Changes)
-
Copy Collected Data in a csv file
-
Remove SSID, MAC Address from csv file
-
Train model using ann_train.py
-
Train model using ann_train.py
-
Train model using ann_train.py
-
Train model using ann_train.py
-
Train model using ann_train.py
(Changes to be made Before Training
X = dataset.iloc[:,0:4].values Y = dataset.iloc[:,4].values
Change value of 4 to number of aps
classifier.add(tf.keras.layers.Dense(units = 25, kernel_initializer= 'uniform', activation = 'relu', input_dim = 4))
classifier.add(tf.keras.layers.Dense(units = 25, kernel_initializer= 'uniform', activation = 'relu'))
classifier.add(tf.keras.layers.Dense(units = 5, kernel_initializer= 'uniform', activation = 'softmax'))
Change input_dim to number of aps
Add more layers if necessary
Change units in last layer to aps + 1
)
-
Model and Transformation will be exported
-
Train model using autoencoder_train.py
(Changes to be made before training
nb_aps = 4
Change nb_aps to number of aps
self.fc1 = nn.Linear(nb_aps, 2)
self.fc2
= nn.Linear(2, 1)
self.fc3
= nn.Linear(1, 2)
self.fc4
= nn.Linear(2, nb_aps)
Here we can more layers and vary number of nodes
In this case, nb_aps/2 i.e, 2 and nb_aps/4 i.e, 1 has been considered
mean_corrector = nb_aps/float(4.0 + 1e-10)
Change 4.0 to number of aps with decimal point
)
-
Model will be exported
-
Create a Dictionary file which will store MAC addresses and its index according to csv file and store it as 'mac_dictions'
For eg.
Key Type Value
1c:5f:2b:da:78:ec str 0
1e:4d:70:af:f8:9d str 3
1e:96:e6:3d:e2:df str 2
a6:ae:12:0e:37:ff str 1
Deployment
-
Install xampp
-
Create Wideep_Project Folder in htdocs and copy Server Files in it
-
Change Server IP in Android Project
-
Run xampp
-
Run rssi_to_point.py on Server
-
Run App to See Respective Positions