Actions
Forum Thread Details
  • State Verified Answer
  • Replies 4 replies
  • Answers 2 answers
  • Subscribers 223 subscribers
  • Views 799 views
  • Users 0 members are here
Related

Wireless modules selection help

koudelad

Hello,

for start, a little introduction: microcontrollers are my hobby. I started with 8bit PICs, did the great online MOOC with Tiva C microcontroller and now playing with incredible PSoCs 4 and 5LP.

My goal is to build a small home/office measurement system. The idea is to have different sensors on different places (temperature, humidity, air pressure, light sensors, PIR motion sensors...) at a house and to be able to display "realtime" readings and also have the possibility to store the values somewhere. Cloud storage may be an option, but I could do with some local db on LAN.

 

The thing is I need help with the idea, how to build a wireless sensor network (technology, topology, HW...). I have seen source codes of some wireless protocols (MiWi, WiFi, ...) but it is way too complicated. I tried ESP-12 module, but the more I have read about it, the less I trust it (especially when it is not supported by chip manufacturers). I have a CC3100 and CC2650 Laundchpad, but they also seem complicated to me. Perhaps Energia could make it a bit easier?

 

One of my ideas: use Raspberry Pi as a hub and connect other sensor boards via Bluetooth (serial port profile?).

Another overkill idea: use Raspberry Pi as a sensor board itself and connect it via WiFi somewhere, where the data could be stored. (Of course, this draws huge amount of current).

 

Are there any easy to use, wireless, point to multipoint modules? With embedded security, supported by manufacturer, upgradeable. My goal is to transfer simple hex/ascii data and do the processing somewhere else, after storing to db. I could live with simple DIY boards including a sensor and a wireless module that would need to be preprogrammed (for example via PC, before use) to communicate to just one endpoint. Encryption and pairing mechanism will be preset and transparent for the managing microcontroller (so some simple low power 8bit MCU could be used?).

 

Thank in advance for any tips.

David

Parents
  • 0

    If this is for home why do you need a standard protocol with embedded security etc ?

     

    I f you have to have everything you can expect it to complicated or expensive, or even both.

     

    The ESP thingies are fine - lots of user and open source support and dead cheap. Add you own security protocol if you think anyone cares about snooping your messages, NSA could break it in a few uS but they could break anything you could afford in a few minutes so who cares. The average drive by snooper isn't going to be able to break even the simplest home brew protocol because obscurity works in your favor  here.

     

    If you don't like the amount of power that the WiFi stuff uses then go for a nice simple sub GHz radio chip - the TI ones are nice, well documented and easy to get going. Once again - do your own protocol and make a serial to radio adapter for your PC and sensor to radio for your sensors. I think TI do cheap bits for all this.

     

    The LauncXL-CC1310 is a nice way to go but pricey at £23.

     

    The Cypress Bluetooth baords are cheaper but BT is more complicated.

     

    My top pick is 903-3062 from RS at £4.10 -  its an 868 MHz transceiver from RF Solutions - use it with a low power demo board from ST (start at £7.65) - or any other that you like.

     

    MK

     

     

     

     

  • 0 in reply to michaelkellett

    Thank you for your reply and all the tips.

     

    Well, I always think about security of things I use. It's not about NSA or any other government agency. It is about using the things in a way that neighbors and script kiddies with software defined radios don't watch (or play with) everything wireless you own. Many manufacturers of consumer electronics don't have budget for a security department and this is why we read about hacked light bulbs, toys, cars etc. almost every week. I don't want to do it the same "poor security" way.

     

    You basically confirmed what I thought. WiFi and Bluetooth are quite complex and 868 MHz transceivers are the easiest to use. I just tried to avoid writing a custom protocol and reinventing the wheel. There are so many wireless protocols with dedicated chips by many manufacturers, but the software stack needed to use it is difficult for a hobbyist to use.

Reply
  • 0 in reply to michaelkellett

    Thank you for your reply and all the tips.

     

    Well, I always think about security of things I use. It's not about NSA or any other government agency. It is about using the things in a way that neighbors and script kiddies with software defined radios don't watch (or play with) everything wireless you own. Many manufacturers of consumer electronics don't have budget for a security department and this is why we read about hacked light bulbs, toys, cars etc. almost every week. I don't want to do it the same "poor security" way.

     

    You basically confirmed what I thought. WiFi and Bluetooth are quite complex and 868 MHz transceivers are the easiest to use. I just tried to avoid writing a custom protocol and reinventing the wheel. There are so many wireless protocols with dedicated chips by many manufacturers, but the software stack needed to use it is difficult for a hobbyist to use.

Children
  • 0 in reply to koudelad

    David Koudela wrote:

     

    Well, I always think about security of things I use. It's not about NSA or any other government agency. It is about using the things in a way that neighbors and script kiddies with software defined radios don't watch (or play with) everything wireless you own. Many manufacturers of consumer electronics don't have budget for a security department and this is why we read about hacked light bulbs, toys, cars etc. almost every week. I don't want to do it the same "poor security" way.

    Being security minded is certainly no bad thing! image

    Based on what you have said, this video of a talk by Ken Munro might interest you: https://www.youtube.com/watch?v=65b201MLyOY

  • 0 in reply to koudelad

    The point I was trying to make is that this is a case where "re-inventing the wheel" is a good thing. If you use your own protocol with the simplest kind of encryption it will be very unlikely that anyone will make the effort required to break it - because the prize is too small. If you use a standard protocol and encryption the one thing you can be sure of is that huge amounts of effort will go into breaking it.

     

    This approach works very well for small footprint projects.

     

    The other thing to keep in mind is that most security breaches arise from poor implementation of a standard approach  - the recent internet camera bot attacks work because the cameras have an unprotected back door.

     

    MK