamp": 1791568328958,
     "user_tz": -180,
     "elapsed": 29,
     "user": {
      "displayName": "Joseph Volkan ÖZCAN",
      "userId": "11389530967736397480"
     }
    },
    "outputId": "8484bcaf-d002-4f45-dfef-749045c54e88"
   },
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "v012 MLP         K1' FAIL (ort 0.0363)  K2' PASS (medyan 0.0)  K3' FAIL (outage 4/24)\n",
      "KKA pilot        K1' FAIL (ort 0.0276)  K2' PASS (medyan 0.0)  K3' PASS (outage 0/24)\n",
      "KKA pilot + önsel K1' PASS (ort 0.0046)  K2' PASS (medyan 0.0)  K3' PASS (outage 0/24)\n",
      "KKA pilotsuz     K1' FAIL (ort 0.0787)  K2' FAIL (medyan 0.0056)  K3' FAIL (outage 11/24)\n",
      "KKA oracle       K1' PASS (ort 0.0018)  K2' PASS (medyan 0.0)  K3' PASS (outage 0/24)\n",
      "v012 (sizin ağ)  K1' FAIL (ort 0.0383)  K2' PASS (medyan 0.0)  K3' FAIL (outage 6/24)\n"
     ]
    }
   ],
   "source": "def K(E):\n    v = np.concatenate([E['harbor'], E['tap']]); hi = np.array(snrs) >= 8\n    inst_b = v[:, hi].mean(1)\n    return dict(K1=v.mean() <= 0.025, K2=np.median(inst_b) <= 0.005, K3=(inst_b > 1e-2).mean() <= 0.15,\n                mean=round(float(v.mean()), 4), med=round(float(np.median(inst_b)), 4), out=f'{int((inst_b>1e-2).sum())}/{len(inst_b)}')\nfor a in R['harbor']:\n    k = K({s: R[s][a] for s in R})\n    print(f\"{a:16s} K1' {'PASS' if k['K1'] else 'FAIL'} (ort {k['mean']})  K2' {'PASS' if k['K2'] else 'FAIL'} (medyan {k['med']})  K3' {'PASS' if k['K3'] else 'FAIL'} (outage {k['out']})\")"
  },
  {
   "cell_type": "markdown",
   "id": "4cbb84af",
   "metadata": {
    "id": "4cbb84af"
   },
   "source": "## 9. Kuantum: nerede yardımcı olabilir, nerede olamaz\n\n- **Olamaz:** Ham dalga biçimi uzayını kuantum ya da klasik hiçbir hesaplama gücü kaplayamaz (§2: 10⁸–10¹⁰ ayrı kanal). Paket başına çalışan alıcı ise FPGA'da gerçek zamanlıdır, kuantum donanımı o döngüye giremez.\n- **Deneyebilir:** Öğrenen parçaların eğitim ve doğrulama geometrilerini, kanalın faz-bağımsız yapısını kaplayacak biçimde seçmek. Bu bir QUBO problemidir ve D-Wave / QAOA ile çözülebilir. Aşağıda rastgele seçim ve açgözlü k-merkez ile dürüstçe karşılaştırılıyor. Bu ortamda QUBO + tavlama, k-merkezi **geçemedi**."
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "da609f21",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-08T19:23:49.017849Z",
     "iopub.status.busy": "2026-10-08T19:23:49.017673Z",
     "iopub.status.idle": "2026-10-08T19:23:51.227957Z",
     "shell.execute_reply": "2026-10-08T19:23:51.226847Z"
    },
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "da609f21",
    "executionInfo": {
     "status": "ok",
     "timestamp": 1791568334838,
     "user_tz": -180,
     "elapsed": 5876,
     "user": {
      "displayName": "Joseph Volkan ÖZCAN",
      "userId": "11389530967736397480"
     }
    },
    "outputId": "04dd13e7-b60b-44b7-abd9-c2c87e557822"
   },
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "en kötü durum mesafesi (küçük = iyi): rastgele 0.493 | QUBO+SA 0.519 | k-merkez 0.364\n"
     ]
    }
   ],
   "source": "from oceanml.qcoreset import structure_features, similarity, build_qubo, solve_sa, greedy_kcenter, coverage\nfrom oceanml.coherence import sample_geom\nrng = np.random.default_rng(3)\npool = [sample_geom(rng) for _ in range(400 if QUICK else 600)]; held = [sample_geom(rng) for _ in range(1500)]\nFp = np.array([structure_features(g) for g in pool]); Fh = np.array([structure_features(g) for g in held])\nS, sig = similarity(Fp); K_ = 40\nrnd = np.mean([coverage(Fp[np.random.default_rng(s).choice(len(Fp), K_, replace=False)], Fh, sig)[1] for s in range(10)])\nkc = coverage(Fp[greedy_kcenter(Fp, K_)], Fh, sig)[1]\nx = solve_sa(build_qubo(S, K_, 0.5), num_reads=20); qb = coverage(Fp[x.astype(bool)], Fh, sig)[1]\nRESULTS['coreset'] = (float(rnd), float(qb), float(kc))\nprint(f'en kötü durum mesafesi (küçük = iyi): rastgele {rnd:.3f} | QUBO+SA {qb:.3f} | k-merkez {kc:.3f}')\n# D-Wave Leap (token gerekir):  from oceanml.qcoreset import solve_dwave; x = solve_dwave(build_qubo(S, K_, 0.5))\n# QAOA (Qiskit, <= ~25 değişken): from oceanml.qcoreset import solve_qaoa"
  },
  {
   "cell_type": "markdown",
   "id": "4cfaf903",
   "metadata": {
    "id": "4cfaf903"
   },
   "source": "## 10. Kuantum yolu: paket-seviyesi QUBO dedektörü (KKA-Q)\n\nKanal ölçüldükten sonra her veri biti $b_k\\in\\{0,1\\}$ kayda $g_{0,k} + b_k d_k$ ekler ($d = g_1 - g_0$, $g_b$ = sözcük $\\ast$ kanal). Bilinen her şey (preamble, pilotlar, $\\sum_k g_{0,k}$) çıkarıldıktan sonra paketin TAMAMI için en olası bit dizisi şudur:\n\n$$\\min_{b}\\ \\Big\\|r-\\sum_k b_k d_k\\Big\\|^2 = \\text{sabit} + \\sum_k Q_{kk}b_k + \\sum_{j<k} Q_{jk}b_jb_k,\\qquad Q_{kk}=G_{kk}-2\\langle r,d_k\\rangle,\\ \\ Q_{jk}=2G_{jk}$$\n\nBu tam olarak bir QUBO'dur ve kuantum tavlayıcının doğal girdisidir: 200 sembollük çerçevede 20 pilotla 180 değişken; bağlaşımlar yankı süresi kadar (40–80 komşu). DFE'den iki farkı var:\n\n1. Her bitin bütün yankı enerjisini kullanır.\n2. Hata yayılımı yoktur.\n\nÇözücüler:\n- **greedy:** DFE'den başlayan yerel iniş.\n- **SA:** klasik benzetimli tavlama.\n- **SB (simulated bifurcation):** kuantumdan esinlenmiş; FPGA'ya taşınabilir.\n- **D-Wave QPU / reverse annealing / Leap hybrid:** kuantum tavlama.\n- **QAOA:** IBM, en belirsiz bitlerin penceresinde.\n\nHepsi aynı ML enerjisiyle puanlanır ve en düşük enerjiyi veren seçilir.\n\n**Önemli:** QUBO, önselle temizlenmiş kanalla kazanç sağlar. Ham kanal tahminiyle optimum yanlış olur; çözücü bu yanlış optimumu yine de kusursuz bulur (exp5 öncesi ölçüm)."
