scos-sensor is a work-in-progress reference implementation of the IEEE 802.15.22.3 Spectrum Characterization and Occupancy Sensing (SCOS) sensor developed by NTIA/ITS. scos-sensor defines a RESTful application programming interface (API), that allows authorized users to discover capabilities, schedule actions, and acquire resultant data.

• Introduction
• Glossary
• Architecture
• Overview of scos-sensor Repo Structure
• Quickstart
• Configuration
• Security
• Actions and Hardware Support
• Development
• References
• License
• Contact

scos-sensor was designed by NTIA/ITS with the following goals in mind:

• Easy-to-use sensor control and data retrieval via IP network
• Low-cost, open-source development resources
• Design flexibility to allow developers to evolve sensor technologies and metrics
• Hardware agnostic
• Discoverable sensor capabilities
• Task scheduling using start/stop times, interval, and/or priority
• Standardized metadata/data format that supports cooperative sensing and open data initiatives
• Security controls that prevent unauthorized users from accessing internal sensor functionality
• Easy-to-deploy with provisioned and configured OS
• Quality assurance of software via automated testing prior to release

Sensor control is accomplished through a RESTful API. The API is designed to be rich enough that multiple heterogeneous sensors can be automated effectively while being simple enough to still be useful for single-sensor deployments. For example, by advertising capabilities and location, an owner of multiple sensors can easily filter by frequency range, available actions, or geographic location. Yet, since each sensor hosts its own Browsable API, controlling small deployments is as easy as clicking around a website.

Opening the URL to your sensor (localhost if you followed the Quickstart) in a browser, you will see a frontend to the API that allows you to do anything the JSON API allows. Relationships in the API are represented by URLs which you can click to navigate from endpoint to endpoint. The full API is discoverable simply by following these links:

Scheduling an action is as simple as filling out a short form on /schedule:

Actions that have been scheduled show up in the schedule entry list:

We have tried to remove the most common hurdles to remotely deploying a sensor while maintaining flexibility in two key areas. First, the API itself is hardware agnostic, and the implementation assumes different hardware will be used depending on sensing requirements. Second, we introduce the high-level concept of "actions" which gives the sensor owner control over what the sensor can be tasked to do. For more information see Actions and Hardware Support.

This section provides an overview of high-level concepts used by scos-sensor.

• action: A function that the sensor owner implements and exposes to the API. Actions are the things that the sensor owner wants the sensor to be able to do. Since actions block the scheduler while they run, they have exclusive access to the sensor's resources (like the signal analyzer). Currently, there are several logical groupings of actions, such as those that create acquisitions, or admin-only actions that handle administrative tasks. However, actions can theoretically do anything a sensor owner can implement. Some less common (but perfectly acceptable) ideas for actions might be to rotate an antenna, or start streaming data over a socket and only return when the recipient closes the connection.

• acquisition: The combination of data and metadata created by an action (though an action does not have to create an acquisition). Metadata is accessible directly though the API, while data is retrievable in an easy-to-use archive format with its associated metadata.

• admin: A user account that has full control over the sensor and can create schedule entries and view, modify, or delete any other user's schedule entries or acquisitions.

• capability: Available actions, installation specifications (e.g., mobile or stationary), and operational ranges of hardware components (e.g., frequency range of signal analyzer). These values are generally hard-coded by the sensor owner and rarely change.

• plugin: A Python package with actions designed to be integrated into scos-sensor.

• schedule: The collection of all schedule entries (active and inactive) on the sensor.

• scheduler: A thread responsible for executing the schedule. The scheduler reads the schedule at most once a second and consumes all past and present times for each active schedule entry until the schedule is exhausted. The latest task per schedule entry is then added to a priority queue, and the scheduler executes the associated actions and stores/POSTs task results. The scheduler operates in a simple blocking fashion, which significantly simplifies resource deconfliction. When executing the task queue, the scheduler makes a best effort to run each task at its designated time, but the scheduler will not cancel a running task to start another task, even one of higher priority.

