Wireless communication and frequency band

Hello,

I have a question pertaining to the frequency bands of carriers in
wireless communication.

So, when a two systems use two carriers of two different frequencies, say
2.5 GHz and 5 GHz then I would expect that overall communication in system
using 5 GHz would be faster as compared to system using 2.5 GHz. Of
course, how fast is this would depend on where the bottleneck is in terms
of communication. That is, in communication channel, Baseband, MAC or
application.

Can someone please comment on this as none of the materials I have
referred to make even an oblique reference to this concept.

Thanks, 
---------------------------------------
Posted through http://www.DSPRelated.com

"Sharan123" <99077@DSPRelated> writes:

> Hello,
>
> I have a question pertaining to the frequency bands of carriers in
> wireless communication.
>
> So, when a two systems use two carriers of two different frequencies, say
> 2.5 GHz and 5 GHz then I would expect that overall communication in system
> using 5 GHz would be faster as compared to system using 2.5 GHz. Of
> course, how fast is this would depend on where the bottleneck is in terms
> of communication. That is, in communication channel, Baseband, MAC or
> application.
>
> Can someone please comment on this as none of the materials I have
> referred to make even an oblique reference to this concept.

Hello,

That is not correct. The carrier frequency used has nothing to do (in
general) with the data rate, assuming "faster" means a higher data rate.

What DOES affect data rate is the bandwidth that is utilized. Also
the power level, or SNR, of the signal affects data rate, both
the maximum possible theoretical error-free data rate and also
the practical data rate a practical system can have.

In fact one of the most important relationships in digital
communications is Claude Shannon's capacity theorem, which says the
maximum possible error-free data rate C (for the "channel capacity") is
a function of the bandwidth B, signal power P, and noise power N:

  C = B * log_2[ (P + N) / P]

Now I qualified at the beginning with "in general" because (in a real
system) the carrier frequency f does limit the maximum possible
bandwidth to B <= 2*f, but for systems transmitted over an RF channel, f
is usually much, much higher than this constraint requires. This limit
comes into play in, e.g., telephone line modems which are constrainted
to use the standard POTS bandwidth of 4 kHz (or thereabouts).
-- 
Randy Yates
Digital Signal Labs
http://www.digitalsignallabs.com

Randy Yates <yates@digitalsignallabs.com> writes:

>   C = B * log_2[ (P + N) / P]

Correction:

  C = B * log_2[ (P + N) / N]
  
-- 
Randy Yates
Digital Signal Labs
http://www.digitalsignallabs.com

On Fri, 28 Aug 2015 14:05:43 -0400, Randy Yates wrote:

> "Sharan123" <99077@DSPRelated> writes:
> 
>> Hello,
>>
>> I have a question pertaining to the frequency bands of carriers in
>> wireless communication.
>>
>> So, when a two systems use two carriers of two different frequencies,
>> say 2.5 GHz and 5 GHz then I would expect that overall communication in
>> system using 5 GHz would be faster as compared to system using 2.5 GHz.
>> Of course, how fast is this would depend on where the bottleneck is in
>> terms of communication. That is, in communication channel, Baseband,
>> MAC or application.
>>
>> Can someone please comment on this as none of the materials I have
>> referred to make even an oblique reference to this concept.
> 
> Hello,
> 
> That is not correct. The carrier frequency used has nothing to do (in
> general) with the data rate, assuming "faster" means a higher data rate.
> 
> What DOES affect data rate is the bandwidth that is utilized. Also the
> power level, or SNR, of the signal affects data rate, both the maximum
> possible theoretical error-free data rate and also the practical data
> rate a practical system can have.
> 
> In fact one of the most important relationships in digital
> communications is Claude Shannon's capacity theorem, which says the
> maximum possible error-free data rate C (for the "channel capacity") is
> a function of the bandwidth B, signal power P, and noise power N:
> 
>   C = B * log_2[ (P + N) / P]
> 
> Now I qualified at the beginning with "in general" because (in a real
> system) the carrier frequency f does limit the maximum possible
> bandwidth to B <= 2*f, but for systems transmitted over an RF channel, f
> is usually much, much higher than this constraint requires. This limit
> comes into play in, e.g., telephone line modems which are constrainted
> to use the standard POTS bandwidth of 4 kHz (or thereabouts).

