| ??? 04/24/08 05:52 Read: times |
#153928 - Nothing I feel Responding to: ???'s previous message |
Richard Erlacher said:
Erik Malund said:
with a mask of zero any unassigned sign should respond.
the mask mask out bits not to be checked, so mask = 0 means no ID bits to be checked. If you have worked througth the mask and match as described there will be no unassigned signs if you ever would get to a mask of zero. Thus, with correct code, there will never be a mask of zero. Actually, by the principle of "belt, suspenders and a piece of rope" it might be an idea to do a mask = 0 at the end just to get the piece of rope used too. Erik Erik, The way in which you've described your approach is not very clear from what little you've said. If you have 64 bits of address, then you have to search through a field of 2^64 (very many) or even 2^48 (still very many, though fewer) possible addresses. If you have no information about which addresses are used and which are not, which, so far, has been the case, unless some more information is necessary, then one has to poll 2^64 addresses in order to discern which ones are occupied and which are not. At the typical maximum of 8 kbps available with cheap RF links, or even the maximal 250 kbps offered under absolutely perfect conditions by, say, Zibgee hardware, that will still be in progress when the sun burns out, right? If you simply issue a mask address, i.e. a subset of the total address range, or anything else other than a single unique address from the available range, knowing nothing about who or what is out there on your frequency band, I don't think you're going to learn much. OTOH, if your slaves can transmit at will without a "stroke" to tell them when, and uniquely, there will be lots of unintelligible traffic, which won't help much. One has to know more about the protocol. You haven't shared much. Nobody else has attempted to explain your approach, since they apparently don't recognize it either. What intelligence do you obtain from a no-response result? How is the "mask" to be interpreted? What sort of price range does your RF link that accommodates a ~450 kbps data rate fit? Is it in the >$5 range that the "low-cost" devices cost? What I understand is http://www.8052.com/forumchat/read.phtml?id=153780 . I think he isnt polling the slaves at all ? to find whos out. Erik Malund said:
nope, it's quite fast. With the proper algorithm you can 'find', say, 64 slaves (each with a 48 bit address) in about 2 seconds. Then assign them an 8 bit 'node address' And with RF data rate of 460k baud , that implies the RF radio layer is fighting for that result to come with some networking options selected . After the CA algorithm delievers data to master , that particular node is masked . Now as we move on the time factor keeps reducing . Ap |
