This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

6LowPan someone ?

Other Parts Discussed in Thread: CC2430, Z-STACK, MSP430F1611, MSP430F5438

Hi,

Anyone working on a 6LowPan Stack or design like LPWRocks asked for Wireless HART ?

JP

  • Yes, Radiocrafts and Sensinode offer 6LoWPAN modules. Check out RC2300 at www.radiocrafts.com and the stack at www.sensinode.com

     

  • Yeah, but what I am seeking is comment about this stack. I was interested about size, complexity, performance, ... versus the Zigbee one but on the TI's chips.

    But thanks anyway !

  • Hi JP,

    We have had someone take a look at this earlier.

    Pro: 

    • TCP/IP style interface, so no need to reinvent the wheel. Great for developers who are familiar with BSD sockets/WinSock etc.
    • Programming-model: event-driven like Z-Stack but main() is located in the application. This makes applications more intuitive.
    • Everything is open source
    • Free development tools, (Example FreeRTOS: www.freertos.org )

    Areas for improvement:

    • Need to improve documentation.
    • Need to improve on the comments in the code

    LPWRocks

  • Is it lighter than Zigbee ?

    The features are comparable ?

    Does-it perform well into an existing IP network ?

     

    Noone have a reliable reference, a website or any scientific paper which present clearly the benefits (and drawbacks) of the stack compared to the others (many, many others !) ?

  • The 6lowpan stack is smaller (as small as 24K) and significantly less complex that Zigbee.  Additionally becuase it doens't have to use all of the complex mac constructs it can perform better.

  •  JP, it is lighter.  Don't expect to run zigbee on anything less than a 64K part and more probably a 128K part.

    Zigbee does has a define security architecture (such as it is - Not good) and does have application profiles defined.  6lowpan does not.

    All of the implementations of 6lowpan that are available today play very well with existing IP networks!  this is the point.   With Zigbee you need complex gateways to translate zigbee to something that

    the internet can understand which then breaks end to end security and end to end reliability.  6lowpan you don't need a gateway, just a simple bridge or multiple bridges (which zigbee can't do either.)

     

  • Hi,

    got some questions:

    How does routing work with 6LowPan? Is there a automatic Self Healing / Route-Discovery like in ZigBee (AODV)?

    Or is it more static (manuell setup)?

     regards,

    OZ

     

  •  HI,

    As you state each standard protocols has its nice feature. I guess the challenge and what everyone is searching for is the "killer" application where the protocol of choice can make the perfect fit.

    What is the killer app for 6LowPan?

    What about the IPv6 packet overhead on top of IEEE 802.15.4?

    BTW: I have found this overview: 

    http://www.archrock.com/downloads/resources/6LoWPAN-tutorial.pdf

    LPWRock 

  •  6lowpan doesn't constrain you to any particular routing protocol.  With 6lowpan you can choose to use a mesh under (layer 2 routing) or a route over (layer 3 routing).  It is your choice that would best fit your needs.  Unlike zigbee you are not locked to any particular protocol.

     There are pros and cons to using mesh under or route over.  Generally with mesh under in order to ensure intreoperability every node in the pan is required to use the same mesh routing protocol (this is the contraint with zigbee), but with mesh under the higher layer architecture is really simple.  The layer 2 mesh makes the many nodes and multiple hops look like a single (albiet slow and lossy) broadcast network.  To IP every node looks like they are on the same subnet.  With mesh under unless you talk the same layer 2 mesh you can't exchange data packets with any other nodes.

    With route over the various nodes could  choose to use different routing protocols and yet every node can exchange packets.  This is the way the internet works today.  there are many routing protocol in use and still nodes can exchange packets because they are just IP packets.  The upside is interoperability.  The down side is that the routing or network topology is not as simple.  This is why there is a new working group (ROLL) working on just this topic.

    One minor nit - Zigbee does not use AODV.  They use their own modified version of AODV - unfortunately.

    With 6lowpan you can build self organizing and self healing multi-hop networks with either mesh under or route over and you can also choose to create manually set up networks.

     

     

  • I think the killer point for 6lowpan is that you don't need to learn new protocols, new apis, build new protocols and new gateways and increase complexity.  6lowpan allows you to build IP networks on top of 15.4 and easily interconnect them to existing IP infrastructure, manage the nodes and network with existing tools and knowledge, protect the networks with existing tools, program the networks with existing experience and knowledge.

     Additionally by using 6lowpan on 15.4 you can build sensor and control networks that make use of multiple phys as needed.  If you need to talk over PLC or 15.4 or even narrow band RF you can do it.  If you want to interconnect these networks with ethernet or wifi, just do it.  You don't need to build tunnels, you don't need to build gateways to translate to and from ip and you don't need to program end nodes to understand proprietary applications (as with Zigbee).

    As for packet overhead, for intra-pan packets the overhead can be as low as 4 bytes (single hop, no fragmentation) and if you are using a layer 2 mesh you would add 4 bytes and if you are sending large packets you would add a couple bytes for the fragmentation header.  With 6lowpan's header design you only "pay for" what you use.  In a route over design the overhead is generally only 4 bytes (much smaller than zigbee's monlithic header).

     

     

  •  Hi 6lp,

    I am new to 6LowPan, but was wondering where to start. Any Open 6LowPan Stack you recommend to use with a TI transceiver to SoC?

    What type of logical devices do you have with 6LowPan? If you compare a ZigBee Coordinator, Router, or Ende Device

    How big is the 6LowPan API? How many function calls for the application layer to setup a network?  

    Hobbyist  

  • Hi 6lp,

    A few more questions. Can 6LowPan be used with SP100 a or b or something? Is 6LowPan made to be used for industrial applications?

    What about sleeping routers? Can this be supported by 6LowPan. ZigBee does not support this. 

    Hobbyist 

  • Actually 6lowpan is part of the ISA100 (SP100) standard.  SP100 uses an SP100 layer 2 self forming self healing mesh and 6lowpan on top of the mesh including IP and UDP.

    SP100 uses and support sleeping routers so, yes you can build biuld a 6lowpan network that supports sleeping routers.

     

     

  • www.6lowpan.org/

    www.archrock.com  (US)

    www.sensinode.com (Finland)

     

    EngiNerd

  •  

    When I was reading materials available on internet I came to know that Zigbee 

    provides support for sleeping end device. But, nothing related to sleeping end 

    device in case of 6LowPAN was mentioned. I want to know whether 6LowPAN 

    supports sleeping end device or not.

     

  • Most of the 6lowpan implementations support sleeping end-nodes and some support sleeping routers.

     

  • Check out Contiki community:

    www.sics.se/contiki

    LPRF Rocks the World

     

  • Is 6LowPAN compatible with MSP5438 chip?

  • Hello,

    You can port the Contiki OS with the uipv6 used on the sky mote using an MSP430F1611 I believe to your MSP430F5438. See URL below for this open source community:

    http://www.sics.se/contiki/download.html

    http://www.sics.se/contiki/uipv6-faq.html

    LPRF Rocks the World