A 20:20 View: Satellite and the Cloud
by Martin Jarrold
London, UK, January 3, 2020--A new addition to the GVF-C21 Partnership’s conferences and roundtables portfolio, Satellite and the Cloud, will be introduced in London on 3rd March 2020. Whilst not “a match made in heaven” the affinity between the two is a match made in – GEO, MEO, LEO – orbit.
My observation at the outset is that it has taken the satellite industry time to realize the full potential of going beyond just the delivery of Cloud-based applications to remote locations via satellite – the way that the Cloud has been making its mark in the satellite sector over the course of the last decade. But now, recognition of the greater opportunities in partnering with the Cloud is working through the entire, and expanding, satellite industry value chain. This is resulting from leveraging reduction in upfront CAPEX, reduced OPEX, rapid scalability, ease of development and ubiquitous accessibility, which is, in a virtuous circle, bringing further evolutionary change to that value chain.
This is a value chain that engages both space segment and ground segment – with a cumulative revenue opportunity of nearly US$17.7 billion by 2028, according to the July 2019 NSR report Big Data Analytics via Satellite (3rd Edition) – and characterized as encompassing data volumes andanalytics products not only arising out of satellite communications network operations, but also generated by NewSpace Earth imaging/observation/remote sensing, and the IoT/IIoT revolution.
Pre-NewSpace, Earth imaging/observation/remote sensing was confined to governments, their various specialized agencies, and their big program budgets. A myriad of private companies – start-ups and spin-offs from academia – driven by commercial competition for customers, are the very fabric of NewSpace. Various flavors of small satellites have rapid revisit times/frequencies for the same part of Earth’s surface, generating many thousands of images, vast quantities of imagery-based geospatial analytics and information to feed into GIS applications. Combining this with IoT/IIoT sensor-based applications (with hundreds of zettabytes of data generated each year by billions of devices) we can begin to comprehend the magnitudes of the big data analytics forecasts, such as that cited above.
Dublin-based Research and Markets define Cloud computing as follows: “…[It] delivers software, platform, and IT infrastructure services via a shared network. In this model, businesses access resources such as hosted software and applications remotely, i.e., via the Internet. The model not only obviates the need for making capital investments in servers and storage, but also results in zero operational expenses for running data centers. Cloud computing not only reduces business costs, but also makes applications accessible from any location, and reacts swiftly to changes in business needs...”
The Cloud benefits users by allowing flexibility in utilizing computing resources needed at any given time, but, due to the open nature of Public Clouds, information security needs have become increasingly complex. Concerns about Public Clouds, as opposed to Private Clouds, have tended to focus more on issues of control rather than on reliability or robustness. Major Cloud providers spend vast sums on cyber security innovations and yet customers in satellite (and in terrestrial) communications have tended to prefer using their own Cloud infrastructures, keeping data centers in-house. Confidence in using Public Clouds, as far as satellite networks are concerned, should, at least in part, be contingent on ensuring that networks which do extend to Public Clouds for critical functions have robust secure links into the Cloud.
Beyond the current traditional mainstream of Public Clouds, Cloud Service Providers are beginning to take note of increased opportunities for Cloud adoption in the satellite industry which is posing both interesting use cases and models for Cloud storage on space-based assets.
From the perspective of the Internet world the diversity of standards in the satellite equipment manufacturer and operator environment may be perceived as potentially impeding the growth of Cloud via satellite services. However, where satellite technology provides a complete end-to-end solution, as is increasingly the case in the solutions market, this obstacle becomes of reduced relevance.
Adjunct to the DVB-S2X satellite standard is Adaptive Coding and Modulation (ACM), a technology which automatically changes the modulation and forward error correction (FEC) of a satellite link to compensate for changes in link conditions, caused by such factors as the weather (e.g. rain fade), but also due to changes in the RF environment. ACM has been generally introduced into satellite networks, allowing operators/users to provide near 100% service availability in the presence of link disruptions. Whilst actual throughput would drop-off during the link recovery process, Internet applications are accustomed to variable data rates.
Satellite teleports have not changed very much in 20 years and often they are still, very largely, racks of hardware. But, are teleports verging on major change? Does it not now make sense for satellite teleports to be considered as data centers or, on the flipside, data centers to be considered teleports?