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "8009d80f",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-08T19:23:51.230548Z",
     "iopub.status.busy": "2026-10-08T19:23:51.230151Z",
     "iopub.status.idle": "2026-10-08T19:24:28.156426Z",
     "shell.execute_reply": "2026-10-08T19:24:28.155835Z"
    },
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "8009d80f",
    "executionInfo": {
     "status": "ok",
     "timestamp": 1791568440095,
     "user_tz": -180,
     "elapsed": 105241,
     "user": {
      "displayName": "Joseph Volkan ÖZCAN",
      "userId": "11389530967736397480"
     }
    },
    "outputId": "ef9140aa-d974-43f9-a814-13cb1d047bb8"
   },
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "tap    -12 dB  DFE+önsel 0.0731  ->  QUBO 0.0185 | en iyi enerjiye ulaşan: {'dfe': 0.4166666567325592, 'greedy': 0.8333333134651184, 'sa': 1.0, 'sb': 0.9166666865348816}\n",
      "tap    -8 dB  DFE+önsel 0.0148  ->  QUBO 0.0005 | en iyi enerjiye ulaşan: {'dfe': 1.0, 'greedy': 1.0, 'sa': 1.0, 'sb': 0.9166666865348816}\n",
      "harbor -12 dB  DFE+önsel 0.0477  ->  QUBO 0.0181 | en iyi enerjiye ulaşan: {'dfe': 0.25, 'greedy': 1.0, 'sa': 1.0, 'sb': 1.0}\n"
     ]
    }
   ],
   "source": "from oceanml.qdetect import KKAQ, QuboBuilder\nkq = KKAQ(KKAPrior(m, prior, pilots=PIL), outer=1)\nfor src, snr in [('tap', -12), ('tap', -8), ('harbor', -12)]:\n    rx, bits, lock, info = make_packets(m, inst[:12], [src]*12, [snr]*12)\n    b, log = kq.run(rx, lock, bits)\n    tb = bits[:, PIL:]\n    reach = {n: float((log['E1'][n] <= torch.stack(list(log['E1'].values()), 1).min(1).values + 1e-6).float().mean()) for n in log['E1']}\n    RESULTS.setdefault('qubo', []).append((src, snr, float(np.mean(log['dfe'][:, PIL:] != tb)), float(np.mean(b[:, PIL:] != tb)), reach))\n    print(f\"{src:6s} {snr:+d} dB  DFE+önsel {np.mean(log['dfe'][:, PIL:] != tb):.4f}  ->  QUBO {np.mean(b[:, PIL:] != tb):.4f} | \"\n          f\"en iyi enerjiye ulaşan: {reach}\")"
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "88e2ddd8",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-08T19:24:28.157953Z",
     "iopub.status.busy": "2026-10-08T19:24:28.157678Z",
     "iopub.status.idle": "2026-10-08T19:24:28.163419Z",
     "shell.execute_reply": "2026-10-08T19:24:28.162863Z"
    },
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "88e2ddd8",
    "executionInfo": {
     "status": "ok",
     "timestamp": 1791568440096,
     "user_tz": -180,
     "elapsed": 7,
     "user": {
      "displayName": "Joseph Volkan ÖZCAN",
      "userId": "11389530967736397480"
     }
    },
    "outputId": "9b0f30e7-377d-4187-e2b2-63be261698d9"
   },
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "DWAVE_API_TOKEN yok: QPU hücresi atlandı (Colab: os.environ[\"DWAVE_API_TOKEN\"] = \"...\")\n"
     ]
    }
   ],
   "source": "# --- D-Wave (kendi hesabınızla): DWAVE_API_TOKEN ortam değişkeni gerekir ---\nUSE_DWAVE = bool(os.environ.get('DWAVE_API_TOKEN'))\nif USE_DWAVE:\n    from oceanml import qhw\n    rx, bits, lock, info = make_packets(m, inst[:2], ['tap']*2, [-12]*2)\n    outs, h = kq.ks.run(rx, lock, true_bits=bits); y = kq.ks.align(rx, lock)\n    Q, _, _ = kq.qb.build(y, h, outs[-1], PIL); Qn = Q.cpu().numpy()\n    emb = qhw.band_embedding(Qn.shape[1], max(qhw.bandwidth(q) for q in Qn))   # bir kez bul, yeniden kullan\n    for p in range(len(Qn)):\n        bq, info_q = qhw.solve_dwave_qpu(Qn[p], embedding=emb)\n        br, Er = qhw.solve_dwave_reverse(Qn[p], outs[-1][p, PIL:], embedding=emb)\n        bh = qhw.solve_leap_hybrid(Qn[p])\n        for name, bb in [('QPU', bq), ('reverse', br), ('hybrid', bh), ('DFE', outs[-1][p, PIL:])]:\n            RESULTS.setdefault('qhw', []).append(dict(packet=p, solver=name, energy=float(qhw.energy(Qn[p], bb)), ber=float(np.mean(bb != bits[p, PIL:])),\n                                                      chain_break=info_q.get('chain_break_fraction') if name == 'QPU' else None))\n            print(p, name, 'E=%.3f' % qhw.energy(Qn[p], bb), 'BER', np.mean(bb != bits[p, PIL:]), info_q.get('chain_break_fraction') if name == 'QPU' else '')\nelse:\n    print('DWAVE_API_TOKEN yok: QPU hücresi atlandı (Colab: os.environ[\"DWAVE_API_TOKEN\"] = \"...\")')"
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "32ef3a98",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-08T19:24:28.164985Z",
     "iopub.status.busy": "2026-10-08T19:24:28.164725Z",
     "iopub.status.idle": "2026-10-08T19:25:24.340125Z",
     "shell.execute_reply": "2026-10-08T19:25:24.339412Z"
    },
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "32ef3a98",
    "executionInfo": {
     "status": "ok",
     "timestamp": 1791568543618,
     "user_tz": -180,
     "elapsed": 103524,
     "user": {
      "displayName": "Joseph Volkan ÖZCAN",
      "userId": "11389530967736397480"