About the only relationship between carrier frequency and speed is that 
-- all else being equal -- a higher carrier frequency makes it _easier_ 
to have a wider signal bandwidth.  For conventional narrowband 
communications (i.e., ones that use a signal bandwidth as narrow as can 
be to match the signal) higher carrier frequencies can make it harder to 
transmit a clean signal -- but most of the current standards for GHz 
radios are spread spectrum, which neatly sidestep that issue.

For example, AM radio has a bandwidth of maybe 10kHz (someone will 
correct me with an exact number).  The normal AM radio broadcast band is 
around 500Hz, but you can transmit AM just fine at 120MHz, and people 
still do (if I have my numbers right, that's aeronautical mobile).

Conversely, wideband, broadcast FM voice radio has a signal bandwidth of 
something like 100kHz.  The normal FM broadcast band is up around 
100MHz.  You can transmit your 10kHz AM signal up there _just fine_ -- 
but if you tried sending a 100kHz wide signal centered around 500kHz 
you'd have all sorts of difficulties just with setting up an antenna that 
could handle the bandwidth, much less with building a decent receiver and 
with the frequency-selective properties of the atmosphere.

-- 

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Tim Wescott wrote:
> [snip]
>
> For example, AM radio has a bandwidth of maybe 10kHz (someone will
> correct me with an exact number).  The normal AM radio broadcast band is
> around 500Hz,

ERRR, suspect a missing k for kilo ;/


>  but you can transmit AM just fine at 120MHz, and people
> still do (if I have my numbers right, that's aeronautical mobile).
> [snip]

Tim Wescott <seemywebsite@myfooter.really> wrote:
> On Fri, 28 Aug 2015 14:05:43 -0400, Randy Yates wrote:
>> "Sharan123" <99077@DSPRelated> writes:

(snip)

>>> So, when a two systems use two carriers of two different frequencies,
>>> say 2.5 GHz and 5 GHz then I would expect that overall communication in
>>> system using 5 GHz would be faster as compared to system using 2.5 GHz.
>>> Of course, how fast is this would depend on where the bottleneck is in
>>> terms of communication. That is, in communication channel, Baseband,
>>> MAC or application.

>>> Can someone please comment on this as none of the materials I have
>>> referred to make even an oblique reference to this concept.
 
>> That is not correct. The carrier frequency used has nothing to do (in
>> general) with the data rate, assuming "faster" means a higher data rate.
 
>> What DOES affect data rate is the bandwidth that is utilized. Also the
>> power level, or SNR, of the signal affects data rate, both the maximum
>> possible theoretical error-free data rate and also the practical data
>> rate a practical system can have.
 
(snip)

> About the only relationship between carrier frequency and speed is that 
> -- all else being equal -- a higher carrier frequency makes it _easier_ 
> to have a wider signal bandwidth.  For conventional narrowband 
> communications (i.e., ones that use a signal bandwidth as narrow as can 
> be to match the signal) higher carrier frequencies can make it harder to 
> transmit a clean signal -- but most of the current standards for GHz 
> radios are spread spectrum, which neatly sidestep that issue.

Also, all else being equal, it is often cheaper to run at lower
frequencies.  Fast transistors are more expensive than slow ones.
(This is less true now than it used to be.)

So, if you don't need the bandwidth, it (used to be) usual not
to go to a high carrier frequency.  (Not to mention the complications
of government allocation of frequency bands.)

> For example, AM radio has a bandwidth of maybe 10kHz (someone will 
> correct me with an exact number).  The normal AM radio broadcast band is 
> around 500Hz, but you can transmit AM just fine at 120MHz, and people 
> still do (if I have my numbers right, that's aeronautical mobile).

