Wireless Topology

Discussion in 'C-Bus Wireless Hardware' started by KevinH, Feb 16, 2005.

  1. KevinH

    KevinH

    Joined:
    Aug 3, 2004
    Messages:
    171
    Likes Received:
    0
    Location:
    Yorkshire. UK
    Just wanting to clarify the topolgy that C-Bus wireless implements. Am I right that a true mesh is created and that wireless units that are out of range of other units then use closer units as relays ? If so is this a dynamically managed routing which copes with changes or is it fixed at install when you add new units ? The illustratons all show point to point.
    I am confused a little perhaps by the apparent associations that are built using 'learn' mode between two fixed units at install. I realise all units go into 'learn' mode simultaneously but is it possible to say associate a remote button that is not in RF range with the intended original which is on the same network ?
    I have one particular install which has very thick walls, I realise that my RF range will be limited but it should work as I am hoping distances can be covered by a chain of devices although response will be slower. Wired isn't an option.
    I am also assuming the 'mesh' only works within the same network identifier such that two RF networks that overlapped at some point wouldn't actually be linked except through any wired gateways, Again the gateway documentation concerns me when it recommends siting the gateway within RF range of all switches. Maybe this is just (sensibly) to reduce network traffic and increase speed.
    I hope this isn't covered elsewhere but I am having a little difficulty finding good information on the C-Bus wireless products. Someone mentioned a specific Clipsal site for information on the new wireless stuff - is there such a thing ?

    Kevin
     
    KevinH, Feb 16, 2005
    #1
  2. KevinH

    Newman

    Joined:
    Aug 3, 2004
    Messages:
    2,203
    Likes Received:
    0
    Location:
    Adelaide, South Australia
    Kevin

    All devices in a C-Bus Wireless network MUST be within range of each other. Commands are NOT repeated from unit-to-unit-to-unit-to..... as you suggest. This is why the instructions show traffic from point-to-point.

    The learn mode process creates a logical grouping of units into a single network. This is why the gateway must be in range of all devices, as must all units in fact, except the remote control which transmits only.

    If you have a large installation or one with poor RF performance then the only solution is to break the single large network up into smaller networks.
     
    Newman, Feb 16, 2005
    #2
  3. KevinH

    ashleigh Moderator

    Joined:
    Aug 4, 2004
    Messages:
    2,393
    Likes Received:
    25
    Location:
    Adelaide, South Australia
    Before unleashing a flame war:


    Truly dymanic automatically routed mesh networks are *extremely* painful to write software code for, and need lots of processing resource.

    (Been & done a subset of that a few years ago and its not nice).

    You will see a lot of mesh network products around, but there's always a catch if you dig deep enough. Its certainly the case that the sustainable bandwidth is only a fraction of that available, and they don't lend themselves to ad-hoc addition of nodes at any time. Routing discovery is painful.

    Remember - its a tiny processor in a tiny space, and for the retrofit 2-wire products there is only a tiny amount of power available.
     
    ashleigh, Feb 17, 2005
    #3
Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.