In this project I am planning to develop an android app which collects data from motion sensors, position sensors, and environment sensors. The idea is to collect all the possible data available from device sensors store them on elastic search with timestamps. The data will be later used to train and predict user behavior by applying machine learning techniques such as random forest using Tensor Flow.
- Collect data from motion sensors, position sensors, and environment sensors
- Store the collected data in Elastic Search
- Apply Machine Learning using Tensor Flow to predict or observe any pattern from the collected data
- Project aims to provide an idea about the information that can be collected from Android Sensors.
- As a user, I can download the app
- As a user, I allow which information from the sensors can be shared with the app
- As a user, I can access Elastic search and see the information which has been sent from my device
- As a user I can login to the website and see the machine learning output from the training set
- As a user I can access only the information sahred by my device.
- As a misuser, I want to access information not shared by my device
- As a misuser, I want to access the machine learning output of training data not shared by my device
- As a misuser, I want to access all the data collected in elastic search without proper authorization
Context Diagram :- Context Diagram.png Container Diagram :- container.pdf Component Diagram:- Component.pdf
This component will collect data from the device. The app will be running in background and sending information to the service every 1 hour.
This component will store all the information collected from the sensor.
This will be a web layout which will allow users to view their information which was collected from their device sensor.
| Component name | Category of vulnerability | Issue Description | Mitigation |
|---|---|---|---|
| Data Collection from Sensor | Data Theft | Application shouldnt allow any third party to collect the data collected from the sensor | App only stores the information and sends it as a batch to ES, stores in local database using Room |
| Elastic search | Data Theft | Elastic Search data is stored in HCC cloud data can be stolen or misused | Elastic Search viewing system is secured and all request will be processed via security token |
| Risk name (value) | Impact | Likelihood | Description | Mitigation |
|---|---|---|---|---|
| Malicious popups being inserted into the app / Injection (56) | 8 | 7 | The ad server selected for this application has not been secured and accepts malicious submissions from ad partner networks. | Ensure that any advertisements are filtered or verified before being allowed access. |
| Information is being stolen from the user/development team (25) | 5 | 5 | The ad server selected for this application has not been secured and allows for user information to be stolen. | Ensure that any advertisements are filtered or verified before being allowed access. |
| Man-in-the-Middle Attack (21) | 7 | 3 | The user or the development team using insecure Wi-Fi while using the app. | Making a note in the instruction sheet to remind users to encrypt their Wi-Fi. Ensure the developer remembers to only use secure Wi-Fi. |
| Bluetooth Sniffing (30) | 6 | 5 | A malicious entity monitors the traffic to and from the android device. | Making a note in the instruction sheet to remind users to turn their Bluetooth connection off when app is not being used. Also to turn the device to a hidden setting if at all possible. |
- You will need the following things properly installed on your computer.
- Git
- Android Emulator
- Android Studio
- Docker or Elastic Search
- Django
git clone <repository-url>this repository Don't use docker as Docker and emulator won't work together.
Download Elastic Search :- https://www.elastic.co/products/elasticsearch
Download Kibana :- https://www.elastic.co/products/kibana
(If you have docker running on different machine, you can skip this process and use docker image for Elastic Search and Kibana)
Once you have downloaded both the softwares, make sure they are running (if not start both of them from command terminal), default ports are 9200 and 5601 for elastic search and Kibana respectively.
If you are running Elastic search on different machine you can use (https://ngrok.com/) and forward the port. ngrok will generate a new url. Next update the url in your source code ( String BASE_URL="http://3e325242.ngrok.io/";) Line number 77 in ESservice with the url obtained from ngrok.
Once you have cloned the app, make sure you have it all setup in the android studio, next launch the app via emulator and soon you can see the logs which prints value from the snsor. Next using the emulator try to change sensor values Temperature, Pressure, Accelerometer and other sensor values. As you change the value you can see the app screen showing the values. Emulator won't show Latitude and Longitude values correctly it will always default to 0.0.