     }
    },
    "outputId": "86e97be4-b4b3-459c-b30e-e811ba0fdddb"
   },
   "outputs": [
    {
     "output_type": "stream",
     "name": "stderr",
     "text": [
      "/usr/local/lib/python3.13/dist-packages/scipy/sparse/linalg/_dsolve/linsolve.py:606: SparseEfficiencyWarning: splu converted its input to CSC format\n",
      "  return splu(A).solve\n",
      "/usr/local/lib/python3.13/dist-packages/scipy/sparse/linalg/_matfuncs.py:707: SparseEfficiencyWarning: spsolve is more efficient when sparse b is in the CSC matrix format\n",
      "  return spsolve(Q, P)\n"
     ]
    },
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "pencere [28, 36, 72, 84, 102, 117, 162, 164] | QAOA = kaba kuvvet optimum: True\n"
     ]
    }
   ],
   "source": "# --- IBM / QAOA: en belirsiz 8 bitlik pencere; yerelde statevector, IBM için backend verin ---\nfrom oceanml import qhw\nrx, bits, lock, info = make_packets(m, inst[:1], ['tap'], [-12])\nouts, h = kq.ks.run(rx, lock, true_bits=bits); y = kq.ks.align(rx, lock)\nQ, _, _ = kq.qb.build(y, h, outs[-1], PIL); Qn = Q.cpu().numpy()[0]; b0 = outs[-1][0, PIL:]\nS = qhw.uncertain_window(Qn, b0, W=8); Qs = qhw.sub_qubo(Qn, b0, S)\nxq = qhw.solve_qaoa(Qs, reps=2)            # IBM: backend=QiskitRuntimeService().least_busy(operational=True)\nxb, _ = qhw.brute_force(Qs)\nRESULTS['qaoa'] = dict(window=S.tolist(), matches_bruteforce=bool((xq == xb).all()), backend='statevector (yerel)')\nprint('pencere', S.tolist(), '| QAOA = kaba kuvvet optimum:', bool((xq == xb).all()))"
  },
  {
   "cell_type": "markdown",
   "id": "5c15b927",
   "metadata": {
    "id": "5c15b927"
   },
   "source": "## 10b. Şile limanı senaryosu (açık kaynak parametreleri)\n\n`oceanml/sile.py`: EMODnet batimetrisi (liman içi 1,5–5 m, ağız 5–9 m, ada çukuru 9–21 m), Karadeniz tuzluluğu 17–18 psu, mevsimlik ses hızı 1457–1515 m/s, kum/silt dip. Tüm değerler yaklaşık; liman içi ölçüm gerekir."
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "cad6d9e8",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-08T19:25:24.343092Z",
     "iopub.status.busy": "2026-10-08T19:25:24.342753Z",
     "iopub.status.idle": "2026-10-08T19:26:38.545762Z",
     "shell.execute_reply": "2026-10-08T19:26:38.545060Z"
    },
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "cad6d9e8",
    "executionInfo": {
     "status": "ok",
     "timestamp": 1791568742284,
     "user_tz": -180,
     "elapsed": 198666,
     "user": {
      "displayName": "Joseph Volkan ÖZCAN",
      "userId": "11389530967736397480"
     }
    },
    "outputId": "d9795154-ed7b-4bb5-9cb9-6babea21937a"
   },
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "liman_ici    -8 dB  MLP 0.2097  KKA 0.0060  QUBO 0.0028  (outage 12/1/1 /12)\n",
      "liman_ici    +0 dB  MLP 0.1546  KKA 0.0000  QUBO 0.0000  (outage 9/0/0 /12)\n",
      "liman_agzi   -8 dB  MLP 0.2111  KKA 0.0361  QUBO 0.0088  (outage 12/7/3 /12)\n",
      "liman_agzi   +0 dB  MLP 0.1259  KKA 0.0005  QUBO 0.0000  (outage 10/0/0 /12)\n",
      "ada_cukuru   -8 dB  MLP 0.2236  KKA 0.0875  QUBO 0.0028  (outage 12/8/1 /12)\n",
      "ada_cukuru   +0 dB  MLP 0.2079  KKA 0.0440  QUBO 0.0000  (outage 12/3/0 /12)\n"
     ]
    }
   ],
   "source": "from oceanml import sile\nrs = {}\nfor zone in sile.ZONES:\n    rng_s = np.random.default_rng(31337); inst_s = [sile.sample(rng_s, zone=zone) for _ in range(12)]\n    rs[zone] = {}\n    for snr in ([-8, 0] if QUICK else [-12, -8, -4, 0, 8]):\n        rx, bits, lock, info = make_packets(m, inst_s, ['sile']*12, [snr]*12)\n        tb = bits[:, PIL:]\n        e_mlp = (predict(net, m.features_v012(rx, lock).reshape(-1, 400)).numpy().reshape(bits.shape)[:, PIL:] != tb).mean(1)\n        b, lg = kq.run(rx, lock, bits)\n        e_k = (lg['dfe'][:, PIL:] != tb).mean(1); e_q = (b[:, PIL:] != tb).mean(1)\n        rs[zone][snr] = {'v012 MLP': (float(e_mlp.mean()), int((e_mlp>1e-2).sum())), 'KKA + önsel': (float(e_k.mean()), int((e_k>1e-2).sum())), 'Paket QUBO': (float(e_q.mean()), int((e_q>1e-2).sum()))}\n        print(f'{zone:12s} {snr:+d} dB  MLP {e_mlp.mean():.4f}  KKA {e_k.mean():.4f}  QUBO {e_q.mean():.4f}  (outage {(e_mlp>1e-2).sum()}/{(e_k>1e-2).sum()}/{(e_q>1e-2).sum()} /12)')\nRESULTS['sile'] = rs"
  },
  {
   "cell_type": "markdown",
   "id": "16a1c2bc",
   "metadata": {
    "id": "16a1c2bc"
   },
   "source": "## 11. Faz 2: hareket (Doppler) ve zor-kanal sınırı\n\nİki ek deney. **Hareket:** alıcı-verici arasında 0,5–2 m/s bağıl hız, kayıt zaman ekseninde (1+v/c) ile yeniden örneklenerek üretiliyor (katı cisim hareketi; faz sürekli döner). `KKADoppler`, pilot bloğunun eşleşmiş süzgeç tepesini en yükselten Doppler katsayısını arar, ilk 60 karar verilmiş sembolle inceltir ve statik adayla paket artık enerjisine göre yarıştırır. **Zor kanal:** Şile ada çukurunda J=2, 40 pilot ve 4096'lık preamble taraması."
