How to simulate WSN with NS-3

[ram...@gmail.com]

I want to simulate WSN with NS-3. The network is formed by n sensor nodes and one sink in a hierarchical architecture. 

I want to draw a curve: throughput= f (number of nodes) 

I am starting to use NS-3. Thank you for your support

[Rohini SharmaOhlan]

use leach protocol to get an idea how sensor nodes works?

--
You received this message because you are subscribed to the Google Groups "ns-3-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ns-3-users+...@googlegroups.com.
To post to this group, send email to ns-3-...@googlegroups.com.
Visit this group at http://groups.google.com/group/ns-3-users.
For more options, visit https://groups.google.com/groups/opt_out.

[Tommaso Pecorella]

How to get two birds with a stone.

LEACH protocol. What's this ? I mean, I perfectly know what it is, the point is... it's paper-ware. Good to write a scientific paper, but no practical standard is using it. Enough said.

About the original question. Analysis of the question.

"The network is formed by n sensor nodes and one sink in a hierarchical architecture."

What do you mean by "hierarchical architecture" ?

802.15.4 could use multi-tree topology, but it's never used practically. Not even the "basic" tree one is used.

You didn't specify the MAC/PHY protocol you'd like to simulate, nor the IP-level topology. Nor if you need clustering or not.

Need more details on what's your intended topology, please.

Cheers,

T.

[ram...@gmail.com]

Thanks,

My network forms a tree. The sink is the root and the sensors are nodes in the tree. I chose this example to train using NS-3.

[Tommaso Pecorella]

Hi,

bad luck. The lr-wpan module implements 802.15.4, but only the ad-hoc part (plus, it's not yet merged). The IPv6 routing level might be a tree, but you'd need the RPL module, which is still unreleased.

By the way, it's a quite harsh way to train ns-3 on unreleased and on not-yet-merged modules... It'd suggest to train yourself with more "stable" stuff, and then move to the development parts.

However, if you're really into WSN, feel free to join the discussions about them on ns-dev mailing list. At the moment me and some other people are woking on improving the lr-wpan module to merge it as soon as possible in the main ns-3 branch.

Cheers,

T.

[ram...@gmail.com]

Thanks Tommaso,

In this case, I can use NS-2? 

What is the best for the WSN NS-2 or NS-3.

[Tommaso Pecorella]

Hi,

definitely ns-3. The lr-wpan module (although still being worked on) it's more precise and complete than the ns2 one. 6LoWPAN and IPv6 aren't in ns2 and RPL is not going to be implemented in ns2. Enough said.

Cheers,

T.

[Anas D.]

Hi,
i am also working on wireless sensor networks. but a novice in ns3 like you.  i started using the simple-adhoc example using 802.11b standard. i will add more nodes and continue from there. 

[Tommaso Pecorella]

On Wednesday, February 12, 2014 3:43:32 PM UTC+1, ram...@gmail.com wrote:

Thanks,

How can I specify a hierarchical topology in the simulation with NS-3? 

There is an implementation of LEACH or HEED protocol with NS-3?

Hi,

no, there's no LEACH or HEED implementation in ns-3. And, about the topology, it depends on what layer you're looking at. MAC or IP ?

Cheers,

T.

[Fernaz Nur]

Hi!! I am very much new in ns3. I have just started understanding the example programs. I want to simulate a tree based WSN and simulate my MAC protocol there. How can I change MAC protocol in wifi-adhoc program?

[Tommaso Pecorella]

Hi,

please study the Wi-Fi module documentation. The module is very complex and it is not possible to "simply" change the MAC protocol without understanding the relationships with all the other module components.

Good luck,

T.

[Vladislav Efremov]

Hi Tommaso,

I'm developing wsn uwb specific model for simulation. 

I took lr-wpan module as a base and I changed phy and mac levels, error model, spectrum  etc code in a way to make what I need. 

But its still LrWpanDevices.

Now I want to create a grid example where I have several nodes and the first node sends packets to the last node (not broadcast). Experiments with just two nodes is not interesting anymore)

And I a little bit confused here because I need a kind of routing protocol which operates on MAC level, I don't want to deal with Network Layer (or even higher) at all and want to avoid it.

I want use CSMA and some routing protocol on MAC level and implement a simple grid example.

Could you please advise is it possible and how?

Ideally in the future I want to compare behavior using TDMA and CSMA mechanism for my network on MAC layer, but it's another question...

[Tommaso Pecorella]

Hi Vladislav,

this could have been a new topic... but it's ok.

I'd suggest to check ISA100 for hints (there's no ns-3 model, but the standard can give you some ideas).

You will need for sure a way to propagate and handle the routing tables (even if they're based on MAC addresses, they're still routing tables) and so on. Basically you need a L2 routing protocol. Unfortunately I work at L3 (IP), so I can't suggest you one in particular.

Once you have the routing tables, you can either complete the work at MAC level, or use 6LoWPAN MESH headers to handle the multi hop forwarding. They're not implemented, but it should be trivial to do.

Anyway, for sure it is possible, but you'll need to implement the routing at L2. Possible, but not avoidable.

Hope this helps,

T.

[Tann Lybear]

regard to WSN with NS-3 discussion ,  I am also interesting.
Sir , could you give advice or recommend  if I want work on MAC protocol with grouping method , How can enable AP to separate STA into each groups ?

Thanks in advance .

Best regards,
libea

[ram...@gmail.com]

[Tommaso Pecorella]

Hi,

if you have an AP and some STAs, you're not working with a WSN. Please start a new thread.

Thanks,

T.

[Jayasree Sengupta]

In NS-3, can I simulate multihop routing in WSN environment ?

[Tommaso Pecorella]

Read the posting guidelines, PLEASE.

T.