Blame view

MLTD/src/OnlinePrediction.py 9.68 KB
0d8c0f816   Thanasis Naskos   initial commit
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
  # -*- coding: utf-8 -*-
  import time, json
  
  import dateutil
  import numpy as np
  import pandas as pd
  
  import Utils
  
  try:
      import cPickle as pickle
  except:
      import pickle
  import PredictionKorvesis as pdm
  import paho.mqtt.client as paho
  import ReportTimeDB
  import logging
  import logging.config
  import yaml
  import os
  import urllib3
  
  urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning)
  
  MYDIR = os.path.dirname(os.path.realpath(__file__))
  LOGGING_CONF_FILE = os.path.join(MYDIR, "logging.yml")
  
  
  class OnlinePrediction:
      def __init__(self):
          self.data_dates = []
          self.data_values = []
          self.read_log_conf(LOGGING_CONF_FILE)
          self.logger = logging.getLogger("mltd-online")
          self.logger.info("Online MLTD is running")
  
      def read_log_conf(self, yaml_file):
          with open(yaml_file) as f:
              logging.config.dictConfig(yaml.safe_load(f))
  
      def on_connect(self, client, userdata, flags, rc):
          if rc == 0:
              self.logger.debug("Connected to broker")
  
              self.Connected = True  # Signal connection
          else:
              self.logger.error("Connection failed")
  
      def on_message(self, client, userdata, message):
          """
          {
              "asset_id": "string",
              "timestamp": "2020-02-27T13:40:18.224Z",
              "event_alarm": [
                  {
                      "event_alarm_id": "string",
                      "event_alarm_char": "string",
                      "name": "string",
                      "source_ip": "string",
                      "source_port": 0,
                      "destination_ip": "string",
                      "destination_port": 0,
                      "priority": 0,
                      "confidence": 0,
                  }
              ],
          }
          :param client:
          :param userdata:
          :param message:
          :return:
          """
          self.logger.debug(
              "Event received: " + str(json.loads(message.payload.decode("UTF-8")))
          )
          json.loads(message.payload)
          data_dates = []
          data_values = []
          measDict = json.loads(message.payload.decode("UTF-8"))
  
          if measDict["asset_id"] == self.asset_id:
              for event in range(len(measDict["event_alarm"])):
                  data_dates.append(measDict["timestamp"])
                  # datetime.datetime.fromtimestamp(measDict["timestamp"]).strftime(
                  #     "%Y-%m-%dT%H:%M:%SZ"
                  # )
                  # )
                  event_alarm_id = measDict["event_alarm"][event]["event_alarm_id"]
                  data_values.append(event_alarm_id)
  
          self.do_the_monitoring(data_dates, data_values)
  
      def do_the_monitoring(self, data_dates=[], data_values=[]):
          predictions = []
  
          data_values = self.data_values + data_values
          data_dates = self.data_dates + data_dates
  
          if len(data_dates) > 0:
              first_date_str = data_dates[0]
              last_date_str = data_dates[-1]
              first_date = dateutil.parser.parse(first_date_str)
              last_date = dateutil.parser.parse(last_date_str)
              duration = (last_date - first_date).total_seconds()
              if duration >= self.ts_seconds:
                  self.data_values = []
                  self.data_dates = []
                  self.logger.info(f"Events Received: {len(data_values)}"
                                   f" - Duration: {round(duration,2)} secs")
                  self.logger.info("Prediction triggered")
                  predictions = self.predict(data_dates, data_values)
              else:
                  self.data_values = data_values
                  self.data_dates = data_dates
  
          if max(predictions) > self.sigmoid_threshold:
              time_db_client = ReportTimeDB.connect(
                  self.time_db_host,
                  self.time_db_port,
                  self.time_db_database,
                  self.time_db_username,
                  self.time_db_password,
                  self.time_db_ssl,
              )
              timeframe = Utils.sigmoid_mins(
                  max(predictions),
                  self.rf_s,
                  Utils.convert_hours_to_mins(Utils.strtime_to_hours(self.rf_midpoint)),
                  self.hours_before,
              )
              self.logger.info(f"A prominent security incident is predicted"
                               f" - Risk level: {round(max(predictions),2)}"
                               f" - Expected timeframe: {round(timeframe,2)} secs")
              ReportTimeDB.report(
                  time_db_client, self.asset_id, max(predictions) * 100, timeframe
              )
              self.logger.info(f"The incident was reportered on TimescaleDB")
          else:
              self.logger.info(
                  f"The predicted risk {round(max(predictions),2)} is "
                  f"below the alarm threshold {round(self.sigmoid_threshold,2)}"
              )
  