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "d3fd26d7",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-08T19:26:38.547827Z",
     "iopub.status.busy": "2026-10-08T19:26:38.547465Z",
     "iopub.status.idle": "2026-10-08T19:27:17.057033Z",
     "shell.execute_reply": "2026-10-08T19:27:17.055690Z"
    },
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "d3fd26d7",
    "executionInfo": {
     "status": "ok",
     "timestamp": 1791568876730,
     "user_tz": -180,
     "elapsed": 134448,
     "user": {
      "displayName": "Joseph Volkan ÖZCAN",
      "userId": "11389530967736397480"
     }
    },
    "outputId": "39f5ccea-429d-4019-f62b-31b440096040"
   },
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "v_rel 0.0 m/s: statik KKA 0.0000 out 0/12 | Doppler KKA 0.0000 out 0/12 | v tahmin hatası medyan 0.000 m/s\n",
      "v_rel 1.0 m/s: statik KKA 0.5301 out 12/12 | Doppler KKA 0.0347 out 1/12 | v tahmin hatası medyan 0.017 m/s\n"
     ]
    }
   ],
   "source": "from oceanml.sim import run_moving, draw_frontend_pair\nfrom oceanml.receiver import KKADoppler\nfrom oceanml.geometry import compile_geom\nkd, ksq = KKADoppler(m, pilots=PIL), KKASeq(m, pilots=PIL)\nrd = {}\nfor v in ([0.0, 1.0] if QUICK else [0.0, 0.5, 1.0, 2.0]):\n    g = inst[:12]; bits = np.random.default_rng(int(v*10)+1).integers(0, m.M, size=(12, m.nsym))\n    chans = [(*compile_geom(gi, 'tap'), draw_frontend_pair(np.random.default_rng(i))) for i, gi in enumerate(g)]\n    rx, info = run_moving(m, bits, chans, [8.0]*12, [v]*12)\n    lock = m.lock(rx)\n    o, _ = ksq.run(rx, lock, true_bits=bits); e0 = (o[-1][:, PIL:] != bits[:, PIL:]).mean(1)\n    o, _ = kd.run(rx, true_bits=bits); e1 = (o[-1][:, PIL:] != bits[:, PIL:]).mean(1)\n    rd[v] = (float(e0.mean()), int((e0 > 1e-2).sum()), float(e1.mean()), int((e1 > 1e-2).sum()))\n    print(f'v_rel {v} m/s: statik KKA {e0.mean():.4f} out {(e0>1e-2).sum()}/12 | Doppler KKA {e1.mean():.4f} out {(e1>1e-2).sum()}/12 | v tahmin hatası medyan {np.median(np.abs(kd.last_v - v)):.3f} m/s')\nRESULTS['doppler'] = rd"
  },
  {
   "cell_type": "markdown",
   "id": "6a746bfb",
   "metadata": {
    "id": "6a746bfb"
   },
   "source": "## 11b. Kuantum avantaj benchmark'ı — zorluğu artırılmış gerçek alıcı QUBO'ları (Exp 13)\n\nSoru: sentetik veriyi **derinleştirip (50–200 m), yoğunlaştırıp (koruma aralığı 200→0, veri hızı ×2), büyütüp (80→1580 değişken), gürültüyü artırıp (−18 dB) ve M=4 sembole geçip** klasik çözücülerin yetişemediği bir rejim yaratabilir miyiz? Yaratabilirsek kuantum avantajı orada aranır.\n\nÖlçüt Rønnow vd. (Science 2014) usulü **TTS99**: optimum enerjiye %99 olasılıkla ulaşma süresi, her çözücünün kendi en iyi bütçesinde. Klasik taraf: SA, tabu, simüle çatallanma (SB) ve M=4 için sembol uzayında Potts-SA. Kuantum tarafı **aynı örnek dosyalarıyla** (`runs/qbench/*.npz`) ve **aynı referans enerjilerle** `qbench_qpu.py` ile koşulur. Karar ölçütleri QPU koşusundan önce `qbench_verdict.py` içinde sabitlendi.\n\nTam tarama 2 çekirdekte ~1,5 saat sürdü; sonuçlar pakette (`runs/exp13_qbench.json`). `RERUN=True` ile Colab'da yeniden koşabilirsiniz."
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "bea2b5d4",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-08T19:27:17.059586Z",
     "iopub.status.busy": "2026-10-08T19:27:17.059284Z",
     "iopub.status.idle": "2026-10-08T19:27:21.709881Z",
     "shell.execute_reply": "2026-10-08T19:27:21.708863Z"
    },
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "bea2b5d4",
    "executionInfo": {
     "status": "ok",
     "timestamp": 1791568889330,
     "user_tz": -180,
     "elapsed": 12601,
     "user": {
      "displayName": "Joseph Volkan ÖZCAN",
      "userId": "11389530967736397480"
     }
    },
    "outputId": "810aa096-401e-4808-c3a8-50bc1ca24bb3"
   },
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "size_M2_n100_g200_geo_snr-12               n=   80 deg=   67 best=     0.17 ms (sa) max=     0.67  easy  sa=0.17 tabu=6.04 sb=4.59\n",
      "size_M2_n200_g200_geo_snr-12               n=  180 deg=  102 best=     0.52 ms (sa) max=     2.98  easy  sa=0.52 tabu=6.06 sb=8.03\n",
      "size_M2_n400_g200_geo_snr-12               n=  380 deg=  107 best=     1.47 ms (sa) max=     4.61  easy  sa=1.47 tabu=7.69 sb=36.35\n",
      "size_M2_n800_g200_geo_snr-12               n=  780 deg=  108 best=     5.81 ms (sa) max=    17.94  easy  sa=5.81 tabu=19.34 sb=168.64\n",
      "size_M2_n1600_g200_geo_snr-12              n= 1580 deg=   90 best=    10.95 ms (sa) max=    48.98  easy  sa=10.95 tabu=62.34 sb=inf\n",
      "density_M2_n400_g100_geo_snr-12            n=  380 deg=  141 best=     4.41 ms (sa) max=     7.16  easy  sa=4.41 tabu=7.23 sb=1279.09\n",
      "density_M2_n400_g0_geo_snr-12              n=  380 deg=  210 best=    10.22 ms (sa) max=    45.92  easy  sa=10.22 tabu=28.01 sb=inf\n",
