Forum

Should we begin to differentiate between the Physical Layer and the RF Layer and add the RF layer as another layer to our modeled thinking? The RF Layer is entirely outside the scope of the current OSI model. We often refer to antennas, RF cabling, even RF behaviors as "physical layer" issues, though they are not so. The physical layer prepares the upper layer payloads for transmission onto the carrier medium. RF cabling, antennas, air, etc. are part of the carrier medium. Obviously, these things affect the physical layer, but they also affect all other layers.

Maybe this reference is not all that common, but I do hear it every now and again. Thoughts?

In the wired world, I've always called the cable part of PHY. If I'm wrong, then oops, maybe you are on to something.

If the wire is part of the PHY like I have always understood, then RF is just part of the PHY IMO.

GT

This is an interesting one. In many popular books etc, you will read of the physical layer in the OSI model as being concerned with "connectors, timing, bits and cabling". However if we look at typical diagrams showing the interconnection of two "stacks" as per the ISO model, you often see a "physical link" connecting the two, almost implying that the actual media is independent [ to a degree ] of the physical layer. Even in 802.11 we have the Physical Medium DEPENDENT layer. I.e. It DEPENDS on the physical medium. Even this is a bit confusing. For example, the normal medium in 802.11 is air. However, 802.11 works just as well [ better in many cases in a vacuum - space...no don't bring up .11n !!]. Howver, if we wanted to, we could put two systems a few feet apart in a container of Jello and we would still get communication. The PMD in all cases has done the same thing...no change of bit patterns, yet the medium has changed in between, so it should really be called the Physical Medium Dependent [ To a degree !! ] layer.

It could be argued that the RF medium is really not actually part of the "802.11 stack" but merely the interconnecting link between the two "layered cakes".

My brain hurts......

Dave

http://farm4.static.flickr.com/3036/2911722148_f972dfa5ae.jpg

We can verify whether a radio performs up to the 802.11 specs using various pieces of equipment. Companies such as Anritsu, Rhode and Schwarz, Keithley, Veriwave, etc. have a good selection of all-in-one and expensive, gear to do this.

The more hard core RF engineers use their Agilent, and other discreet scopes, power meters, and Network (pre LAN technology) Analysers, etc. to do the same thing (and more).

With equipment like this we can quantitatively answer questions about power output and sensitivity as defined in the spec.

However, by themselves they don't answer questions about design, physical alignment, assembly practices, and the like.

For example, why does a radio work wonderfully when connected directly (a conducted measurement), and then terribly when the same signals are sent through a wonderfully designed antenna (a radiated measurement).

We don't usually hear about RF cable routing for example. Whether a cable approaches the antenna board at a 90 degree angle or whether it lazily gets there on some haphazard path determined during final assembly which affects the ground plane and therfore the antenna pattern.
Cable loss and length are the only things usually discussed. But factors such as this can make a really big difference in the ultimate performance of a device. Other than talking about tight and waterproof connectors, we don't usually see much about these issues.

I would definitely be in favor of separating out RF design issues and speaking to them directly, but I honestly don't know whether to call it another layer.

Relating this to the latest Ipad discussions, I dont think you'll find it's reported problems related to it's physical (ie antenna) layout.    More likely it would be the actual design of its antenna, or the quality of its radio.

It's still hard to tell from the pictures I've seen how its radio, for example, sits in proximity to its other parts.     I need to take one apart!

If I should win one (ha ha) I let you know what I find out.

Sorry to me all of the above is still considered the Physical Layer of the OSI Model.

Basically it is the equivalent of the Cable or even of the CSU/DSU but not really much higher.

physical layer is the main layer in wireless ase need to know how packets and signals send and modulated and add spicific information to it