As someone noted, 540kHz to 1600kHz (or maybe 1700kHz now).
I believe AM allows for 10kHz spacing, usually with nearby
transmitters not on adjacent frequencies. AM double sideband,
so 5kHz on either side of the carrier.

> Conversely, wideband, broadcast FM voice radio has a signal bandwidth of 
> something like 100kHz.  The normal FM broadcast band is up around 
> 100MHz.  

The spacing is 200kHz. The carrier frequencies are odd multiples
of 100kHz. The bandwidth might be 150kHz or so. (Theoretically
it is infinite, with a long tail.)

> You can transmit your 10kHz AM signal up there _just fine_ -- 
> but if you tried sending a 100kHz wide signal centered around 500kHz 
> you'd have all sorts of difficulties just with setting up an antenna that 
> could handle the bandwidth, much less with building a decent receiver and 
> with the frequency-selective properties of the atmosphere.

Well, first the carrier has to be higher than the bandwidth.
(For most modulation methods.) 

In addition, it is a waste of a nice lower frequency band.

At the beginning of broadcast television, it wasn't so easy
to get to the higher frequencies, so they started with channel 2
at 54MHz to 60MHz (6 MHz spacing). 

Now that nice fast transistors are fairly cheap, (a big reason why
we have the cell phone system we have) high carrier frequencies
aren't so hard to use.  The analog cell phone system runs with
fairly narrow bandwidth in the 900MHz region. (What used to be
the higher UHF TV channels.) In this case, it isn't the individual
bandwidth but the total bandwidth of many cell channels.  

-- glen

On Fri, 28 Aug 2015 12:31:05 -0500, "Sharan123" <99077@DSPRelated>
wrote:

>Hello,
>
>I have a question pertaining to the frequency bands of carriers in
>wireless communication.
>
>So, when a two systems use two carriers of two different frequencies, say
>2.5 GHz and 5 GHz then I would expect that overall communication in system
>using 5 GHz would be faster as compared to system using 2.5 GHz. Of
>course, how fast is this would depend on where the bottleneck is in terms
>of communication. That is, in communication channel, Baseband, MAC or
>application.
>
>Can someone please comment on this as none of the materials I have
>referred to make even an oblique reference to this concept.
>
>Thanks, 
>---------------------------------------
>Posted through http://www.DSPRelated.com

What do you mean by "faster"?   More throughput?   Less delay?

Neither throughput or delay are dependent on frequency, so I'm not
sure what you're really asking.


Eric Jacobsen
Anchor Hill Communications
http://www.anchorhill.com

On Fri, 28 Aug 2015 14:04:15 -0500, Tim Wescott
<seemywebsite@myfooter.really> wrote:

>On Fri, 28 Aug 2015 14:05:43 -0400, Randy Yates wrote:
>
>> "Sharan123" <99077@DSPRelated> writes:
>> 
>>> Hello,
>>>
>>> I have a question pertaining to the frequency bands of carriers in
>>> wireless communication.
>>>
>>> So, when a two systems use two carriers of two different frequencies,
>>> say 2.5 GHz and 5 GHz then I would expect that overall communication in
>>> system using 5 GHz would be faster as compared to system using 2.5 GHz.
>>> Of course, how fast is this would depend on where the bottleneck is in
>>> terms of communication. That is, in communication channel, Baseband,
>>> MAC or application.
>>>
>>> Can someone please comment on this as none of the materials I have
>>> referred to make even an oblique reference to this concept.
>> 
>> Hello,
>> 
>> That is not correct. The carrier frequency used has nothing to do (in
>> general) with the data rate, assuming "faster" means a higher data rate.
>> 
>> What DOES affect data rate is the bandwidth that is utilized. Also the
>> power level, or SNR, of the signal affects data rate, both the maximum
>> possible theoretical error-free data rate and also the practical data
>> rate a practical system can have.
>> 
>> In fact one of the most important relationships in digital
>> communications is Claude Shannon's capacity theorem, which says the
>> maximum possible error-free data rate C (for the "channel capacity") is
>> a function of the bandwidth B, signal power P, and noise power N:
>> 
>>   C = B * log_2[ (P + N) / P]
>> 
>> Now I qualified at the beginning with "in general" because (in a real
>> system) the carrier frequency f does limit the maximum possible
>> bandwidth to B <= 2*f, but for systems transmitted over an RF channel, f
>> is usually much, much higher than this constraint requires. This limit
>> comes into play in, e.g., telephone line modems which are constrainted
>> to use the standard POTS bandwidth of 4 kHz (or thereabouts).
>
>About the only relationship between carrier frequency and speed is that 
>-- all else being equal -- a higher carrier frequency makes it _easier_ 
>to have a wider signal bandwidth.  For conventional narrowband 
>communications (i.e., ones that use a signal bandwidth as narrow as can 
>be to match the signal) higher carrier frequencies can make it harder to 
>transmit a clean signal -- but most of the current standards for GHz 
>radios are spread spectrum, which neatly sidestep that issue.