      "density_M2_n400_g200_sile_snr-12           n=  380 deg=  114 best=     5.34 ms (sa) max=    25.73  easy  sa=6.49 tabu=8.55 sb=59.74\n",
      "density_M2_n400_g0_sile_snr-12             n=  380 deg=  231 best=    14.05 ms (tabu) max=    33.95  easy  sa=19.59 tabu=22.50 sb=inf\n",
      "density_M2_n400_g200_deep_snr-12           n=  380 deg=  112 best=     3.23 ms (sa) max=   172.07  easy  sa=3.23 tabu=17.85 sb=inf\n",
      "density_M2_n400_g0_deep_snr-12             n=  380 deg=  206 best=    11.91 ms (sa) max=    34.90  easy  sa=18.40 tabu=14.57 sb=inf\n",
      "noise_M2_n400_g200_geo_snr-6               n=  380 deg=   51 best=     0.57 ms (sa) max=     0.70  easy  sa=0.57 tabu=7.16 sb=0.72\n",
      "noise_M2_n400_g200_geo_snr-18              n=  380 deg=  137 best=     2.08 ms (sa) max=    23.73  easy  sa=2.08 tabu=7.19 sb=inf\n",
      "mary_M4_n50_g200_geo_snr-12                n=  120 deg=  119 best=     0.36 ms (potts) max=     1.42  easy  sa=inf tabu=6165.33 sb=inf potts=0.36\n",
      "mary_M4_n100_g200_geo_snr-12               n=  320 deg=  319 best=    10.34 ms (potts) max=    30.49  easy  sa=inf tabu=1307.18 sb=inf potts=10.34\n",
      "mary_M4_n200_g200_geo_snr-12               n=  720 deg=  521 best=   344.83 ms (potts) max=   687.10  HARD  sa=inf tabu=inf sb=inf potts=344.83\n",
      "extreme_M2_n800_g0_deep_snr-18             n=  780 deg=  264 best=    32.85 ms (sa) max=   155.99  easy  sa=55.21 tabu=71.72 sb=inf\n",
      "extreme_M2_n1600_g0_deep_snr-18            n= 1580 deg=  284 best=   193.09 ms (sa) max=  1119.94  HARD  sa=692.11 tabu=638.17 sb=inf\n",
      "extreme_M4_n400_g0_deep_snr-18             n= 1520 deg= 1061 best=      inf ms (potts) max=      inf  HARD  sa=inf tabu=inf sb=inf potts=inf\n",
      "extreme_M2_n400_g0_deep_snr-18             n=  380 deg=  251 best=    22.88 ms (tabu) max=   294.70  easy  sa=39.76 tabu=22.88 sb=inf\n",
      "extreme_M2_n1200_g0_deep_snr-18            n= 1180 deg=  280 best=    78.83 ms (tabu) max= 22737.17  easy  sa=269.83 tabu=81.93 sb=inf\n",
      "slopes (log10 TTS per variable): {'classical_slope': 0.0011, 'slope_sa': 0.0011, 'slope_tabu': 0.00071, 'slope_sb': inf}\n",
      "best-classical TTS ~ n^1.45 (power-law fit, size axis)\n",
      "V0 hard tiers: ['mary_M4_n200_g200_geo_snr-12', 'extreme_M2_n1600_g0_deep_snr-18', 'extreme_M4_n400_g0_deep_snr-18']\n",
      "V1: None\n",
      "V2: None\n",
      "\n"
     ]
    }
   ],
   "source": "import subprocess, json\nRERUN = True\nif RERUN or not os.path.exists('runs/exp13_qbench.json'):\n    subprocess.run([sys.executable, 'exp13_qbench.py', '2', '2' if QUICK else '4', 'size_M2_n' if QUICK else ''], check=True)\nout = subprocess.run([sys.executable, 'qbench_verdict.py'], capture_output=True, text=True); print(out.stdout[-4000:])\nRESULTS['qbench'] = json.load(open('runs/exp13_verdict.json'))"
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "87dd3489",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-08T19:27:21.712154Z",
     "iopub.status.busy": "2026-10-08T19:27:21.711890Z",
     "iopub.status.idle": "2026-10-08T19:27:21.717980Z",
     "shell.execute_reply": "2026-10-08T19:27:21.717371Z"
    },
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "87dd3489",
    "executionInfo": {
     "status": "ok",
     "timestamp": 1791568889331,
     "user_tz": -180,
     "elapsed": 6,
     "user": {
      "displayName": "Joseph Volkan ÖZCAN",
      "userId": "11389530967736397480"
     }
    },
    "outputId": "1947306a-359a-4f4f-c67b-7b1c62a14dcd"
   },
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "DWAVE_API_TOKEN yok: QPU sütunu boş kalır; klasik taraf ve ön-kayıtlı ölçütler raporda.\n"
     ]
    }
   ],
   "source": "# --- Aynı örnekler D-Wave QPU'da (kendi hesabınızla). Tahmini QPU süresi: kademe başına ~4 örnek x 5 tavlama süresi x 1000 okuma ---\nif os.environ.get('DWAVE_API_TOKEN'):\n    subprocess.run([sys.executable, 'qbench_qpu.py', 'dwave', 'size_M2_n'], check=True)     # önce boyut ekseni (V2 ölçekleme)\n    subprocess.run([sys.executable, 'qbench_qpu.py', 'dwave', 'density_M2_n400_g0'], check=True)\n    out = subprocess.run([sys.executable, 'qbench_verdict.py'], capture_output=True, text=True); print(out.stdout[-4000:])\n    RESULTS['qbench'] = json.load(open('runs/exp13_verdict.json'))\nelse:\n    print('DWAVE_API_TOKEN yok: QPU sütunu boş kalır; klasik taraf ve ön-kayıtlı ölçütler raporda.')"
  },
  {
   "cell_type": "markdown",
   "id": "636a728b",
   "metadata": {
    "id": "636a728b"
   },
   "source": "## 12. Anlatımlı sonuç raporu\n\nBu hücre, yukarıdaki koşunun bütün çıktılarını toplar ve teknik uzmanlara yönelik Türkçe bir rapor üretir: ne ölçüldü, ne anlama geliyor, ne başarıldı, ne başarılamadı. Her sayı hücrelerden okunur; cümleler sonuca göre değişir (PASS/FAIL, kazanç var/yok). Rapor `runs/rapor_<tarih>.md` ve `.html` olarak kaydedilir; Drive bağlıysa oraya da kopyalanır."
