REGISTER | LOGIN
Breaking News
Blog

IoT Calls for Simpler Protocol

View Comments: Newest First | Oldest First | Threaded View
Page 1 / 2   >   >>
LarryM99
User Rank
Author
Re: Processing Power for Chirps vs IP
LarryM99   9/30/2014 7:24:02 PM
NO RATINGS
@fdacosta, that type of hierarchy could add value to a set of sensors. For example, it could perform fusion of the data from a set of colocated sensors. That would be relatively difficult to do with each sensor being queried independently. But the hierarchy could also be flattened into the sensor package itself, which would remove the dependence on a central node. An example would be a set of sensors in the substrate of a bridge or road. More sophisticated routing would allow them to self-organize into an ad hoc reporting network and aggregate data based on relative position between them.

Larry M.

fdacosta
User Rank
Author
Re: Processing Power for Chirps vs IP
fdacosta   9/30/2014 6:53:49 PM
NO RATINGS
@Larry. Agreed. The Internet of Things is about connectivity but its also about Things.  Billions on them. What would be an alternative approach to putting networking stacks on every one of these things? Thats where the device-propogator-integrator model I describe in my book comes in.  Francis 

LarryM99
User Rank
Author
Re: Processing Power for Chirps vs IP
LarryM99   9/30/2014 6:05:32 PM
NO RATINGS
If a hundred of those wireless temperature sensors are built into a bridge then they very well might want to have a full networking stack in them. If the sensors are strictly for very local consumption then it almost certainly makes sense to have a minimal interface, but the whole point of IoT is to make sensors accessible across a network.

Larry M.

fdacosta
User Rank
Author
Processing Power for Chirps vs IP
fdacosta   9/30/2014 11:40:14 AM
NO RATINGS
The reason I have been advocating chirps is that IP compression and its variants all require processing power at the end point that simply won't be there, if we plan on building billions of such simple, purpose built devices. IP makes sense when we are talking about more expensive general purpose devices like smart phones. Does it really make sense for a wireless temperature sensor ? The power of Chirp is that it minimizes the resources needed _at the end point_. That's why I think it matters.

Francis 

 

Sheetal.Pandey
User Rank
Author
Re: Minimalism
Sheetal.Pandey   9/30/2014 3:06:08 AM
NO RATINGS
IoT specifications and needs are different. Also there will lot of data from sensors. It needs a much simpler protocol.

prabhakar_deosthali
User Rank
Author
re:
prabhakar_deosthali   9/30/2014 2:39:16 AM
NO RATINGS
In my opinion the devices in the classification -Internet of Things- do not need a global reach , they will basically be a cluster of local devices served/monitored by a local server , the server itself being part of the global internet.

So it is something like LAN and if some protocol is able to truncate some of the packet information ( by padding zeros whereever required for processing) then the problem could be solved.

WW Thinker
User Rank
Author
Re: Minimalism
WW Thinker   9/30/2014 1:29:14 AM
NO RATINGS
In honest, the Chirp serves nothing but framentates the commonly used IP protocol.  A custom-made protocol may sound interesting at first sight.  But, I am sorry to say that we don't need another protocol unless it has convinncing merits.  At the end of the day, we don't need elegant solution.  A workable solution which can be understood and ever epxerience by engineers universally will be the best.  In conclusion, let's stick with IPv4 or IPv6 or ..., but definitely not Chirp.

WW Thinker
User Rank
Author
Re: Minimalism
WW Thinker   9/30/2014 1:29:12 AM
NO RATINGS
In honest, the Chirp serves nothing but framentates the commonly used IP protocol.  A custom-made protocol may sound interesting at first sight.  But, I am sorry to say that we don't need another protocol unless it has convinncing merits.  At the end of the day, we don't need elegant solution.  A workable solution which can be understood and ever epxerience by engineers universally will be the best.  In conclusion, let's stick with IPv4 or IPv6 or ..., but definitely not Chirp.

LarryM99
User Rank
Author
Minimalism
LarryM99   9/29/2014 3:42:58 PM
NO RATINGS
I understand the allure of a minimalistic approach to IoT devices, but consider the longer-term and wider range aspects of IoT rather than the history of networked sensors. IoT devices will need to participate in arbitrary network configurations via ad hoc routing protocols. They may also need to qualify the access rights of a data request. A 40-byte IPv6 header is clearly not necessary, given the multiple compression techniques such as 6LOWPAN which adapt full network capability to low-power sensor networks. The software stacks for these protocols would have been a real stretch for earlier controllers, but they are well within the capability of current ones.

Larry M.

zeeglen
User Rank
Author
Re: The tweet of IoT
zeeglen   9/29/2014 12:53:29 PM
NO RATINGS
The technical term "Chirp" has already been taken long ago to describe a frequency shift in a signal.  Rather than yet another re-use of a technical term that will generate confusion due to multiple meanings, may I suggest something slightly different, such as "Chirrup" or even "Twit" or "Twyte"? 

Page 1 / 2   >   >>
Most Recent Comments
rick merritt
 
Tim R Johnson
 
ewertz
 
antedeluvian
 
ewertz
 
perl_geek
 
R_Colin_Johnson
 
perl_geek
 
R_Colin_Johnson
Most Recent Messages
11/19/2017
3:31:42 PM
Like Us on Facebook
EE Times on Twitter
EE Times Twitter Feed