LEO Terminals – Pointing the Way Forward
There was a certain buzz created in the build-up to the latest event in the long-running GVF series, presented in partnership with Connectivity Business News, with almost 500 individuals from 87 countries registered. It surely has lived up to expectations packing the ‘virtual room’, with 326 viewers tuning in to the live event to hear the discussion.
The dialogue began with a definition of a ‘LEO terminal’, including an exploration of the degree of integration of LEO terminal components and the more software-centric facets of their design. The conversation went on to explore SpaceX/Starlink as “the elephant in the room”, asking “How can the rest of the industry compete?”
The conversation then turned to answering “What Metrics ‘Matter’?” A detailed analysis drew-out the fundamental partial conclusion that what essentially matters is the ability to provide the service the customer wants.
Further points addressed the connected issues of terminal inter-operability and standardisation, and finally turned to answering the question, “Where can players go from here in building the (LEO) connectivity ecosystem?”
If you missed this “live” event, catch-up with the recording now!
Q & A continued….
Thank you to our audience for taking an active part by asking many questions. We ran out of time to respond to the questions below but our panellists were kind enough to answer after the event ended…
1. Can you please define “CIR”?
Kartik Seshadri (Hughes): This is a requirement for many enterprise customers who expect a certain amount of bandwidth available at all times.
2. What is your opinion on the use of LEO constellations in the commercial aviation sector? Do you feel like Starlink will be ahead of the competition due to their know-how in LEO terminals?
Will Mudge (Speedcast): Starlink is ahead of the other LEOs. They have more experience with the mobility required and have already launched an airline product.
Kartik Seshadri (Hughes): This is to be seen and Starlink for aero is still nascent but they do have a head start. LEO capacity is just one component for aero connectivity solution as this requires STCs, integrations etc.
3. How is the ‘jitter effect’ fixed on some heavy applications like Oracle, and what is the result also on RTP traffic like VoIP and visio, especially on SS7 or other signalling protocols for GSM backhauling?
Will Mudge (Speedcast): We haven’t seen a jitter issue on applications, but we have seen these applications struggling with the changing public IP.
Kartik Seshadri (Hughes): Jitter can impact VoIP/RTP traffic but many VoIP phones already have a jitter buffer inbuilt to handle jitter in RTP traffic. Similarly for GSM cell backhauls we typically configure committed bandwidth approaches to ensure that jitter over satellite is minimized.
4. Is there a place for LEO operators through transponder leases? Starlink, so far, has the widest offerings for geographic coverage.
Will Mudge (Speedcast): We haven’t seen transponder leases with LEO given how they fundamentally operate.
5. Is latency important only for real time applications which are less percentage compared to overall requirement of broadband connectivity?
Will Mudge (Speedcast): Latency has effect on more than just real-time applications.
Blad Stavropoulos (Intellian): Latency is important on an overall approach, if you are able to deliver traffic faster the demand for bandwidth is reduced.
6. Now the LEO is more or less used by Starlink. And I guess the vendor will also moved in that direction Do you think as Vendors that you will more invest in LEO or still continuing invest in GEO and MEO?
Will Mudge (Speedcast): We are still seeing investment in GEO satellites and supporting materials. Hispasat is launching Nexus in about 10 days as an example.
7. If it’s not economical as compared to GEO (HTS), it will be hard for it to survive?
Will Mudge (Speedcast): We really have to look at the value systems like GEO bring and then monetize accordingly.
8. What will be the latency in Starlink? Secondly, will there be any outage or delay during satellite reconciliation or switching? Like if I take this solution on GSM network backhauling then how better it will be instead of GEO service.
Will Mudge (Speedcast): Latency is around 50ms, very similar to your home or business internet from a cabled provider. We have not seen outages due to satellite switching, but there are outages.
Kartik Seshadri (Hughes): Satellite switches in LEOs should be seamless. Applications like cell backhaul require adaptive committed bandwidth to operate seamlessly.
9. So, (non)autopoint of GEO VSAT is the problem for GEO?
Will Mudge (Speedcast): GEO does have autopoint, it is telling it when to switch from LEO to GEO and what to point at that is the trick.
Blad Stavropoulos (Intellian): Hope I understand the question properly. Autopoint is more challenger than GEO. NGSO requires more accurate pointing and more precise tracking than GEO.
Kartik Seshadri (Hughes): For fixed site applications GEOs don’t need autopoint – and for mobility GEO autopoints work well and have been demonstrated.
10. Can you explain what “Make Before Break” is?
Will Mudge (Speedcast): “Make Before Break” means that a second connection is established and passing data before the primary link in use is disconnected.
Blad Stavropoulos (Intellian): “Make Before Break” is the acquisition of the next rising satellite and be able to transmit on that satellite before the falling satellite disconnects from the remote.
11. What about the end user’s data security on LEO? Will the network layers get exposed?
Will Mudge (Speedcast): Security is a big question and should be discussed with the respective provider. Each has their own approach and it depends on which network layer you are asking about.
Kartik Seshadri (Hughes): Network layer security doesn’t guarantee end user data security.
12. LEO and MEO will deliver lot of capacity (Gbps) with low latency (<100ms). Considering the relatively small size of flat antennas, what is the maximum uplink and downlink data rate that can be achieved per terminal? How does it perform compared to GEO terminals with the same real estate?
Will Mudge (Speedcast): LEO will outperform GEO performance on the same panel (we have tried and tested that). The uplink performance is also linked to how many MHz a constellation allows for its links, so it can vary from one LEO to another. Right now on average we are seeing about 20 Mbps being a normal value.
Blad Stavropoulos (Intellian): LEO/MEO constellation have the benefit to be closer to earth, which is gives them better G/T to each terminal. This results in better throughput and bandwidth delivery. The maximum uplink would be defined by the network efficiency and network operation ratio. Any number we provide today would be just a guess.
