Brett Lee
January 2003
Muchos Gracias to James Houser
=============================================================================


		**************************************
		***   Everything you wanted to     ***
		***   know about dBm but were      ***
		***        afraid to ask.          ***
		**************************************



It's all about POWER !!!!!  


INTRODUCTION

Below is information compiled from listening to James presentation
on dBm and stuff like that.


DEFINITIONS

f	Frequency	# of cycles in 1 second (measured in Hz)
Hz	Hertz		One cycle per second
T	Period		1/f
~\	Wavelength	v/f, where v = velocity ~ speed of light
dBm	Decibels < 1 mW	This is a measurement of power loss with 
			1 milliwatt as the transmission reference.
P	Power Density	watts/m^2, where m^2 is square meters
W	Watt		I*V, where I = current and V = voltage


1.  FCC regulations at 2.4 GHz DSSS - Gubment numbers
----------------------------------

FCC maximum power input INTO of our antenna's is 1 watt.
	(termed insertion power - good name)
	No more than 1 watt out the intentional radiator.
FCC maximum power output OUT of our 6 dBi antenna's is 4 watts.
	(termed passive radiation)

As you can guess, "somehow" the power increases.  How, I dunno.

On another point, however, 30 dBm is equal to 1 watt.
(Note: dBm can be negative, but wattage cannot - see Power Chart)

A exponential relationship exists between dBm and wattage, in that an
	increase in dBm by 3 doubles (roughly) the wattage.
	Therefore, 33 dBm is 2 watts, and 36 dBm is 4 watts.
	(Actually, an increase of 10 dBm increases the wattage by a 
		factor of 10, therefore, 40 dBm is 10 watts.)
So, the maximum antenna power output as governed by the FCC is 36 dBm.

*** The caveat ***

The power that actually enters the antenna cannot be greater than 30 dBm,
but it can be less.  And, for every dBm less on the input side, we can
go 1 dBm more on the output side.  So, if 20 dBm goes in, 46 dBm can go out.

A practical way to look at this is for every 3 dBi you increase the gain 
of the antenna, you have to decrease the power into it by 1 dBm.

So, More power to ya !!!


2.  Loss of Power - it happens to all of us...
----------------------------------------------------

Power is lost due to a variety of reasons.  I think this loss is called
attenuation.  If so, then attenuation can be attributed to cabling, 
connectors, travel distance of the signal and more.  Lets investigate.

A general formula for 2.4 GHz signal loss follows:

	Signal loss = 96.6 + 20log(f) + 20log(D) 

		where:	f= frequency 2.4 GHz, and
		D=distance in miles

Here's some discussion on the factors that go into this formula:

A.  LMR cable:

LMR400 cable loses 6.8 dBm for every 100 feet of cable length.  So, if the
signal travels 200 feet between the radio and the antenna, its gonna 
lose 13.6 dBm.  The signal loss for LMR6000 cable is less at 4.42 dBm.

B.  Connectors:

Depending on the type of connectors used, the signal loss will vary.  
Currently, the 4 connectors used are assumed to have a total loss of 1.5 dBm.

C.  Airborne:

As soon as the signal leaves the feeder horn and enters the atmosphere, it
takes a big hit!  The loss is 96.6 dBm.  Harsh.

D.  Distance it travels.

The further the distance, the more the signal fades.  Dunno how much.

E.  Flat fading / Rayleigh fading

- Flat fading (sometimes referred to as Rayleigh fading) occurs in narrowband 
	systems.  This is due to refraction.  
- In wideband (2.4, 5.8, ...) systems, fading can be flat OR selective.
- Then, of course, there is Multipath fading.  Adaptive equalization 
	techniques are used to overcome this.

F.  Atmosphere

What about rain and snow and dark of night?

- H2O and O2 are the main factors in absorbing electromagnetic (EM) energy.

	- O2 resonance occurs at 60GHz 
		- that's pretty high for us to worry about

	- H2O resonance occurs at 23GHz
		- below 10 GHz, the effect is insignificant
		- below 5 GHz, the effect is negligable

		- yes, rain does cause the signal to scatter, but once
			again, the effects are negligable

		- mist, fog, snow and dust are negligable as compared
			to rain attenuation


3.  Increase - Take back that power !!!
---------------------------------------------

Strength can be added to the signal.  This can be done by powered amplifiers
and by antenna's.  Powered amplifiers somehow take electricity and magnify
some part of the signal (we better ask Rick about this one).  Antenna's 
add power (signal strength) by focusing the signal in a narrower beam.

