Subscribe
  • Home
  • /
  • Networking
  • /
  • Why satellite can't work... or at least, why the market thinks so

Why satellite can't work... or at least, why the market thinks so

Debunking the key myths about satellite services.


Johannesburg, 05 Sep 2018

I haven't even started the presentation and already the technical director in the customer team has proudly informed the other decision-makers that satellite is slow, expensive and has a latency bug that makes computer systems fail. That's it, game over. I am supposed to pack my bags and go home.

My satellite network solution will not work for his very critical business applications, he thinks. Thinking quickly, I decide to relax, close my laptop and start a courtesy discussion. Being polite gentlemen, they can at least let me finish the coffee that is being served, says Dr Dawie de Wet, CEO of Q-KON.

"Big Springbok rugby game on this week-end," I start. "So pleased that DStv has secured the rights so that we can watch the game even at the lodge we are going to." Everybody agrees and we have some light comments about teams, players and the game. I guide the discussion a bit and say that I wonder about the aspect of "latency". Will this mean I see the big moments later than my friends in the city, or how come I can watch a game on satellite TV every split second of the way, even with slow motion details, and not be affected by this latency issue? The technical director looks at me with a question in his eyes and his body language saying he is partly irritated, partly confused.

We continue to have a different discussion and they allow me to explain some principles of satellite technology. In the end, they became a valuable customer that can apply satellite in the correct niche applications to support their business.

Allow me to also share some context on the three major myths of satellite services: It's expensive, it's slow and applications don't work due to latency.

Let's start with latency. Yes, geostationary satellites are at 35 000km above earth, and yes data packets do take 550msec to travel to the remote sites and back, and no it doesn't mean applications can't work. What the DStv example shows is that satellite networks must simply be designed as an end-to-end service. What doesn't work is to run an application using fibre network connectivity, replace the circuit with satellite and then expect the application to still operate. Some adjustments in the application layer are necessary to adjust for the expected satellite transmission latency. This means latency is more an application issue than a satellite technology issue.

Next, let's talk about speed. Firstly, what is slow in today's technical expectations? 1Mbps? 10Mbps? 100Mbps? Satellite networks are very good at doing video distribution: your DStv terminal at home receives the equivalent of 72Mbps. Is this slow? The perspective that satellite is slow comes from the initial broadband services where customer profiles were offered at 512kbps. Yes that is slow, but it is also history. Today's low-cost 1.2m 3W terminals can provide uplink data rates of 10Mbps and receive rates of 100Mbps. That's not slow anymore.

Finally, there's cost. Uncapped fibre to the home at R499... yes, against that reference, satellite will be expensive. What about providing a point-of-sale service at less than R50 per month while the current service offers are R500 per month, or a branch backup service of 4Mbps at only R1 500 per month? For these applications, satellite services are actually the better cost option, and dare we say, cheaper. The key point is that satellite services are no different than any other technology. Applied in the correct application, it will offer good value for money. For applications such as content distribution networks, packet data and on-demand services, satellite will be the better cost option. The principle is that satellite networks are a specialist niche connectivity service and must be implemented with an appreciation of their respective strengths and value aspects.

The satellite industry is rapidly developing and a range of new platforms have entered the market in recent years. These include Ka-band services, the new EPIC high-throughput satellites from Intelsat, and planned medium orbit constellations, etc. In addition, most technology marketing messages are that what is new is better, faster, cheaper and that what is old needs to be replaced and upgraded. In a digital technology world where anything that is older than two years is considered obsolete, such a view might be logical and expected.

For network operators that use satellite network infrastructure, the position on replacement programmes is often more complex, and the decision to replace and make obsolete very painful. One of the key advantages of using satellite solutions is that the technology is extremely reliable and teleport infrastructure often operates for 10 years and more without major maintenance and upgrade requirements. For these reasons, there are operators that have C-band teleport infrastructure that are very much still service ready, while the market requirements have moved to a need for lower cost smaller Ku-band remote terminals.

The good news is that the Q-KON system engineering team has now successfully demonstrated the user advantages of using the latest Ku-band satellite technologies for user terminals while operating from a previous generation C-band teleport infrastructure. This solution effectively gives operators the best of both worlds. It is a win-win in all directions: re-purpose the previous C-band infrastructure to service high demand market sectors using low-cost Ku-band remote user terminals.

In this particular case, the Q-KON engineering team, under the leadership of Riaan Roux, Head of Engineering, has reconfigured a C-band hub network to connect to Ku-band remote user terminals over the Intelsat EPIC cross-connected platform. Although in theory this is expected to be a relatively standard procedure, Roux's teams have learnt all the small details and technical requirements to actually get such a solution operational. "Being able to calculate and set the customised network parameters required to fully implement this C-to-Ku network is just another example of the depth and experience of our engineering team," says Roux.

For the end-use, this solution enables a completely new business case with significant operational advantages. This user network can now leverage a previous redundant C-band teleport infrastructure and continue to provide business backup services, using the same remote Ku-band terminals while increasing the data rate capacity from 2Mbps to 10Mbps. This is a clear example of the potential of the Intelsat EPIC HTS satellites linked with the iDirect Evolution platform and integrated through an advanced end-to-end system engineering process.

Share

Editorial contacts

Mariette Scholtz
Q-KON Africa
(+27) 82 762 8346
mscholtz@qkon.com