Eh, not really.   Satellite comm at any band, L-Band, C-Band, K-band,
are all basically the same single-carrier techniques.  There are some
instances where spreading is used, but the reasons aren't
frequency-related.

It's a little easier to get long range as frequency increases due to
gain of directional antennas, and other than various atmospheric or
physical propagation effects that's are frequency dependent, that's
about it.

>For example, AM radio has a bandwidth of maybe 10kHz (someone will 
>correct me with an exact number).  The normal AM radio broadcast band is 
>around 500Hz, but you can transmit AM just fine at 120MHz, and people 
>still do (if I have my numbers right, that's aeronautical mobile).
>
>Conversely, wideband, broadcast FM voice radio has a signal bandwidth of 
>something like 100kHz.  The normal FM broadcast band is up around 
>100MHz.  You can transmit your 10kHz AM signal up there _just fine_ -- 
>but if you tried sending a 100kHz wide signal centered around 500kHz 
>you'd have all sorts of difficulties just with setting up an antenna that 
>could handle the bandwidth, much less with building a decent receiver and 
>with the frequency-selective properties of the atmosphere.

It's not that bad, actually.   At 500kHz a ferrite rod antenna is a
good choice, and their BW is not too bad.   That's a good frequency
for underground communications and I've been playing around with stuff
down there for a while.

>
>-- 
>
>Tim Wescott
>Wescott Design Services
>http://www.wescottdesign.com

Eric Jacobsen
Anchor Hill Communications
http://www.anchorhill.com

Sharan123 wrote:
> Hello,
>
> I have a question pertaining to the frequency bands of carriers in
> wireless communication.
>
> So, when a two systems use two carriers of two different frequencies, say
> 2.5 GHz and 5 GHz then I would expect that overall communication in system
> using 5 GHz would be faster as compared to system using 2.5 GHz.

Nope.

> Of
> course, how fast is this would depend on where the bottleneck is in terms
> of communication. That is, in communication channel, Baseband, MAC or
> application.
>

Okay then.


See also "bandwwith" and how it relates to channel capacity.


then comes the latency from the hard work of decoding and marshalling on 
the recieving end.


> Can someone please comment on this as none of the materials I have
> referred to make even an oblique reference to this concept.
>

I know, right?

> Thanks,
> ---------------------------------------
> Posted through http://www.DSPRelated.com
>

-- 
Les Cargill

On Fri, 28 Aug 2015 15:02:45 -0500, Richard Owlett wrote:

> Tim Wescott wrote:
>> [snip]
>>
>> For example, AM radio has a bandwidth of maybe 10kHz (someone will
>> correct me with an exact number).  The normal AM radio broadcast band
>> is around 500Hz,
> 
> ERRR, suspect a missing k for kilo ;/

Indeed.  10kHz centered around 500Hz would be a trick, wouldn't it?

-- 

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com