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "23aad26e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-08T19:27:21.720373Z",
     "iopub.status.busy": "2026-10-08T19:27:21.720131Z",
     "iopub.status.idle": "2026-10-08T19:27:21.805317Z",
     "shell.execute_reply": "2026-10-08T19:27:21.803970Z"
    },
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "id": "23aad26e",
    "executionInfo": {
     "status": "ok",
     "timestamp": 1791568889412,
     "user_tz": -180,
     "elapsed": 83,
     "user": {
      "displayName": "Joseph Volkan ÖZCAN",
      "userId": "11389530967736397480"
     }
    },
    "outputId": "8e446ed6-132d-4523-d6a8-9579b40120f7"
   },
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "rapor Drive klasorune kopyalandi: MyDrive/rapor_20261009_1801.md\n"
     ]
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ],
      "text/markdown": "# OceanML koşu raporu — 09.10.2026\n\nOrtam: GERÇEK sözlük; hızlı mod (QUICK=True, küçük örneklem). Kanal modeli OceanChannel v1.1 §4 görüntü yöntemi.\n\n## 1. Özet\n\nEğitimde hiç görülmemiş 24 kurulumda (liman + tap kanalı, 5 SNR noktası) referans v012 MLP 4/24 outage verdi; kanal-koşullu alıcı (KKA) 0/24. Ortalama BER 0,0363 → 0,0046. KKA, koşudan önce sabitlenen K1'–K3' ölçütlerinin üçünü de geçti.\n\nNe başardık: alıcı geometriyi ezberlemek yerine her paketin kanalını ölçüyor. Kuantum yolu düşük SNR'de kararı iyileştiriyor. Kuantum donanım henüz koşulmadı; kanal statik.\n\n## 2. Teşhis: 30 cm bölge neden yetmiyor\n\n- matlab modunda yankı korelasyonu 0,5'in altına düştüğü kayma: r 3,0 cm, zs 1,0 cm, zr 1,0 cm, D 0,8 cm. 44 kHz'de dalga boyu 3,4 cm; bir yankının yol farkı λ kadar değişince fazı tam tur dönüyor.\n- physics modunda yankı korelasyonu 0,5'in altına düştüğü kayma: r test aralığında (10 cm) düşmedi, zs 1,5 cm, zr 1,5 cm, D test aralığında (10 cm) düşmedi. 44 kHz'de dalga boyu 3,4 cm; bir yankının yol farkı λ kadar değişince fazı tam tur dönüyor.\n- matlab: yankı yayılımı medyan 14,7 ms; alınan enerjinin %36'i MLP'nin 2 sembollük (0,8 ms) penceresinin dışındaki sembollerden geliyor.\n- physics: yankı yayılımı medyan 4,2 ms; alınan enerjinin %8'i MLP'nin 2 sembollük (0,8 ms) penceresinin dışındaki sembollerden geliyor.\n\nYorum: sorun veri eksikliği değil. 30 cm'lik bir hücrenin içinde kanal defalarca değişiyor ve MLP yankıların çoğunu hiç görmüyor. Daha fazla geometri eğitmek bu iki engeli kaldırmaz.\n\n## 3. Yöntem\n\n1. **Kanal tahmini.** Preamble + 20 pilot sembol (8 ms) ile nedensel maskeli en küçük kareler (eşlenik gradyan); her 40 sembolde kararlarla yeniden tahmin; paket sonunda tam kayıtla son tur.\n2. **Öğrenen kanal önseli.** 40 kHz temel bantta çalışan küçük bir CNN; rastgele sürekli geometrilerle eğitildi, geometri değil kanalın ortak yapısını öğreniyor. Düşük SNR'de tahmini temizliyor.\n3. **Karar.** (a) Gecikmeli-karar DFE, 2 sembollük pencere, 4 hipotez; (b) paket-seviyesi QUBO: paketin tüm bitlerine birlikte karar, kuantum tavlayıcı girdisi; çözücüler yerel iniş / SA / simulated bifurcation / D-Wave / QAOA.\n\nReferans MLP (400→32→1, BCE, Adam 2e-3, 8 epoch) sürekli geometriyle eğitildi; eğitim hatası 0,0783. \n\n## 4. Sonuçlar: görülmemiş geometriler\n\n| Alıcı | K1' ort. BER ≤ 0,025 | K2' kurulum medyanı ≤ 0,005 | K3' outage ≤ %15 |\n|---|---|---|---|\n| v012 MLP | 0,0363 — FAIL | 0,0000 — PASS | 4/24 — FAIL |\n| KKA pilot | 0,0276 — FAIL | 0,0000 — PASS | 0/24 — PASS |\n| KKA pilot + önsel | 0,0046 — PASS | 0,0000 — PASS | 0/24 — PASS |\n| KKA pilotsuz | 0,0787 — FAIL | 0,0056 — FAIL | 11/24 — FAIL |\n| KKA oracle | 0,0018 — PASS | 0,0000 — PASS | 0/24 — PASS |\n| v012 (sizin ağ) | 0,0383 — FAIL | 0,0000 — PASS | 6/24 — FAIL |\n\n- harbor kanalı, SNR -10…14 dB BER eğrileri — v012 MLP: 0,100 0,035 0,021 0,002 0,000; KKA pilot: 0,117 0,007 0,000 0,000 0,000; KKA pilot + önsel: 0,015 0,000 0,000 0,000 0,000; KKA pilotsuz: 0,147 0,078 0,050 0,039 0,034; KKA oracle: 0,011 0,000 0,000 0,000 0,000; v012 (sizin ağ): 0,092 0,035 0,007 0,001 0,001\n- tap kanalı, SNR -10…14 dB BER eğrileri — v012 MLP: 0,118 0,050 0,016 0,008 0,012; KKA pilot: 0,137 0,013 0,000 0,002 0,000; KKA pilot + önsel: 0,031 0,000 0,000 0,000 0,000; KKA pilotsuz: 0,139 0,084 0,061 0,085 0,069; KKA oracle: 0,007 0,000 0,000 0,000 0,000; v012 (sizin ağ): 0,117 0,058 0,014 0,018 0,041\n\nYorum: kanal bilindiğinde (oracle) ortalama BER 0,0018; KKA ile fark 0,0029. Fark küçük: kalan hatalar kanal tahmininden geliyor, karar kuralından değil.\n\n## 5. Öğrenen kanal önseli\n\n| Kanal | SNR | KKA | KKA + önsel |\n|---|---|---|---|\n| harbor | -10 dB | 0,1009 | 0,0157 |\n| harbor | -6 dB | 0,0204 | 0,0005 |\n| harbor | -2 dB | 0,0005 | 0,0000 |\n| tap | -10 dB | 0,1880 | 0,0431 |\n| tap | -6 dB | 0,0657 | 0,0037 |\n| tap | -2 dB | 0,0088 | 0,0000 |\n\nYorum: önsel düşük SNR'de hatayı düşürüyor; 0 dB üstünde kazanç kalmıyor, orada kapatılabilir.