• schedule entry: Describes a range of scheduler tasks. A schedule entry is at minimum a human readable name and an associated action. Combining different values of start, stop, interval, and priority allows for flexible task scheduling. If no start time is given, the first task is scheduled as soon as possible. If no stop time is given, tasks continue to be scheduled until the schedule entry is manually deactivated. Leaving the interval undefined results in a "one-shot" entry, where the scheduler deactivates the entry after a single task is scheduled. One-shot entries can be used with a future start time. If two tasks are scheduled to run at the same time, they will be run in order of priority. If two tasks are scheduled to run at the same time and have the same priority, execution order is implementation-dependent (undefined).

• signals: Django event driven programming framework. Actions use signals to send results to scos-sensor. These signals are handled by scos-sensor so that the results can be processed (such as storing measurement data and metadata).

• task: A representation of an action to be run at a specific time. When a task acquires data, that data is stored on disk, and a significant amount of metadata is stored in a local database. The full metadata can be read directly through the self-hosted website or retrieved in plain text via a single API call. Our metadata and data format is an extension of, and compatible with, the SigMF specification - see sigmf-ns-ntia.

• task result: A record of the outcome of a task. A result is recorded for each task after the action function returns, and includes metadata such as when the task started, when it finished, its duration, the result (success or failure), and a freeform detail string. A TaskResult JSON object is also POSTed to a schedule entry's callback_url, if provided.

When deploying equipment remotely, the robustness and security of software is a prime concern. scos-sensor sits on top of a popular open-source framework, which provides out-of-the-box protection against cross site scripting (XSS), cross site request forgery (CSRF), SQL injection, and clickjacking attacks, and also enforces SSL/HTTPS (traffic encryption), host header validation, and user session security.

scos-sensor uses a open source software stack that should be comfortable for developers familiar with Python.

• Persistent metadata is stored on disk in a relational database, and measurement data is stored in files on disk.
• A scheduler thread running in a Gunicorn worker process periodically reads the schedule from the database and performs the associated actions.
• A website and JSON RESTful API using Django REST framework is served over HTTPS via NGINX, a high-performance web server. These provide easy administration over the sensor.

A functioning scos-sensor utilizes software from at least three different GitHub repositories. As shown below, the scos-sensor repository integrates everything together as a functioning scos-sensor and provides the code for the user interface, scheduling, and the storage and retrieval of schedules and acquisitions. The scos-actions repository provides the core actions API, defines the radio interface that provides an abstraction for all signal analyzers, and provides basic actions. Finally, using a real radio within scos-sensor requires a third scos-<signal analyzer> repository that provides the signal analyzer specific implementation of the radio interface where <signal analyzer> is replaced with the name of the signal analyzer, e.g. a USRP scos-sensor utilizes the scos-usrp repository. The signal analyzer specific implementation of the radio interface may expose additional properties of the signal analyzer to support signal analyzer specific capabilities and the repository may also provide additional signal analyzer specific actions.

• configs: This folder is used to store the sensor_definition.json file.
• docker: Contains the docker files used by scos-sensor.
• docs: Documentation including the documentation hosted on GitHub pages generated from the OpenAPI specification.
• entrypoints: Docker entrypoint scripts which are executed when starting a container.
• gunicorn: Gunicorn configuration file.
• nginx: Nginx configuration template and SSL certificates.
• schemas: JSON schema files.
• scripts: Various utility scripts.
• src: Contains the scos-sensor source code.
  • actions: Code to discover actions in plugins and to perform a simple logger action.
  • authentication: Code related to user authentication.
  • capabilities: Code used to generate capabilities endpoint.
  • handlers: Code to handle signals received from actions.
  • schedule: Schedule API endpoint for scheduling actions.
  • scheduler: Scheduler responsible for executing actions.
  • sensor: Core app which contains the settings, generates the API root endpoint.
  • static: Django will collect static files (JavaScript, CSS, …) from all apps to this location.
  • status: Status endpoint.
  • tasks: Tasks endpoint used to display upcoming and completed tasks.
  • templates: HTML templates used by the browsable API.
  • conftest.py: Used to configure pytest fixtures.
  • manage.py: Django’s command line tool for administrative tasks.
  • requirements.txt and requirements-dev.txt: Python dependencies.
  • tox.ini: Used to configure tox.
• docker-compose.yml: Used by docker-compose to create services from containers. This is needed to run scos-sensor.
• env.template: Template file for setting environment variables used to configure scos-sensor.

