FHSS versus DSSS, which is best?

Problems, experiences or just something to say about RC gear? Say it here.

Re: FHSS versus DSSS, which is best?

Postby errsemr » Sat May 29, 2010 9:57 pm

May be this http://www.spektrumrc.com/DSM/Technolog ... spx#calOrr is what Engineer is refering to?
errsemr
 
Posts: 23
Joined: Tue May 25, 2010 2:47 pm

Re: FHSS versus DSSS, which is best?

Postby Engineer » Sun May 30, 2010 3:57 am

One more time:

http://www.spektrumrc.com/Content/Im...SPMarticle.pdf

That's interesting. The link works from the Flying Giant forum but it wont work here.

http://www.spektrumrc.com/Content/Im...SPMarticle.pdf
Engineer
 
Posts: 14
Joined: Sat May 29, 2010 4:04 am

Re: FHSS versus DSSS, which is best?

Postby RCModelReviews » Sun May 30, 2010 5:13 am

I'm not sure I believe *anything* that comes from the Spektrum website or Horizon.

Why?

Well read this.
RCModelReviews.com, just the facts.
User avatar
RCModelReviews
 
Posts: 2120
Joined: Tue May 04, 2010 3:40 am

Re: FHSS versus DSSS, which is best?

Postby Engineer » Sun May 30, 2010 9:34 pm

I am sure Spektrum holds this site in equal regard.
Engineer
 
Posts: 14
Joined: Sat May 29, 2010 4:04 am

Re: FHSS versus DSSS, which is best?

Postby RCModelReviews » Sun May 30, 2010 10:22 pm

That's the great thing about the internet. People get to see all sides to a story.
RCModelReviews.com, just the facts.
User avatar
RCModelReviews
 
Posts: 2120
Joined: Tue May 04, 2010 3:40 am

Re: FHSS versus DSSS, which is best?

Postby errsemr » Mon May 31, 2010 8:03 am

Engineer wrote:One more time:

http://www.spektrumrc.com/Content/Im...SPMarticle.pdf

That's interesting. The link works from the Flying Giant forum but it wont work here.


The software on this board seems to remove part of your link. This should work:

http://www.spektrumrc.com/Content/Images/Products/calOrrSPMarticle.pdf

I have use the
Code: Select all
[url]
tagging as a workaround.

Let's hope this thread gets back to a pure technical discussion of spread spectrum merits.
errsemr
 
Posts: 23
Joined: Tue May 25, 2010 2:47 pm

Re: FHSS versus DSSS, which is best?

Postby Engineer » Mon May 31, 2010 8:24 pm

RCModelReviews wrote:
Engineer wrote:There seems to be a miss understanding about how 2.4 works.
A channel is not limited to just one system at a time. There can be many system on the same channel at a time. How many? A channel channel is 1 MHZ wide. If we were operating on 72mhz rules, and had separation of frequencies of 10kc there would be 100 separate frequencies to operate on.
In the testing done by Cal Orr there seems to be a limit of 16 systems before the range starts to degrade. Thus we could have 16 x 40 = 640 Spectrum DSSS radios happily operating at one time.

What you're referring to is temporal density.

Because 2.4GHz systems don't transmit all the time (like the old MHz systems did), you can "share" a single part of the band with other transmitters.

In theory, if you timed it just right, a transmitter which each transmitted for just 5% of the time could co-exist with 19 other transmitters of the same type -- so long as no two tried transmitting simultaneously.

In the computer networking world the same thing happens. After all, how else could you get a dozen or so computers all sharing the same WiFi connection?

These systems use what's called carrier sense multiple access (CSMA) which means that before any transmitter sends a signal, it listens to make sure that the channel is free and not already in use. If it hears another transmitter using the band, it waits for a random length of time (a few mS) and then listens again.

Obviously, if you're dealing with non-realtime or soft-realtime data like a file-transfer between two computers then this isn't a problem. However, with RC model planes there would be a problem.

Imagine if your model is on the back-side of a loop, heading towards the ground and you need to give it up elevator at exactly the right time to avoid hitting the ground. The last thing you want/need is a transmitter that listens to the part of the band it's using and courteously waits for it to be free before sending your "up elevator" command to the model.

So RC systems generally don't use CSMA to control access to the band.

This means that we can never really achieve the theoretical maximum number of users on a single part of the band.

So, in the case of the Spektrum, it has only a very limited ability to share the same part of the band with other Spektrum RC systems. The more Spektrum radios that try to share the same part of the band, the more "collisions" occur because one or more radios transmit at the same time.

