Full Duplex Wifi Simulation

[dabench]

I'm trying to simulate a full duplex wireless network for a school project. Do you think it would be a good approximation if I had the following setup: Dual Band Router that has a built in NAS and two laptops. Laptop A connected at 2.4ghz writing a large file to to the NAS and laptop B connected to 5ghz reading a differnt large file from the NAS. I know it wouldn't be a perfect simulation but the router would almost be doing "full" duplex because it would be receiving data on the 2.4ghz network while simultaneously transmitting data on the 5ghz network. Please let me know what you think.

 

[theevilsharpie]

It's hard to tell if this meets your needs without knowing the requirements of your school project.

 

[dabench]

It's hard to tell if this meets your needs without knowing the requirements of your school project.

There are no requirements, its for a research paper, so I get to make the requirments.

 

[DanFungus]

FYI I was getting ~5MBps throughput over wireless OR wired when I had an external hard drive hooked up to my router with USB port/NAS. With a separate ethernet-based NAS box I get 60-90MBps sustained over wired, and 20-30MBps sustained over wireless. The router is set up for 802.11n at 5GHz and gigabit wired. In short, you will use a fraction of the overall bandwidth if you use the built-in USB port port for NAS - not sure if that matters for your experiment.

If you want a *real* simulation, check out NS2 (free), NS3 (free), or QualNet (not free). Although there is a pretty big learning curve for these guys.

 

[dabench]

FYI I was getting ~5MBps throughput over wireless OR wired when I had an external hard drive hooked up to my router with USB port/NAS. With a separate ethernet-based NAS box I get 60-90MBps sustained over wired, and 20-30MBps sustained over wireless. The router is set up for 802.11n at 5GHz and gigabit wired. In short, you will use a fraction of the overall bandwidth if you use the built-in USB port port for NAS - not sure if that matters for your experiment.

If you want a *real* simulation, check out NS2 (free), NS3 (free), or QualNet (not free). Although there is a pretty big learning curve for these guys.

Thanks. I'm already planning to use NS2 for windows as a software simulation but I also want to do a real world type experiment.

 

[imagoon]

Not sure what why you are simulating but the tech already exists. Couple of Dragonwaves can do do 50+MBps / full duplex.

http://www.dragonwaveinc.com/products.asp

Most home / office 802.11(something) techs don't do it because people don't want to pay for it.

 

[drebo]

Wireless is a shared media access network. It is not full duplex, and never will be.

In your scenario, each radio is still operating at half duplex.

Full duplex would imply that one radio was sending and one radio was receiving to the same client.

Basing your "research paper" on a faulty premise doesn't seem like a very good idea.

There are point-to-point radios (generally extremely expensive) that do full duplex, as pointed out above. These operate by using a channel width twice the size of standard wireless and sending in one band while at the same time receiving in another band. So, where your consumer gear uses a 40mhz channel to achieve 300mbps half duplex, the Dragonwave gear would use an 80mhz channel to achieve the same base bandwidth in both directions. The encoding type is generally different, to ensure better reliability rather than better speed, and so you'll generally see around 100mbps out of an 80mhz channel...depending on dish size and frequency. A 6ghz link using 6ft dishes over a couple of miles will be much faster than an 11ghz link using 2ft dishes over 20 miles. Both could be full duplex, though.