      def predict(self, data_dates=[], data_values=[]):
          dataset = pd.DataFrame({"Timestamps": data_dates, "Event_id": data_values})
          predictions = pdm.predict(
              self.regr, dataset, self.time_segments, self.feature_importance.index
          )
          self.logger.debug(f"Risk predictions: {predictions}")
          return predictions
  
      def form_dataset(self, dates_list, events_list, feature_importance):
          # Create a Pandas dataframe with all the non zero event ids
          # TODO handle differently the zero event ids based on some policy
          loc = 0
          dataset = pd.DataFrame(columns=["Timestamps", "Event_id"])
  
          if len(events_list) != abs(sum(events_list)):
              for i in range(len(events_list)):
                  if i < len(dates_list) and feature_importance.index.contains(
                      events_list[i]
                  ):
                      dataset.loc[loc] = pd.Series(
                          {"Timestamps": dates_list[i], "Event_id": events_list[i]}
                      )
                      loc += 1
  
          if not dataset.empty:
              # dropping ALL duplicate values
              dataset.drop_duplicates(subset="Timestamps", keep="first", inplace=True)
              dataset.set_index(
                  pd.to_datetime(dataset["Timestamps"]), drop=False, inplace=True
              )
              self.logger.debug(f"Formed dataest: {dataset}")
          return dataset
  
      def load_data(self, filename):
          infile = open(filename, "rb")
          pat_length = pickle.load(infile)
          weak_bins_mapping = pickle.load(infile)
          mp = pickle.load(infile)
          train_dataset_values = np.array(pickle.load(infile))
          regr = pickle.load(infile)
          feature_importance = pickle.load(infile)
          artificial_events_generation = pickle.load(infile)
          infile.close()
          return (
              pat_length,
              weak_bins_mapping,
              mp,
              train_dataset_values,
              regr,
              feature_importance,
              artificial_events_generation,
          )
  
      def start_online_prediction_MQTT(
          self,
          trainID,
          broker_address,
          port,
          mqtt_topic,
          prediction_threshold,
          report_time_db_host,
          report_time_db_port,
          report_time_db_username,
          report_time_db_password,
          report_time_db_database,
          report_time_db_table,
          report_time_db_ssl,
          report_asset_id,
      ):
          self.sigmoid_threshold = prediction_threshold
          self.time_db_host = report_time_db_host
          self.time_db_port = report_time_db_port
          self.time_db_username = report_time_db_username
          self.time_db_password = report_time_db_password
          self.time_db_database = report_time_db_database
          self.time_db_table = report_time_db_table
          self.time_db_ssl = report_time_db_ssl
          self.asset_id = report_asset_id
          sql_conn = Utils.create_sqlite_connection("pdm.sqlite")
          self.time_segments = Utils.select_model_attribute(
              sql_conn, trainID, "time_segments"
          )
          self.rf_s = Utils.select_model_attribute(sql_conn, trainID, "rf_s")
          self.rf_midpoint = Utils.select_model_attribute(
              sql_conn, trainID, "rf_midpoint"
          )
          self.hours_before = Utils.select_model_attribute(
              sql_conn, trainID, "hours_before"
          )
          ts_hours = Utils.strtime_to_hours(self.time_segments)
          self.ts_seconds = ts_hours * 3600
          # load the data from pickle (binary) files - should consider to move to a database solution(?)
          (
              self.pat_length,
              self.weak_bins_mapping,
              self.mp,
              self.dataset_values,
              self.regr,
              self.feature_importance,
              self.artificial_events_generation,
          ) = self.load_data("train_" + str(trainID) + ".dat")
  
          self.Connected = False
          client = paho.Client(
              "Prediction_client" + str(time.time())
          )  # create new instance
          client.on_connect = self.on_connect  # attach function to callback
          client.on_message = self.on_message  # attach function to callback
  
          client.connect(broker_address, port=port)  # connect to broker
  
          client.loop_start()  # start the loop
  
          while self.Connected != True:  # Wait for connection
              time.sleep(0.1)
  
          client.subscribe(mqtt_topic)
  
          try:
              while True:
                  time.sleep(1)
          except KeyboardInterrupt:
              self.logger.debug("exiting")
              client.disconnect()
              client.loop_stop()
  
  
  if __name__ == "__main__":
      op = OnlinePrediction()
      op.start_online_prediction_MQTT(
          10,
          "localhost",
          1884,
          "auth/incidents",
          0.1,
          "83.212.116.5",
          5432,
          "postgres",
          "xs?Z7HsY",
          "kea",
          "mltd",
          False,
          "server",
      )