This section describes how to spin up a production-grade sensor in just a few commands.

We currently support Ettus USRP B2xx software-defined radios out of the box, and any Intel-based host computer should work. ARM-based single-board computers have also been tested, but we do not prepare pre-built Docker containers for them at this time.

1. Install git, Docker, and docker-compose.

2. Clone the repository.

git clone https://github.com/NTIA/scos-sensor.git
cd scos-sensor

3. Copy the environment template file and modify the copy if necessary, then source it. The settings in this file are set for running in a development environment on your local system. For running in a production environment, many of the settings will need to be modified. See Configuration section. Also, you are strongly encouraged to change the default ADMIN_EMAIL and ADMIN_PASSWORD before running scos-sensor. Finally, source the file before running scos-sensor to load the settings into your environment.

cp env.template env
source ./env

4. Run a Dockerized stack.

docker-compose up -d --build  # start in background
docker-compose logs --follow api  # reattach terminal

When running in a production environment or on a remote system, various settings will need to be configured.

As explained in the Quickstart section, before running scos-sensor, an environment (env) file is created from the env.template file. These settings can either be set in the environment file or set directly in docker-compose.yml. Here are the settings in the environment file:

• ADMIN_EMAIL: Email used to generate admin user. Change in production.
• ADMIN_PASSWORD: Password used to generate admin user. Change in production.
• BASE_IMAGE: Base docker image used to build the API container.
• CALLBACK_SSL_VERIFICATION: Set to “true” in production environment. If false, the SSL certificate validation will be ignored when posting results to the callback URL.
• DEBUG: Django debug mode. Set to False in production.
• DOCKER_TAG: Always set to “latest” to install newest version of docker containers.
• DOMAINS: A space separated list of domain names. Used to generate ALLOWED_HOSTS.
• GIT_BRANCH: Current branch of scos-sensor being used.
• GUNICORN_LOG_LEVEL: Log level for Gunicorn log messages.
• IPS: A space separated list of IP addresses. Used to generate ALLOWED_HOSTS.
• FQDN: The server’s fully qualified domain name.
• MAX_DISK_USAGE: The maximum disk usage percentage allowed before overwriting old results. Defaults to 85%. This disk usage detected by scos-sensor (using the Python shutil.disk_usage function) may not match the usage reported by the Linux df command.
• POSTGRES_PASSWORD: Sets password for the Postgres database for the “postgres” user. Change in production.
• REPO_ROOT: Root folder of the repository. Should be correctly set by default.
• SECRET_KEY: Used by Django to provide cryptographic signing. Change to a unique, unpredictable value. See https://docs.djangoproject.com/en/3.0/ref/settings/#secret-key.
• SSL_CERT_PATH: Path to server SSL certificate. Replace the certificate in the scos-sensor repository with a valid certificate in production.
• SSL_KEY_PATH: Path to server SSL private key. Use the private key for your valid certificate in production.

This file contains information on the sensor and components being used. It is used in the SigMF metadata to identify the hardware used for the measurement. It should follow the sigmf-ns-ntia Sensor Object format. See an example below. Overwrite the example file in scos-sensor/configs with the information specific to the sensor you are using.