After all – leveraging any more advanced satellite-Cloud service synergies aside – more and more of the satellite industry’s long-standing (and more recent) end-user customer-base sectors are themselves undergoing rapid digitization and – particularly where these sectors operate in mobile or remote environments, traditional sectors like shipping and oil & gas (and sectors like farming and mining) – want their IoT/IIoT and other-sourced data available and accessible wherever it is needed. The ground infrastructure companies which provide hardware and software for teleports — used by satellite operators to distribute and control their services — increasingly recognize that such facilities could be made a lot more efficient by relying on Cloud servers to virtualize network functions.
Running data analytics offerings on virtual systems has many benefits: information and insight extraction for end-users becomes much less expensive, and the focus moves away from building the underlying infrastructure to dealing with the influx of big data and solving actual problems.
Additionally, of course, the (Public) Cloud providers want to get to their customers, and to get their customers data to everywhere around the world, which is more easily achievable using high throughput broadband satellites. When this data is – as is increasingly the case – very bandwidth-intensive, such as video streaming, the imperative for Public Cloud providers like Amazon, Google and Microsoft is all the greater.
Some perspectives highlight certain limitations of satellite networks, particularly bandwidth and latency, and how these interact with application acceleration. Satellite links are now capable of delivering higher and higher bandwidth, and at lower and lower cost, with high throughput satellites (HTS). Traditional FSS spacecraft operating in GEO, with throughput measured in the high Mbps range are giving way to newer generations of GEO satellites, with throughput approaching the Tbps range. In the course of this generational shift more traditional satellites – with a capability of 1-3 Gbps, and a CAPEX of US$180-200 million – have a Gbps price tag in excess of US$60 million; the newest upcoming generation of HTS – with capabilities in the 500-1000 Gbps range, and a CAPEX of US$600-700 million – come with a Gbps price tag of less than US$0.7 million.
In addition to the new GEO generations, an expanded MEO constellation, and upgraded and new LEO mega-constellations, will bring vast additional bandwidth capability to orbit. This additional throughput will also provide lower latency. Whilst this will further facilitate the continuing rise of fully managed satcom services, it should be noted that most applications offered via Software as a Service are not real-time but rather capabilities that rely upon servers and databases. Since most of the latency is due to processing and database look-up, satellite delay is secondary in the user’s experience. Managed Platform as a Service solutions can combine VSAT hubs, teleport/data center uplinks, and a terrestrial network enabling easy deployment of high-throughput connectivity in customer locations across the globe.
With constellations in LEO and MEO resolving latency issues satellite network operators and service providers now look to further improve their bandwidth delivery services to meet demand. They look to such Cloud technologies as dynamic Software Defined Wide Area Networking solutions that allow for real-time network optimization, understood in terms of the virtualization seen in the Cloud computing world.
Some facets of the Cloud will be impacted by the widespread arrival of 5G, although the full extent of this cannot be known until this new mobile broadband standard is rolled out and being used by corporates, enterprises, and governments as part of a communications “network of networks” that uses satellite in a wholly integrated functional role with terrestrial systems. However, it is possible to extrapolate from certain key features of 5G.
5G will effectively eliminate latency, allowing devices to connect nearly instantly. In extremis this might suggest the nadir of some aspects of Cloud computing as a whole. One of the main reasons the Cloud is so beneficial is that numerous devices – either in an organization for a Private Cloud or any user with an Internet connection for a Public Cloud – can connect to and transmit data with a central machine or hard drive located on the Cloud. But, if devices can connect with only milliseconds of latency and a minimum connection speed of 20 Gbps down and 10 Gbps up, 10 gigabyte video files can be transferred from user to user in about eight seconds and there’s no need to use a Cloud server as a repository. The Cloud will still have significant use cases in a 5G world – especially as Cloud providers are ready to adapt – as new use cases in Cloud computing and IoT/IIoT are driving greater-than-ever demand for high-capacity, low-latency connectivity.
The satellite industry is catching up in its adoption of the Cloud. Irrespective of particular application, Cloud computing reduces barriers to market for new start-ups in the satellite industry. As noted above, incorporating the Cloud into satellite businesses in network operations, or in Earth imaging/observation/remote sensing, brings several advantages. Perhaps soon, just like the telecoms groups comprising the broader telecommunication ecosphere, satellite operators and will pivot more towards becoming software companies.
Time will tell.
----------------------
Martin Jarrold is Vice-President of International Program Development of GVF. He can be reached at: martin.jarold@gvf.org
