64APSK is that for real?

Thursday 13 June 2013
64APSK is that for real?


By Dave Suffys, self-proclaimed DVB-S2 evangelist.

With the advent of the extensions to the DVB-S2 standard comes among other improvements the addition of 64APSK as a modulation mode.

The question is: “is 64APSK usable?”

Over the last 15 years we have seen a tremendous evolution in satellite communication. Whereas in the early days of satellite, powers were low and antennas had to be very big (remember those beautiful 32 m type-A C-band antennas), nowadays 64APSK with reasonable antennas is within reach.

Imagine having to install a 2.4 m C-band dish to watch television

In the beginning of satellite communication there was C-band (4/6 GHz). Back in the days we already knew that higher frequencies would provide better links since the gain of antennas increases when the frequency does (anyone ever seen a 32 m Ku-band antenna?). But working at higher frequencies is difficult and requires state-of-the-art technology. Moreover for commercial deployment, the hardware must be affordable. While technology evolved, Ku and Ka band equipment became possible.

Using higher frequencies pushed up the achievable carrier-to-noise in a satellite link, hence higher modulation modes and efficiencies came into reach. That is what we saw with the breakthrough of DTH services in Ku-band. In the beginning maybe not so much the increase in efficiency but the smaller DTH dishes allow for a cost effective and practical deployment of a network. Imagine having to install a 2.4 m C-band dish to be able to watch television (maybe for you as an engineer that would be cool but I bet your wife wouldn’t agree). No, nowadays, DTH dishes are as small as 50 cm.

Bigger rockets allow for more powerful satellites

As satellite technology improved, so did the satellites themselves. With the first satellites having an EIRP (a measure for the power that they beam down) in the range of 35 to 40 dBW, we see now satellites having EIRP’s in excess of 50 dBW (remember going from 40 to 50 dBW is a tenfold increase in power). On top of that the electronics in the satellites became more powerful, generating far less additional noise, which also contributes to better links. Even advancements in rockets have contributed; more powerful satellites have a bigger launch mass. Now easily launching payloads in excess of 6 tonnes, bigger satellites can be launched with more transponders, bigger solar arrays and more powerful HPAs. Bigger rockets allow for more powerful satellites.

I’m not a firm believer in Ka-band

There are also more satellites. More satellites means that spotbeams are possible. Now I know that spotbeams are usually referred to when talking about Ka-band. But (as some of you might know) I’m not a firm believer in Ka-band. When I first started hearing about Ka with those huge EIRP’s and G/T’s I could see a fantastic link budget in front of my eyes. But then came the interference that kind of killed the dream. But that is maybe a good subject for another discussion. I think Intelsat is playing it smart here. With the EPIC satellites, there will also be spotbeams, but they will be in Ku and will be much larger. So we are looking at all of the transponder power beamed down in a nice spotbeam.

My personal mission is to evangelize DVB-S2

Other thing that does increase the efficiency is newer codings. Have a look at my LinkedIn page… it says that I’m a DVB-S2 aficionado and my mission is to evangelize DVB-S2. One of my favorite lines (back in the days when DVB-S2 was new) was that DVB-S2 is “the last standard”. This because we were so close to the Shannon limit and maybe also because we didn’t want to scare the industry to have to throw away their investments in new equipment because a new standard would be available in a couple of years.

Well, I need to say that I might have been a little short-sighted.

Extensions to DVB-S2 are no longer negligible

The extensions to the DVB-S2 standard provide gains that are no longer negligible. Are we talking about the new “last standard”? Who knows? We are getting pretty close to that brick wall built by mister Shannon.

But there is something else, let’s go hard-core technical: when you have a close look at the typical curves for C/N versus efficiency for different MODCODs, we see that the curves tend to fall off for higher FECs. Quite normal: as lower, more robust FECs are what they are: very robust, the higher FECs (like 8/9 or 9/10) don’t provide that much protection and require more carrier-to-noise. But the increment in gain when going from 2/3 to ¾ is bigger than when going from 8/9 to 9/10. So what we see is that when you run a link at 8PSK 9/10, a 16APSK 2/3 is requiring less C/N and is more efficient.

Any link running 32APSK 9/10 will happily accept 64APSK.

Now back to reality, we see that about 50% of the ACM links we deploy jump up to 32APSK 9/10. In those cases legacy (relatively big) antennas are used and the ACM helps to really “squeeze” the last out of these links. It’s kind a like driving at full speed and having the needle pressed against the end of the dial. But, taking into account the above, any link running 32APSK 9/10 will happily accept 64APSK.

So, is 64APSK a reality and really usable? Yes, I do think so. The codings provided in the DVB-S2 extensions are more powerful and having links available running in 32APSK will certainly allow 64APSK.

We need 256APSK

Having done some tests recently, we did a link on a pretty powerful satellite. A 72 MHz transponder having 52 dBW was used and 9 m Ku-band antennas. As we tweaked the link we reached 64APSK 9/10 with another 2 dB in margin.

Needless to say that I looked at my colleague and we both immediately said “we need 256APSK!”. Not that that is going to happen tomorrow but with some new technology launched soon, I’m sure that it will be possible.

My 2 cents? The new DVB-S2 extension standard will certainly allow 64APSK on pretty decent satellite links. The satellites are being built right now, the (Newtec) hardware is ready!

What do you think?


PS: I’ll let you in on a little secret, the new standard even proposes up 1024APSK!



What a well-written and informative article by someone who obviously knows their stuff. Thank you.

I guess my only comments are that these modulation schemes impose tighter and tighter constraints on the phase noise requirements of the ground terminals, and the non-linear effects of ground station SSPAs, as well as the intermodulation in the transponders.

That said - nice work!

Aravind Kilaru wrote

Hello Dave,

If possible please update few obtained result.

What is the margin on 64APSK in monsoon season or rainy climate?

what is the maximum System Noise Temperature during the experiment?

What the maximum Noise Temperature Due to rain recorded in the experiment?

What is the elevation angle during the experiment?

Congratulation, for the successful results on 64APSK.

Ben D'Auvre wrote

Nice article Dave... it has a nice approach from the technical side. But what we will be the effect on the commercial side. Who will win most by using 64APSK?


Hi Bob,

Thanks for the nice comments. As for the phase noise well, 64APSK will usually be used on professional links with decent quality BUCs and LNBs where the phase noise performance is pretty good and standard, there is no need for super-class (read very expensive) equipment.

For the non-linear effects indeed, these higher modulation modes require a bit more back-off for linearity but still there is plenty of gain.

Intermodulation shows itself as additional noise, so that plays for any type of modulation. True, with higher C/Ns needed, any excessive interference could bring down the linkbudget. That's why we won't see any modulation modes that will require C/Ns in excess of 25 dB (IMHO) because the interference components in a link are minimally of that magnitude. But with 1024APSK needing only 22~24 dB, I think this will still be in reach for some specific cases.

Take care, Dave.