{
    "id": "",
    "sensor_spec": {
        "id": "",
        "model": "greyhound"
    },
    "antenna": {
        "antenna_spec": {
            "id": "",
            "model": "L-com HG3512UP-NF"
        }
    },
    "signal_analyzer": {
        "sigan_spec": {
            "id": "",
            "model": "Ettus USRP B210"
        }
    },
    "computer_spec": {
        "id": "",
        "model": "Intel NUC"
    }
}

This section covers authentication, permissions, and certificates used to access the sensor, and the authentication available for the callback URL. Two different types of authentication are available for authenticating against the sensor and for authenticating when using a callback URL. Note that the certificate authorities (CAs), SSL certificates, private keys, and JWT public keys used in this repository are for testing and development purposes only. They should not be used in a production system.

The sensor can be configured to authenticate using OAuth JWT access tokens from an external authorization server or using Djnago Rest Framework Token Authentication.

This is the default authentication method. To enable Django Rest Framework Authentication, make sure AUTHENTICATION is set to TOKEN in the environment file (this will be enabled if AUTHENTICATION set to anything other than JWT).

A token is automatically created for each user. Django Rest Framework Token Authentication will check that the token in the Authorization header ("Token " + token) matches a user's token.

To enable OAuth 2 JWT Authentication, set AUTHENTICATION to JWT in the environment file. To authenticate, the client will need to send a JWT access token in the authorization header (using "Bearer " + access token). The token signature will be verified using the public key from the PATH_TO_JWT_PUBLIC_KEY setting. The expiration time will be checked. Only users who have an authority matching the REQUIRED_ROLE setting will be authorized.

The token is expected to come from an OAuth2 authorization server. For more information, see https://tools.ietf.org/html/rfc6749.

The NGINX web server requires an SSL certificate to use https. The certificate and private key should be set using SSL_CERT_PATH and SSL_KEY_PATH in the environment file. Note that these paths are relative to the configs/certs directory.

Optionally, client certificates can be required. To require client certificates, uncomment ssl_verify_client on; and ssl_ocsp on; in nginx/conf.template. Set the CA certificate used for validating client certificates using the SSL_CA_PATH (relative to configs/certs) in the environment file.

It is recommended to create your own CA for testing. For production, make sure to use certificates from a trusted CA. For testing, you can use the certificates and keys in configs/certs/test or you can use scripts/create_certificates.py to create the test CA certificate, test server certificate, and test client certificate. This script can also be used with an existing CA. Here are the instructions to use create_certificates with an existing CA.

1. To configure the create_certificates.py script, use create_certificates.ini. In create_certificates.ini, set ca_private_key_path and ca_certificate_path to the path of your CA private key and certificate. Configure the remaining parameters as desired. The SAN (subject alternative name) parameters will need to be set to the appropriate IP addresses and DNS names of your server and client.

2. While in scos-sensor root directory, run the create_certificates.py script passing the following arguments in the listed order:

   • ini_path - path to the create_certificates.ini file.
   • ini_section - section of the INI file to use.
   • key_passphrase - Passphrase to use to encrypt private keys. Set to None to disable encryption.

   The following certificates will be generated:

   • sensor01_private.pem - sensor private key.
   • sensor01_certificate.pem - sensor certificate.
   • sensor01_client_private.pem - client private key.
   • sensor01_client.pem - client certificate.

3. Copy sensor01_private and sensor01_certificate to the computer where the scos-sensor will run. If you are using client certificates, also copy the CA certificate used to generate the certificates. Make sure the certificates are somewhere in configs/certs, and that SSL_CERT_PATH and SSL_KEY_PATH (in the environment file) are set to the paths of the certificates relative to configs/certs. If you are using client certificates, set SSL_CA_PATH to the path of the CA certificate relative to configs/certs.

4. Run scos-sensor. If you are using client certificates, use sensor01_client_private.pem and sensor01_client to connect to the API.

The create_certificates.py script can also generate a new CA and use it for generating the certificates. To run create_certificates.py this way, comment out ca_private_key_path and ca_certificate_path in create_certificates.ini, make sure ca_private_key_save_path and the other parameters are set as desired, then repeat steps 2-4 above. The CA private key file (saved to ca_private_key_save_path) and the CA public key (scostestca.crt) will be generated in addition to the files listed in step 2 above.