A.  Gain:

The gain of an antenna is a measure of the amount of dBm it increases the
signal by.  For example, a 6 dBm antenna would add 6 dBm to the signal, 
whereas a 24 dBm antenna would add 24 dBm.

B.  Our Antennas

	Omni				6 & 9 dBm	$150-$200
	Patch Panel			8 & 16 dBm	$200-$300
	Grid				18 & 24 dBm	$180-$220
	High Performance Parabolic	21 dBm		$900-$1800
	Sectorial			16 dBm		$500-$1000


C.  Amplifiers

	A 9000 amplifier expects about 10-13 dBm and always adds 11 dBm.
	A shelf mount amplifier tries to put out a fixed amount of power
		and therefore requires a fixed amount of power.

4.  How much is enough
--------------------------

How much power can we really use.  I mean, if you could have all you wanted,
would you want a hundred kazillion bigowatts?  For this, we need some, yes,
sensitivity training.


	Receive Sensitivity		Data Transfer Rate
	---------------------		--------------------
	-82 dBm				11 mb/s
	-85 dBm				5  mb/s
	-89 dBm				2  mb/s
	-91 dBm				1  mb/s


So, as you can see, we would like to receive at least -82 dBm.  But, with
tolerances and differences, a safety margin (aka. fade margin) of 10 dBm
is built in.  Therefore, our systems are designed to receive -72 dBm to 
the radio.


5.  Calculations 
----------------------

Ok math majors, here's a problem for you:  Figure out how much power (in dBm)
would make it from station A to station B.

Station A                                         Station B



    //      < ---------  signal  -------------->      //
    /                                                  /
    /                                                  /
------------                                      ------------
|          |                                      |          |
|          |                                      |          |
------------                                      ------------

Radio: 1 dBm                                         1 dBm
Cable: 100 m                                         10 m
Antenna:  Grid, gain=24				     Grid, gain=24


Would this support 11 mb/s ?


6.  Artifacts 
---------------------

A.  A radio signal has "zones" around it called Fresnal (pronounced "Frey-nel") 
zones. *1  These zones are important to the functioning of the radio 
communication.  For example, just because you "see" the other tower, it is not
enough to actually make a connection.

The TOTAL blockage of the of the fresnal zones cannot be more than about 40%.
That means that if the upper part of the fresnal zone is unblocked (as is the 
case with most situations) then the lower part of the zone (which represents
half of the total zone) needs to be less than 80% blocked.

*1.  See Huygen's (pronounced "Hoy-gen") for the fascinating discussion about
     how a wavefront is in reality an infinite number of wavelets.  Or maybe NOT.


B.  Side Lobes:

Side lobes are Fresnal zones outside of the main lobe.  We design our radios to 
ONLY work with the main lobe.  You can get lucky and actually work with one of 
the others, but they are NOT supported.

Side lobes are lobes which work in a power range -20 dBm LESS than the main lobe,
so they are possible to detect via signal monitoring.

Antennas are not perfect, so side lobes occur in varying places.  Can't trust em.


C.  Back Lobes:

Signals have a back lobe as well.  So, when putting two antennas back to back,
it is best to put them at different heights.  If this is not possible, use
different channels or put at least 10 feet between them


D.  Channels

In the 2.4 GHz frequency there can be defined 11 channels (see Wave Wireless
Conversion Table for Up-Converters).  Since we use at 22 GHz spread spectrum 
the signal overlaps into other channels.  For example, Channel 1 is at 2.412,
but a 22 GHz spread spectrum means that 11 GHz is on each side of the middle
of the channel (namely 2.412).  So, Channel 1 actually covers from 2.401 to
2.423.  The next channel up that won't interfere with 1 is 6, and after that
comes 11.


E.  Cross Polarization:

Antennas can be polarized vertically or horizontally.  For that matter, I don't
know why they can't be polarized something in the middle, but that wasn't brought
up.

At any rate, most cell phone antennas operating in the 2.4 GHz range are 
vertically polarized.  No, this is not the same thing as vertically challenged, 
but it can be challenging as they can monopolize all the signal in that range.

For some reason, changing the polarization from vertical to horizontal opens up
a whole new playing field, and even though the vertical is full, the horizontal
has plenty of room.  

The moral of the story - try to get horizontal when you polarize.