\n\n## 6. Kuantum yolu: paket QUBO'su\n\n| Kanal | SNR | DFE + önsel | Paket QUBO | En iyi enerjiye ulaşan çözücüler |\n|---|---|---|---|---|\n| tap | -12 dB | 0,0731 | 0,0185 | dfe %42, greedy %83, sa %100, sb %92 |\n| tap | -8 dB | 0,0148 | 0,0005 | dfe %100, greedy %100, sa %100, sb %92 |\n| harbor | -12 dB | 0,0477 | 0,0181 | dfe %25, greedy %100, sa %100, sb %100 |\n\nYorum: QUBO 3/3 noktada DFE'ye eşit veya daha iyi. Her bitin bütün yankı enerjisini kullanıyor, hata yayılımı yok. Çözücülerin 'en iyi enerjiye ulaşma' oranı düştükçe problem zorlaşıyor; kuantum tavlayıcının yarışacağı rejim bu. Not: bu koşuda kuantum donanım kullanılmadı (SA/SB/yerel iniş klasik veya kuantumdan esinlenmiş); D-Wave/IBM hücreleri kimlik bilgisi gerektirir.\n\n## 7. Kuantum donanım koşusu\n\nQAOA (statevector (yerel)), 8 bitlik en belirsiz pencerede kaba kuvvet optimumunu buldu.\n\nEğitim geometrisi seçimi (QUBO coreset): en kötü durum mesafesi rastgele 0,493, QUBO+SA 0,519, k-merkez 0,364 — klasik k-merkez daha iyi; bu alt problemde kuantum avantajı yok.\n\n## 8. Kuantum avantaj benchmark'ı: zorlaştırılmış gerçek alıcı QUBO'ları\n\nHer satır, gerçek sözlük + kanal + gürültü + pilot/önsel kanal kestirimiyle alıcının çözücüye verdiği QUBO örnekleri. TTS99 = optimum enerjiye (E_ref) %99 olasılıkla ulaşma süresi, her çözücünün en iyi bütçesinde, tek CPU çekirdeği. Gerçek zaman bütçesi bir paket süresi: 83 ms.\n\n| Kademe | Değişken | Ort. bağlantı | En iyi klasik TTS99 (ms) | Kazanan | SA | Tabu | SB | Potts-SA | Zor? | QPU TTS (ms) |\n|---|---|---|---|---|---|---|---|---|---|---|\n| size_M2_n100_g200_geo_snr-12 | 80 | 67 | 0.17 | sa | 0.17 | 6.04 | 4.59 | — | hayır | — |\n| size_M2_n200_g200_geo_snr-12 | 180 | 102 | 0.52 | sa | 0.52 | 6.06 | 8.03 | — | hayır | — |\n| size_M2_n400_g200_geo_snr-12 | 380 | 107 | 1.47 | sa | 1.47 | 7.69 | 36.35 | — | hayır | — |\n| size_M2_n800_g200_geo_snr-12 | 780 | 108 | 5.81 | sa | 5.81 | 19.34 | 168.64 | — | hayır | — |\n| size_M2_n1600_g200_geo_snr-12 | 1580 | 90 | 10.95 | sa | 10.95 | 62.34 | ∞ | — | hayır | — |\n| density_M2_n400_g100_geo_snr-12 | 380 | 141 | 4.41 | sa | 4.41 | 7.23 | 1279.09 | — | hayır | — |\n| density_M2_n400_g0_geo_snr-12 | 380 | 210 | 10.22 | sa | 10.22 | 28.01 | ∞ | — | hayır | — |\n| density_M2_n400_g200_sile_snr-12 | 380 | 114 | 5.34 | sa | 6.49 | 8.55 | 59.74 | — | hayır | — |\n| density_M2_n400_g0_sile_snr-12 | 380 | 231 | 14.05 | tabu | 19.59 | 22.50 | ∞ | — | hayır | — |\n| density_M2_n400_g200_deep_snr-12 | 380 | 112 | 3.23 | sa | 3.23 | 17.85 | ∞ | — | hayır | — |\n| density_M2_n400_g0_deep_snr-12 | 380 | 206 | 11.91 | sa | 18.40 | 14.57 | ∞ | — | hayır | — |\n| noise_M2_n400_g200_geo_snr-6 | 380 | 51 | 0.57 | sa | 0.57 | 7.16 | 0.72 | — | hayır | — |\n| noise_M2_n400_g200_geo_snr-18 | 380 | 137 | 2.08 | sa | 2.08 | 7.19 | ∞ | — | hayır | — |\n| mary_M4_n50_g200_geo_snr-12 | 120 | 119 | 0.36 | potts | ∞ | 6165.33 | ∞ | 0.36 | hayır | — |\n| mary_M4_n100_g200_geo_snr-12 | 320 | 319 | 10.34 | potts | ∞ | 1307.18 | ∞ | 10.34 | hayır | — |\n| mary_M4_n200_g200_geo_snr-12 | 720 | 521 | 344.83 | potts | ∞ | ∞ | ∞ | 344.83 | EVET | — |\n| extreme_M2_n800_g0_deep_snr-18 | 780 | 264 | 32.85 | sa | 55.21 | 71.72 | ∞ | — | hayır | — |\n| extreme_M2_n1600_g0_deep_snr-18 | 1580 | 284 | 193.09 | sa | 692.11 | 638.17 | ∞ | — | EVET | — |\n| extreme_M4_n400_g0_deep_snr-18 | 1520 | 1061 | ∞ | hiçbiri | ∞ | ∞ | ∞ | ∞ | EVET | — |\n| extreme_M2_n400_g0_deep_snr-18 | 380 | 251 | 22.88 | tabu | 39.76 | 22.88 | ∞ | — | hayır | — |\n| extreme_M2_n1200_g0_deep_snr-18 | 1180 | 280 | 78.83 | tabu | 269.83 | 81.93 | ∞ | — | hayır | — |\n\nÖlçekleme: en iyi klasik TTS99 ≈ n^1.45 (boyut ekseni), n^1.43 (en zor ayarlar).\nÖn-kayıtlı karar: V0 zor kademe = mary_M4_n200_g200_geo_snr-12, extreme_M2_n1600_g0_deep_snr-18, extreme_M4_n400_g0_deep_snr-18; V1 (mutlak, QPU) = ölçülmedi; V2 (ölçekleme, QPU) = ölçülmedi.\n\nBoyut 80→1580 değişkene çıkınca en iyi klasik çözücünün TTS99'u yaklaşık n^1.45 büyüdü: polinom, üstel değil. İkili (M=2) kademelerin en zoru extreme_M2_n1600_g0_deep_snr-18: medyan 193.1 ms, en kötü örnek 1119.9 ms; 83 ms paket bütçesinin ÜSTÜNDE. Her ayar en zor konumdayken (derin 50–200 m, koruma 0, −18 dB) ölçekleme n^1.43: yine polinom. M=4 tek-sıcak (one-hot) QUBO'da genel QUBO çözücüleri (SA, SB) cezalı enerji yüzeyinde takılıyor, hiçbir örnekte optimuma ulaşamadılar. Aynı problemi doğrudan sembol uzayında tavlayan klasik Potts-SA: 120 değişkende 0.4 ms, 320 değişkende 10.3 ms, 720 değişkende 344.8 ms, 1520 değişkende çözemedi. Küçük boyutta zorluk kodlamadan geliyor; 720+ değişkende Potts-SA da yavaşlıyor (ayarlanmamış çizelge; paralel tavlama ile yeniden denenmeli). Neden: alıcı QUBO'su bantlı (bağlaşım yalnızca yankı süresi içindeki sembollerle) ve köşegeni güçlü (her bitin eşleşmiş süzgeç kanıtı); enerji yüzeyi neredeyse tek-vadili, spin-camı değil. Derinlik, yoğunluk ve gürültü problemi büyütüyor ama camsılaştırmıyor. Kritik bulgu: klasik açıdan zorlaşan kademelerde, optimum çözüm ile 83 ms bütçesine sığan ucuz klasik çözüm neredeyse aynı hata oranını veriyor (sembol hata oranı, optimum / ucuz: 0.024 / 0.025; 0.258 / 0.258; 0.411 / 0.411). −18 dB kademelerinde bağlantı zaten kopuk (%26–41 hata); M=4, −12 dB kademesinde kazanç binde 1,4. Optimuma daha hızlı ulaşan herhangi bir çözücü (kuantum dahil) hata oranını anlamlı değiştiremez; sınırı gürültü ve kanal koyuyor, hesaplama değil. Ayrıca zor kademelerin bant genişliği 170–687 değişken ve boyutları 720–1580: bugünkü D-Wave QPU'larına doğrudan gömülemeyecek kadar yoğun; bunlar ancak Leap hybrid ile koşar, o da saf kuantum iddiasına izin vermez. Doğrudan QPU'ya sığan kademeler (≤~380 değişken, bant 85) ise klasik olarak 0,5–5 ms'de çözülüyor. Ön-kayıtlı V0'a göre zor kademeler: mary_M4_n200_g200_geo_snr-12, extreme_M2_n1600_g0_deep_snr-18, extreme_M4_n400_g0_deep_snr-18. Hesaplama açısından kuantumun yarışabileceği rejim var; ancak yukarıdaki iki engel (hata oranı değişmiyor, QPU'ya sığmıyor) yüzünden bu alıcıda iletişim kazancına dönüşmesi beklenmiyor.