The API requires the user to either have an authority in the JWT token matching the the REQUIRED_ROLE setting or that the user be a superuser. New users created using the API initially do not have superuser access. However, an admin can mark a user as a superuser in the Sensor Configuration Portal. When using JWT tokens, the user does not have to be pre-created using the sensor's API. The API will accept any user using a JWT token if they have an authority matching the required role setting.

OAuth and Token authentication are supported for authenticating against the server pointed to by the callback URL. Callback SSL verification can be enabled or disabled using CALLBACK_SSL_VERIFICATION in the environment file.

A simple form of token authentication is supported for the callback URL. The sensor will send the user's (user who created the schedule) token in the authorization header ("Token " + token) when posting results to callback URL. The server can then verify the token against what it originally sent to the sensor when creating the schedule. This method of authentication for the callback URL is enabled by default. To verify it is enabled, set CALLBACK_AUTHENTICATION to TOKEN in the environment file (this will be enabled if CALLBACK_AUTHENTICATION set to anything other than OAUTH). PATH_TO_VERIFY_CERT, in the environment file, can used to set a CA certificate to verify the callback URL server SSL certificate. If this is unset and CALLBACK_SSL_VERIFICATION is set to true, standard trusted CAs will be used.

The OAuth 2 password flow is supported for callback URL authentication. The following settings in the environment file are used to configure the OAuth 2 password flow authentication.

• CALLBACK_AUTHENTICATION - set to OAUTH.
• CLIENT_ID - client ID used to authorize the client (the sensor) against the authorization server.
• CLIENT_SECRET - client secret used to authorize the client (the sensor) against the authorization server.
• OAUTH_TOKEN_URL - URL to get the access token.
• PATH_TO_CLIENT_CERT - client certificate used to authenticate against the authorization server.
• PATH_TO_VERIFY_CERT - CA certificate to verify the authorization server and callback URL server SSL certificate. If this is unset and CALLBACK_SSL_VERIFICATION is set to true, standard trusted CAs will be used.

In src/sensor/settings.py, the OAuth USER_NAME and PASSWORD are set to be the same as CLIENT_ID and CLIENT_SECRET. This may need to change depending on your authorization server.

"Actions" are one of the main concepts used by scos-sensor. At a high level, they are the things that the sensor owner wants the sensor to be able to do. At a lower level, they are simply Python classes with a special method __call__. Actions are designed to be discovered programmatically in installed plugins. Plugins are Python packages that are designed to be integrated into scos-sensor. The reason for using plugins to install actions is that different actions can be offered depending on the hardware being used. Rather than requiring a modification to scos-sensor repository, plugins allow anyone to add additional hardware support to scos-sensor by offering new or existing actions that use the new hardware.

Common action classes can still be re-used by plugins through the scos-actions repository. The scos-actions repository is intended to be a dependency for every plugin as it contains the actions base class and signals needed to interface with scos-sensor. These actions use a common but flexible radio interface that can be implemented for new types of hardware. This allows for action re-use by passing the radio interface and the required hardware and measurement parameters to the constructor of these actions. Alternatively, custom actions that support unique hardware functionality can be added to the plugin.

The scos-actions repository can also be installed as a plugin which uses a mock signal analyzer.

scos-sensor uses the following convention to discover actions offered by plugins: if any Python package begins with "scos_", and contains a dictionary of actions at the Python path package_name.discover.actions, these actions will automatically be available for scheduling.

The scos-usrp plugin adds support for the Ettus B2xx line of software-defined radios. It can also be used as an example of a plugin which adds new hardware support and re-uses the common actions in scos-actions.

For more information on adding actions and hardware support, see scos-actions.

The following techniques can be used to make local modifications. Sections are in order, so "Running Tests" assumes you've done the setup steps in “Requirements and Configuration”.