Because the Spektrum radios use the same spreading code, two Spektrums transmitting on the same frequency at the same time *will* interfere with each other. Unlike an FM radio, this interference won't produce a glitch -- it will simply result in a loss of the data. Lose enough data and the receiver goes into lockout/failsafe mode.

As a result, although "in theory" you might be able to have as many as 16 Spektrum systems all coexisting on the same part of the band, it simply doesn't work that way in practice.

I'm unaware of Cal Orr's comments but I suspect he's based them on the way that CSMA networks work -- and RC isn't a CSMA network -- it doesn't look before it transmits so even two Spektrum radios on the same channel may produce lockouts, if they try to send too many packets of data at exactly the same time.

Interestingly enough, a Spektrum and a Corona (or other-brand DSSS system) are far *less* likely to interfere with each other when using the same part of the band. This is because they use different spreading codes and therefore can (to some extent) actually both transmit at the same time without losing too much data.

Spread Spectrum is a different beast to FM and there are many details that need to be considered.

However, bearing all this in mind, in an RC context, FHSS (or to be more accurate - constantly agile DSSS) does offer more resilience and a far more progressive degradation when trying to deal with very noisy environments.


The following is a copy of Wikipedia's explanation of DSSS

"Direct-sequence spread-spectrum transmissions multiply the data being transmitted by a "noise" signal. This noise signal is a pseudorandom sequence of 1 and −1 values, at a frequency much higher than that of the original signal, thereby spreading the energy of the original signal into a much wider band.

The resulting signal resembles white noise, like an audio recording of "static". However, this noise-like signal can be used to exactly reconstruct the original data at the receiving end, by multiplying it by the same pseudorandom sequence (because 1 × 1 = 1, and −1 × −1 = 1). This process, known as "de-spreading", mathematically constitutes a correlation of the transmitted PN sequence with the PN sequence that the receiver believes the transmitter is using.

For de-spreading to work correctly, the transmit and receive sequences must be synchronized. This requires the receiver to synchronize its sequence with the transmitter's sequence via some sort of timing search process."

Since two independent transmitters would not be synchronized, they would look white noise to one another and would not detect each other.
Engineer
 
Posts: 14
Joined: Sat May 29, 2010 4:04 am

Re: FHSS versus DSSS, which is best?

Postby RCModelReviews » Mon May 31, 2010 9:43 pm

Yes, the two do have to be synchronized -- but therein lies a problem... If the transmitters are using the same spreading code (as appears to be the case with the Spektrum), how does the receiver know which one to sync to?

Spreading works fine so long as the PN codes are different -- not so good when the PN codes are identical.

I was reading the stuff from Carl Orr on the Spektrum site -- he says "FOUR" spektrum transmitters on the same channel is about the limit -- which sounds feasible -- given a temporal density of 10%. This would mean about a 50% packet-loss due to collisions, which is an acceptable amount.
RCModelReviews.com, just the facts.
User avatar
RCModelReviews
 
Posts: 2120
Joined: Tue May 04, 2010 3:40 am

Re: FHSS versus DSSS, which is best?

Postby Engineer » Tue Jun 01, 2010 2:53 am

The receivers are synchronized to the transmitter during the binding process.
A PRN code is a long series of numbers that are repeated. Synchronizing them means getting them to start in the same position of the code. Two transmitters would never be in sync.

How many systems that can co-exist on one channel is a design choice probably controlled by the noise detector.
Eagletree claims more than 1000 of their systems can co-exist on the same channel. They use a random number for the PRN seed.
Engineer
 
Posts: 14
Joined: Sat May 29, 2010 4:04 am

Re: FHSS versus DSSS, which is best?

Postby JiB » Wed Jun 02, 2010 10:43 pm

Engineer wrote:The receivers are synchronized to the transmitter during the binding process.
A PRN code is a long series of numbers that are repeated. Synchronizing them means getting them to start in the same position of the code. Two transmitters would never be in sync[...]


Assuming similar specced crystals in each unit, the two prn streams would move in and out of sync, depending on the slight differences in frequency of their crystals. So a few times per second, those two prn streams would be in sync, meaning that the two radios would be interfering with each other, making reception harder/impossible. So yes, it is likely two same-tape transmitters would interfere because of using the same chirping sequence.
However, there is no need for the chirping sequence to be based directly on the output of a pseudo random number generator. Once bound, it could use the binding ID to step through the random stream (or XOR-ing the output with said ID). That would avoid such a in-synch problem. Once one cirping sequence is not equal to the other, the streams cannot synchronise, thus not recurring interference.
JiB
 
Posts: 11
Joined: Tue Jun 01, 2010 7:58 pm

PreviousNext

Return to Radios and Servos

Who is online

Users browsing this forum: No registered users and 6 guests