LEACH Protocol
-
From my understanding, LEACH protocol would replace the MySensors protocol, so technically it would not be an additional feature.
I do think it would work just fine on the same hw the MySensors protocol supports, but it would be a completely separate network which would not leverage any mysensors functionality. But perhaps some of the drivers could be used.
It is not in the roadmap of MySensors to support multiple protocols. The MySensors protocol is the core of the whole MySensors library. -
@SebastianB For what I know of LEACH its abstraction kind of replaces the MySensors layer. MySensors handles the network topology (routing, repeating, forwarding) . I guess you can learn a lot from how MySensors handles the different radio protocols (rfm24 etc.) but choose to build on the radio interface of rfm24/69
-
@AWI Some function of mySensors are very similar - like they are hierarhical. Tell me - is there is a phisical difference between nodes and reapeting nodes? Or is this only "software"?
@SebastianB Software only but like @Anticimex mentioned LEACH would replace the MySensors networking. However you could use the serial protocol or MQTT for message encoding
-
From my understanding, LEACH protocol would replace the MySensors protocol, so technically it would not be an additional feature.
I do think it would work just fine on the same hw the MySensors protocol supports, but it would be a completely separate network which would not leverage any mysensors functionality. But perhaps some of the drivers could be used.
It is not in the roadmap of MySensors to support multiple protocols. The MySensors protocol is the core of the whole MySensors library.@Anticimex Yes that is correct. I just want to change drivers and sensors to be phisically the same - but implement rounds, gathering informations in different clusters, and add changing functions of nodes - in one round they are cluster head nodes in another a normal node....
Yes I understand. I just want to make LEACH phisical - because most of research is about simulations, and many people are creating garbage articles about it...
If I may ask you what harware would you propose? -
@gohan Ok that make my work easier. Thanks for your reply.
-
@Anticimex Yes that is correct. I just want to change drivers and sensors to be phisically the same - but implement rounds, gathering informations in different clusters, and add changing functions of nodes - in one round they are cluster head nodes in another a normal node....
Yes I understand. I just want to make LEACH phisical - because most of research is about simulations, and many people are creating garbage articles about it...
If I may ask you what harware would you propose?@SebastianB you can see on the main site what hardware is supported by mysensors. I am not familiar with the LEACH protocol so I don't know if the has particular requirements on drivers and hardware. This site and forum is specific to mysensors so there is not much information on other technologies (for natural reasons).
-
@Anticimex Yes that is correct. I just want to change drivers and sensors to be phisically the same - but implement rounds, gathering informations in different clusters, and add changing functions of nodes - in one round they are cluster head nodes in another a normal node....
Yes I understand. I just want to make LEACH phisical - because most of research is about simulations, and many people are creating garbage articles about it...
If I may ask you what harware would you propose?@SebastianB Keep in mind that a repeater node can not sleep or it will not receive messages and so it will use more power and if I understood correctly how LEACH works all nodes are "repeaters" on a rotation basis. I don't know if it will be a winning strategy. Also commercial products like zwave use a mesh network but only powered nodes act as "repeaters".
-
@SebastianB Keep in mind that a repeater node can not sleep or it will not receive messages and so it will use more power and if I understood correctly how LEACH works all nodes are "repeaters" on a rotation basis. I don't know if it will be a winning strategy. Also commercial products like zwave use a mesh network but only powered nodes act as "repeaters".
@gohan LEACH works like this -> the head nodes are choosen, normal nodes connect to them creating clusters. The normal nodes have special "time slots" (TDMA). Normal nodes transmit data in time that was set for them by head nodes. So for the time where nobody is transmiting - the antenas are turned off to save the energy. It is all mostly based on synchronization. After this step another group of heads nodes are choosen.... It may be - I also can check it and send you my observation. The "repeates" like you call them - they are choosen in every round - differend sensor - so the use of the energy is balanced. If the "repeater" die in one round in next round someone else will take his place.... My project is to create a wirelss sensor network that is created from cheap nodes - to observe enviroment... Implementation - could be by plane - so we will not know where they really are but only know the region - they will create a grid and try to gather efficiently information... If they die we just drop another group of sensors - for example to gather a information about fire in forests....