It is highly recommended that you first initialize a virtual development environment using a tool such a conda or venv. The following commands create a virtual environment using venv and install the required dependencies for development and testing.

python3 -m venv ./venv
source venv/bin/activate
python3 -m pip install --upgrade pip # upgrade to pip>=18.1
python3 -m pip install -r src/requirements-dev.txt

Ideally, you should add a test that covers any new feature that you add. If you've done that, then running the included test suite is the easiest way to check that everything is working. In any case, all tests should be run after making any local modifications to ensure that you haven't caused a regression.

scos-sensor uses pytest and pytest-django for testing. Tests are organized by application, so tests related to the scheduler are in ./src/scheduler/tests. tox is a tool that can run all available tests in a virtual environment against all supported versions of Python. Running pytest directly is faster, but running tox is a more thorough test.

The following commands install the sensor's development requirements. We highly recommend you initialize a virtual development environment using a tool such a conda or venv first.

cd src
pytest          # faster, but less thorough
tox             # tests code in clean virtualenv
tox --recreate  # if you change `requirements.txt`
tox -e coverage # check where test coverage lacks

The docker-compose file and application code look for information from the environment when run, so it's necessary to source the following file in each shell that you intend to launch the sensor from. (HINT: it can be useful to add the source command to a post-activate file in whatever virtual environment you're using).

cp env.template env     # modify if necessary, defaults are okay for testing
source ./env

Then, build the API docker image locally, which will satisfy the smsntia/scos-sensor and smsntia/autoheal images in the Docker compose file and bring up the sensor.

docker-compose down
docker-compose build
docker-compose up -d
docker-compose logs --follow api

Running the sensor API outside of Docker is possible but not recommended, since Django is being asked to run without several security features it expects. See Common Issues for some hints when running the sensor in this way. The following steps assume you've already set up some kind of virtual environment and installed python dev requirements from Requirements and Configuration.

docker-compose up -d db
cd src
./manage.py makemigrations
./manage.py migrate
./manage.py createsuperuser
./manage.py runserver

• The development server serves on localhost:8000, not :80
• If you get a Forbidden (403) error, close any tabs and clear any cache and cookies related to SCOS Sensor and try again
• If you're using a virtual environment and your signal analyzer driver is installed outside of it, you may need to allow access to system sitepackages. For example, if you're using a virtualenv called scos-sensor, you can remove the following text file: rm -f ~/.virtualenvs/scos-sensor/lib/python3.6/no-global-site-packages.txt, and thereafter use the ignore-installed flag to pip: pip install -I -r requirements.txt. This should let the devserver fall back to system sitepackages for the signal analyzer driver only.

Besides running the test suite and ensuring that all tests are passed, we also expect all Python code that's checked in to have been run through an auto-formatter.

This project uses a Python auto-formatter called Black. You probably won't like every decision it makes, but our continuous integration test-runner will reject your commit if it's not properly formatted.

Additionally, import statement sorting is handled by isort.

The continuous integration test-runner verifies the code is auto-formatted by checking that neither isort nor Black would recommend any changes to the code. Occasionally, this can fail if these two autoformatters disagree. The only time I've seen this happen is with a commented-out import statement, which isort parses, and Black treats as a comment. Solution: don't leave commented-out import statements in the code.

There are several ways to autoformat your code before committing. First, IDE integration with on-save hooks is very useful. Second, there is a script, scripts/autoformat_python.sh, that will run both isort and Black over the codebase. Lastly, if you've already pip-installed the dev requirements from the section above, you already have a utility called pre-commit installed that will automate setting up this project's git pre-commit hooks. Simply type the following once, and each time you make a commit, it will be appropriately autoformatted.

pre-commit install

You can manually run the pre-commit hooks using the following command.

pre-commit run --all-files

• SCOS Control Plane API Reference
• SCOS Data Transfer Specification
• SCOS Actions
• SCOS USRP

See LICENSE.

For technical questions about scos-sensor, contact Justin Haze, [email protected]