13. Are the LEO networks going to be able to support high density hubs, given that they were designed for more rural applications? Are there fundamental regulatory PFD requirements that will limit the amount of capacity LEOs can provide in a geographic region?
Will Mudge (Speedcast): Yes, there are fundamental limitations on the LEOs that will have effect on the amount of bandwidth to an area. So, the definition of area becomes the next question and that changes with the beam shape of the LEO which is much different OneWeb to Starlink.
Kartik Seshadri (Hughes): The amount of bandwidth delivered to a region will vary based on the constellation and number of satellites etc.
14. Will LEO services replace the most of the GEO services in near future or what will be the overall impact on GEO market? What’s your view on that?
Will Mudge (Speedcast): LEO and GEO both have their places. GEO satellites continue to be launched and there is a fleet of satellites available. The need for CIR will drive the marketshare between GEO and LEO at the moment.
Blad Stavropoulos (Intellian): There is a significant amount of services complete capable to continue their operation on GEO platforms. Those must like would stay over GEO services as the capacity cost and bandwidth efficiency would enable them to stay there. That it is why we continue to see GEO satellite launches these days.
15. Can you guys also discuss about the LEO services for IFC market, Maritime services… the pros/cons etc… Would love to hear that. Thanks
Will Mudge (Speedcast): There are a lot of pros/cons for LEO in these markets. This is a much longer discussion question. Please feel free to reach out if you would like to discuss.
Blad Stavropoulos (Intellian): LEO has it’s own niche, which will expand the ability globally to provide internet across areas where services were not available before. In there, LEO would take any giving business GEO had in there. The speed, latency and capacity delivery by LEO are not comparable on this segment of the service delivery. MEO, on the other hand, has this area of delivery with limitation on look angles, but those would be overcome soon when MEO includes the incline orbits satellite for polar coverages. Then we will have Global services in MEO/LEO and GEO with their own unique benefits and limitations.
16. Can both GEO and LEO coexist to give a complete solution?
Kartik Seshadri (Hughes): Both LEOs and GEOs will co-exist.
17. Just to be sure, in a parabolic antenna solution, do you still use two parabolic dishes for each user terminal? Being maritime or cellular backhaul?
Will Mudge (Speedcast): OneWeb has a parabolic solution, but Starlink does not. So, it is still an option in some cases.
Blad Stavropoulos (Intellian): Yes, parabolas continue to have a large implementation across LEOs. We see parabolic terminals being more efficient and with better reliability in Maritime, as it is able to compensate ocean motion and other factors. We see flat panel for small cellular backhauls but parabolics in areas where the uplink is more demanding. It is our belief each type of terminal has their own pro and cons and we just have to understand your application and be able to provide the proper solution for that.
Kartik Seshadri (Hughes): Parabolic solutions are predominant in maritime but ESAs will make their way there as well. Dual parabolics in OneWeb are used for make before break.
18. Is the system prone to rainfade and local storms?
Will Mudge (Speedcast): Yes. all satellite are. Right now the LEOs are Ku and while the constellation being closer helps, there is still an effect.
19. Would you expect the industry to replace big 2.4 meter antennas that are GEO and MEO capable with a few LEO terminals at one tenth of the cost?
Will Mudge (Speedcast): It’s possible, yes. But there may be a need for the 2.4m based on the use case. CIR as an example with a MEO solution, or C-band’s ability to work in rain better than a flat panel.
20. Do you foresee LEO becoming more accepted by government or large enterprise as a solution over GEO?
Will Mudge (Speedcast): Absolutely. Many governments have their own LEO projects and even Starlink announced a government specific solution after success in the Ukraine.
Bill Marks (Kymeta): LEO services will be location and price dependent. We are seeing that the government new requirements are calling for both GEO and non GEO capabilities from one terminal.
21. Does the use-case dictate that efficacy of terminal type or technology? For example – fixed consumer broadband vs aviation?
Will Mudge (Speedcast): Yes, absolutely. Use-case is key to choosing the terminal/service that fits best for you.
22. Is there any solution for the high latency problem of the VSAT communication?
Will Mudge (Speedcast): Yes, we use accelerators like Riverbed, Silverpeak, Xiplink, and others to accelerate traffic and provide an experience closer to a business application.
23. Is cell handset direct to satellite, e.g. iPhone to Globalstar, a significant industry trend? Will cell handsets become a large share of LEO terminals in the future?
Will Mudge (Speedcast): In my opinion, yes it is, as it can cover the cellular backhaul applications or reduce the need for cell towers, thus increasing coverage and reducing CapEx to provide service to a region. My opinion on handsets is they will be from the manufacturers like Apple, but will be compatible with a service from a LEO like we have seen with iPhone 14.
24. Will there be any outage or delay during satellite constellation or switching? If we take this solution on GSM network backhauling then how better it will be instead of GEO service. How effective this technology is in terms of LB’s and BW optimization?
Will Mudge (Speedcast): We have not seen an outage or delay in switching. Those without make-before-break, even that is not noticeable in our testing.
25. What’s about the reliability of FPA vs classic dishes?
Will Mudge (Speedcast): It is still early to tell on this one, but in theory an ESA should be more reliable than a parabolic in the same scenario because there are no moving parts that could fail. But, compared to a fixed parabolic antenna, then it comes down to a comparison of the reliability of the electronics like the BUC and LNB.
Bill Marks (Kymeta): Flat panel antennas have no moving parts, which means there is no need for annual service of wear and tear on things such as gears and belts like traditional dishes. LEO operation requires tracking across the sky. Since flat panel antennas have no moving parts they can seamlessly track and connect which ensures higher reliability.