Actions
Forum Thread Details
  • State Not Answered
  • Replies 6 replies
  • Subscribers 222 subscribers
  • Views 1360 views
  • Users 0 members are here
Related

ZigBee, would this Ad Hoc network idea work?

Former Member

 

I’m trying to get a general understanding for how a mesh Ad Hoc network could be created. ZigBee seems as a cheap solution so that’s why I’m going for it. However if you guys have a better suggestion (i.e. Bluetooth) feel free to suggest it.

 

Ok, here goes. My basic idea is this. I have moving objects (large area indoors) and I want to construct a simple warning system. When one device detects another I basically just want to play a tune or make a light flash, nothing more. The approach will probably be using RSSI but that is not important right now. I want this to be cheap and I do not want any kind of fixed infrastructure (disqualifying wifi). The devices would randomly come into contact, do some kind of handshake, talk while they are close and finally move out of range.


Is it even possible to create such a thing with ZigBee? The standard requires a ZigBee Coordinator so I was thinking that one moving object could be this and the rest ZigBee routers. Would this approach even work? Perhaps one could equip a few Raspberry Pi’s with “XBee ZB ZigBee RF Modules” and then take it from there.

 

 

  • 0 in reply to gadget.iom

     

    I’m trying to create a warning system so the further away one node can detect another, the better. We are talking running speed so guesstimating the reaction time… let’s say at least 5-10 meters/yards. If I can’t get a connection in time it would defeat the purpose of the whole idea.

     

  • 0 in reply to Former Member

    Have a look at these modules...

    http://www.dx.com/p/433mhz-rf-transmitter-module-receiver-module-link-kit-for-arduino-arm-mcu-wl-green-220194#.ViDtvrzvFHg

     

    You could have one of each in your devices.

    The transmitter sends out a pulse every second for example.

    The receiver can listen for any responses.

     

    You would probably need to disable the local receiver while transmitting, or the device would react to its own transmitter.

    A better approach would be to transmit a code unique to each device, so the receiver can continuously listen, but ignore its own code.

    This could also prove useful in team based games.

  • 0 in reply to gadget.iom

     

    I’m more interested in creating a WPAN. I want to be able create and evaluate some kind of Ad Hoc network providing both a medium access control and the physical layer. Something of the shelf. The modules you are referring to would require me to do all the low layer work. For example when multiple nodes are close, transmitting and receiving uncontrollably, it would all (excuse the language) go to hell.

    My question is more about the architecture of ZigBee. Can you have one node being the ZigBee Coordinator (this part is a must) and all the others being ZigBee Routers?

    From what I see ZigBee end devices can only     talk to the parent node (either the Coordinator or a Router) meaning that if two ZED’s meet they wouldn’t be able to talk


  • 0

    How many nodes would be the maximum?

    t seems you could use any low-cost radio with a small associated CPU for this application as long as collisions can be detected. It probably wouldn't need a big complicated ad-hoc protocol like Zigbee.

    • all nodes periodically broadcast their own ID - if a collision is detected, they re-transmit at a random time later
    • all nodes evaluate RSSI of all received broadcasts to determine who is close and how close
    • the periodic transmission time and the random re-transmission range of times are dictated by how fast the nodes can approach each other and how many nodes are active
    • if you have a lot of nodes, some careful statistical calculations will be needed to determine how many nodes can be handled before the congestion becomes a problem
  • 0

     

    I want this to work for 4-5 moving nodes initially. I also wanted this to be a scalable system (up to say a 100 nodes) that can handle data transfer between nodes when they meet. Step 1 was detection. Only using RSSI is definitely not a good solution, I just mentioned it to give some background info without going into details. I will probably use RSSI but I will need some kind of radio propagation model (perhaps a Log-distance model) and data preprocessing – Kalman or a particle filter. But then again, a may go a different route and use TOA, TDOA or AOA combined with additional sensors and skip or use RSSI as an extra help. This is why I need some options.

     

     

    I’m more interested in the way ZigBee works. I keep hearing about it but I don’t understand if it would work in an environment where everything is moving? Bluetooth works, but only in a star topology with 1-2s delay time for each pairing. That is not acceptable.

     

    There is a 802.11p/1609 standard for car-to-car communication. This would probably be better (regarding moving nodes) but it’s not something suited for indoor use and I bet this an expensive solution.

     

     

    If you guys say that ZigBee doesn’t work and building everything from scratch is the simplest solution then perhaps that is the way to go.