\n\n## 9. Şile limanı senaryosu (açık kaynak parametreleri)\n\n| Bölge | SNR | v012 MLP | KKA + önsel | Paket QUBO |\n|---|---|---|---|---|\n| liman_ici | -8 dB | 0,2097 (12) | 0,0060 (1) | 0,0028 (1) |\n| liman_ici | +0 dB | 0,1546 (9) | 0,0000 (0) | 0,0000 (0) |\n| liman_agzi | -8 dB | 0,2111 (12) | 0,0361 (7) | 0,0088 (3) |\n| liman_agzi | +0 dB | 0,1259 (10) | 0,0005 (0) | 0,0000 (0) |\n| ada_cukuru | -8 dB | 0,2236 (12) | 0,0875 (8) | 0,0028 (1) |\n| ada_cukuru | +0 dB | 0,2079 (12) | 0,0440 (3) | 0,0000 (0) |\n\nYorum: değerler BER (parantezde outage). Şile kanalları eğitim kutusunun dışında; MLP çöküyor, KKA liman içi/ağzında çalışıyor, uzun yankılı ada çukuru hem KKA'nın sınırını hem QUBO'nun kazancını gösteriyor. Parametreler yaklaşık; liman içi ölçüm gerekli.\n\n## 10. Faz 2: hareketli kanal (Doppler)\n\n| |v_rel| | Statik KKA BER (outage) | Doppler KKA BER (outage) |\n|---|---|---|\n| 0.0 m/s | 0,0000 (0) | 0,0000 (0) |\n| 1.0 m/s | 0,5301 (12) | 0,0347 (1) |\n\nYorum: hareket varken Doppler telafisi 1/1 hız noktasında outage'ı düşürdü; hareketsizken statik alıcıyla aynı sonucu verdi (zarar yok). Kalan outage'lar hız kestiriminin ~0,05 m/s hassasiyetinin yetmediği uzun yankılı kanallar; çözüm paket içi sürekli faz izleme (bir sonraki geliştirme).\n\n## 11. Sınırlar ve dürüst notlar\n\n- Hareket, katı cisim Doppler'ı olarak modellendi (tüm yollar aynı hızda); yüzey dalgası ve yol-başına farklı Doppler yok.\n- 20 pilot sembol çerçeveyi %10 uzatıyor; pilotsuz mod bazı kanallarda yanlış sabit noktaya oturabiliyor.\n- Kuantum donanım bu koşuda kullanılmadı; 'kuantum SA'yı geçer' iddiası henüz ölçülmedi.\n- Hızlı modda örneklem küçük; 'BER 0' = o kadar bitte hata yok demek, alt sınır değil.\n\n## 12. Öneriler\n\n1. Gerçek sözlük, GEO_TABLE ve net_v012_table ile QUICK=False koşusu; bu raporu yeniden üretin.\n2. Pilot kararı (20 pilot mu, pilotsuz mu) ve findPreambleLock v2'nin ilk varışı seçtiğinin doğrulanması.\n3. Paket QUBO'sunu D-Wave (düz + reverse annealing) ve IBM QAOA'da SA ile aynı paketlerde kıyaslayın; SB'yi FPGA'ya taşıyın.\n4. Şile'de ölçüm: liman içi batimetri, dip örneği, yaz/kış CTD, 10–50 kHz gürültü.\n5. Faz 2 devamı: paket içi sürekli faz izleme (PLL / Kalman) ile Doppler artığını kapatın; yol-başına Doppler ve yüzey hareketi ekleyin.\n6. Uzun menzilli sığ kurulumlarda yüzey yansıması doğrudan yolu söndürüyor (Lloyd aynası); frekans veya alıcı-derinliği çeşitliliği değerlendirin.\n"
     },
     "metadata": {}
    }
   ],
   "source": "from oceanml.report import build\n\nfrom IPython.display import Markdown, display\nrapor = build(RESULTS, glossary_real=(words is not None), quick=QUICK)\nos.makedirs('runs', exist_ok=True)\nstamp = time.strftime('%Y%m%d_%H%M')\nopen(f'runs/rapor_{stamp}.md', 'w').write(rapor)\ntry:\n    import markdown\n    html = \"<html><head><meta charset='utf-8'><style>body{font-family:sans-serif;max-width:900px;margin:2em auto;line-height:1.5}table{border-collapse:collapse}td,th{border:1px solid #999;padding:4px 8px}</style></head><body>\" + markdown.markdown(rapor, extensions=['tables']) + \"</body></html>\"\n    open(f'runs/rapor_{stamp}.html', 'w').write(html)\nexcept ImportError:\n    pass\nif IN_COLAB and os.path.isdir('/content/drive/MyDrive'):\n    import shutil; shutil.copy(f'runs/rapor_{stamp}.md', '/content/drive/MyDrive/')\n    print(\"rapor Drive klasorune kopyalandi: MyDrive/rapor_%s.md\" % stamp)\ndisplay(Markdown(rapor))"
  },
  {
   "cell_type": "markdown",
   "id": "349604e3",
   "metadata": {
    "id": "349604e3"
   },
   "source": "## 13. Sonraki adımlar\n\n1. **Gerçek sözlük ve kanallarla yeniden koşu.** §3'teki yollar doldurulup 6–8 tekrarlanır. Özellikle 5 ve 9 numaralı geometri-sönümlü kurulumlarda KKA'nın davranışı görülmelidir.\n2. **Pilot kararı.** 20 pilot sembol çerçeveyi %10 uzatır. Alternatif olarak, pilotsuz yinelemeli KKA'nın tuzağa düştüğü kanallar için daha uzun bir preamble veya preamble sonrası sessiz aralık değerlendirilebilir.\n3. **Donanım eşlemesi.** Mevcut Zynq-7020 %98 dolu. KKA'nın kanal tahmini PS'de (ARM, paket başına) yapılabilir. DFE + hipotez karşılaştırması ise seyrek tap'larla PL'de, sembol başına J+1 pencerede yürütülür. Kaynak tahmini ayrı bir çalışma gerektirir.\n4. **OceanChannel donanım kanıtı.** K-em4 (donanım BER = MATLAB BER) için aynı kurulumlar `.och.json` ile emülatöre yüklenir ve KKA modemle karşılaştırılır."
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.13.16"
  },
  "colab": {
   "provenance": []
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}