Introduction: Intel Edison Based Mobile Sensor Network for the Internet of Things
Introduction and Overview
The Intel Edison is a nifty little computer that combines many advantages of other maker plattforms. It can be combined with breakout boards of various sizes and purposes.
In this article we will learn what components are needed to turn our Intel Edison into our own base transceiver station (BTS) to run our own local mobile sensor network.
But before we start getting all the hardware and tools ready, we need to think about a couple things:
(1) Use Case
What do we want to do with the network? Why and what data do we want to collect. Are we just receiving data or do we also want to remotely control things?
(2) Mobile Broadband Coverage
Since we plan to use mobile broad band for connecting us with the outside world, we need to check which providers are covering the planned location for our Intel Edison BTS. If there is more than one mobile operator in range, check to get the best offer. All we need for our application is simple GSM. Paying for 3G (UMTS) or 4G (LTE) would be a waste of money and ressources.
(3) Long Range or Short Range
Since we are planning to set-up a mobile ultra-narrow bandwith network we need to think about the area we would like to cover as well. For things (sensors / actuators) being indoor or beneath the surface we have to choose a different radio technology than for things being in sight and close to our BTS.
(4) Choosing the right frequency
Without going to much into detail, we can assume the following:
The lower the frequency is
- the smaller is the bandwith,
- the longer is the range,
- the deeper the signal gets under the surface or inside buildings,
- the lower is the power consumption.
The higher the frequency is
- the more bandwidth is possible,
- the shorter is the range,
- the more a free line of sight is needed for the signal to be received.
In addition we need to keep in mind that some frequency bands like the ISM band are license free but are subject to usage regulations. Other frequencies have been licensed and we are bound to use the owners hardware to build our infrastructure.
(5) Power Management and Installation Details
If we have a powerline available we do not need to care about our consumption to much but often we want to connect things in remote locations. Therefore we need to plan and consider our power consumption and management carefully. Sometimes even changing a battery every now and then might not be an option. We might want to check in advance, if we can use solar power, wind energy or if we can harvest energy somehow.
Last, but not least, depending on our use cases and the location of our things we further need to think about the installations to be robust (vibration, climate, destruction, ...)
Step 1: Assumptions and Prerequisites
Let us assume we are in a smart living context, we would like to connect things in a range of 5km around our Intel Edison base tranceiver station, we have a mostly free line of sight, we have a generic GSM coverage, we are in Europe and we can use soler power where needed.
First, we need to put together a minimal hardware setup.
For the base station we need:
- Intel Edison
- Breakout board for the Intel Edison. (We chose the Arduino breakout board for easy prototyping)
- Power adapter
- 2 micro USB cables
- GSM module (We chose the Adafruit FONA)
- RF module (We chose DigiKey XBee 868 Mhz)
- Maybe: Serial to USB adapters
For each group of sensors or actuators we need
- the sensors and actuators we are going to connect (Especially for prototyping the sensors and actuators from Seeed Grove come in handy)
- the RF module (the same as above!)
- any kind of power supply (e.g. solar, battery, wind, power line)
- In case of a more complex use case / sensor node we might need to add e.g. some microcontrollers
For prototyping additonal tools and material is helpful, like breadboards, jumper wires, soldering equipment, multimeter, ...
For programming, configuring and monitoring the network we need a computer.
Step 2: Setup and Installations
Step 3: Putting It All Together for Lab Testing
When putting it all together we should do it one by one.
First connect the GSM module to the Intel Edison board. The easiest way would be using a serial-to-USB adapter and attaching it to the USB port using a hub. Try some demo programms and make sure your GSM credentials are correct and you can connect to the mobile broadband.
Second connect the RF module the same way you. Connect another RF module to your PC and try getting a serial connection between the Intel Edison and your computer.
The third and most interesting part is the assembly and testing of your "thing", be it a sensor or an actuator. I highly recommend connecting a RF module to the computer first and try to connect to the thing before trying it with the Intel Edison directly.
When everything is working seperately, but everything together for further integration, lab testing and troubleshooting. If you are still just sending SMS to yourself with the collected data, it is now the right time to register your Intel Edison board to one of the ecosystems. Please check out Intel Cloud Analytics.
Step 4: Finalising and Network Installation
Now that everything is tested, up and running stable, we are ready to establish our IoT network, assemble and install our things, take care of housing and power management.
If we can get hold of a 3D printer we print individual housings, otherwise we can build our own with plywood.
For my base transceiver station I chose a big solar panel, the Seeed Studio Lipo Rider Pro and a 3,7 LiPo rechargable battery, the box is still prototype, since this is a showcase. The sensor node I put in simple plastic electronics box.
Is everything working as epected? Congratulations! You are now your own mobile network operator.
And what do you do with all the data now?
Step 5: Connecting to the Cloud and Resumee
Depending on the use cases we had in mind when we started our network we might need to use different ways to get our data to our applications. One easy way is to choose a cloud ecosystem and push the data via mobile broadband to the cloud. Using e.g. REST we can now integrate the collected information to our processes and workflows.
In simpler cases a dashboard might just be goog enough.
If we take a closer look at the work we have done, we can easily see, that we are covering the whole IoT process end-2-end. We have sensors and actuators, aggregate data, transmit it via our own radio network, have our own gateway to the cloud where our application now makes use if our information.
We have a be nice policy.
Please be positive and constructive.