Cloud computing is a new way of harnessing the information at our disposal in an information-dominated world, and the satellite sector can play a role in the development and operation of the cloud.

There is no one set definition for cloud computing. Mike Tippets, vice president, Hughes Solutions Group says, “Cloud computing is the sharing of computing resources. You can branch that off and talk about cost savings and about access to applications and so forth. At the starting point, it is about sharing resources.”

Duncan McCarthy, research scientist, National Geospatial Intelligence Agency (NIGA) says, “Definitions of cloud computing vary considerably. Some consider simply running virtual machines a form of cloud computing. Others consider Web services or the Internet to be the essence of ‘cloud computing,’ and finally others consider cloud computing to be the simultaneous use of many machines together to run a single program much like what historically has been called ‘cluster computing’ or ‘grid computing.’ … The cloud computing definition I am most comfortable speaking about is closest to the grid computing or cluster computing idea. It is the idea that one computing job can be distributed amongst many different and potentially heterogeneous computers to dramatically reduce the execution time of that program. This technique can be applied to geospatial computing to answer questions that are impractical to answer on a single machine because the processing time for huge data volumes is simply too great.”

Kuni Takahashi, an independent analyst, takes it a step further. “I think it’s important to segment cloud computing into private and public clouds. Private clouds are internal corporate data centers that utilize cloud based technologies and services. Historically, satellite technology has had limited presence in the enterprise data center, so I don’t see tremendous overlap between satellite technology and the private cloud. Public clouds are third-party, hosted environments where ‘rent’ is based on usage only,” he says. 

The Big Question

The big question is what could this move to Internet and cloud-based computing mean for the satellite industry? Is it an opportunity or a threat? Tippets “definitely sees an opportunity for the satellite industry. There are broad opportunities for satellite companies. If you focus your paradigm of cloud computing on word processing, spreadsheets and similar business applications, the opportunity may seem limited, but if you consider applications that involve video and other rich media (Flash, Mpeg-4), cloud computing becomes very interesting, and satellite companies have a very compelling value proposition there,” he says. “I would define the ability for the CEO to address his or her company with a video link as a cloud computing service because they can avoid installing costly infrastructure and just take advantage of services outside of the organization. There is an opportunity for satellite companies to provide that external communication service.” Another potential application is company training “and the ability for an organization to use different types of training media, again video and rich media, as well as the written word and graphics. The ability to have that hosted centrally and allow anytime, anywhere access for training content. In my opinion, that is a cloud service that satellite companies can offer,” he says.

Tippets also says companies can look at securing the cloud. “Another area of opportunity is security and threat management. Threat management is an umbrella term for intrusion detection, anti-virus, malware and spam protection. Satellite companies are in a unique position because we bring the customers’ Internet traffic back through our facility. We are in a position to offer these services at the customers’ premise or at the satellite providers facility. If we do the latter, the customer is putting that functionality in the cloud and getting their security protection as a service offering.” Defining Hughes’ role in this new information technology paradigm, Tippets says, “Our first priority is to make sure our broadband customers have secure, reliable access to the cloud computing services and applications that they desire. We will provide a network that offers remote connectivity to cloud services through satellite, wireline or wireless links. For us, it is the opportunity for our customers to share those computing resources and our responsibility to them is providing them that access, no matter where their remote presence may be.”

Mark Dankberg, CEO of ViaSat, also sees “a huge opportunity” for the satellite industry. “One interesting thing about an enterprise cloud computing network is that it is not enormously different from the mainframe type of network when you go back 20 to 30 years. There are a lot of similarities between that and cloud computing — although the network that connects the computing resources to the end user is much, much better and more pervasive now. VSAT did very well with those enterprise networks. I don’t think cloud computing in itself creates opportunities that weren’t there or undermines opportunities that were. The opportunity arises from company’s interest in rebuilding enterprise networks to capture the cost savings and ubiquitous access that cloud computing offers.”

The onset of cloud computing also will make satellite companies more efficient, says Dankberg. “If you look at the cloud computing world, there is a whole new sector in the networking industry of WAN acceleration and application delivery appliances. There are three leaders in this area — Cisco, Riverbed and Bluecoat. What they are doing is selling appliances you can put at your data center and your remote sites that essentially makes the network look faster and deals with latency and congestion. Basically, those appliances are doing the same kinds of functions that VSAT networks have always had to do to deal with latency over satellite and make applications work. The enterprise applications being run are the same types of applications that companies have always run. One of the biggest issues in the VSAT world has been latency. This can make those applications run slower. WAN acceleration appliances, and the software clients that go with them, are dealing with those issues in a very fundamental way that no longer means you have to fix every application to make it work with latency of any form, including VSAT.” 

Impact on Satellite Broadband Market

Takahashi believes a company like O3b Networks could also benefit from the move to cloud computing. “O3b Networks is a perfect example of a satellite company positioning themselves to benefit from cloud computing. The ‘other 3 billion’ will undoubtedly be using many Web 2.0 and other cloud-based applications, which will require a low-latency, low-cost and high-bandwidth network that O3b networks is building. I would imagine that Google, one of O3b’s investors, is also thinking along these lines,” he says. Broadband providers such as Hughes and ViaSat ALSO can benefit from the move towards cloud computing. “The satellite broadband market is positively impacted as more computing resources move into the cloud. I think the cloud adoption cycle is very synergistic to the target market for satellite broadband. If you look at the Software-as-a-Service (SaaS) adoption trend, branch offices, remote workers, home offices and small- and medium-sized businesses (SMBs) have been early adopters. I suspect many of these remote users are already using satellite broadband, or will begin to, if they want to have fast access to their SaaS applications. High-speed Internet is essential for an acceptable end-user experience when using cloud-based applications,” says Takahashi.

Tippets adds, “Users are going to have access to these services and applications in the cloud. I have said this a number of times, as a broadband connectivity company, it is not a question of satellite versus other types of connectivity. Satellite connectivity is one more component of the WAN infrastructure. What opportunities does this provide for satellite companies? Readily available, high-bandwidth connectivity will be helpful to the developing cloud services infrastructure,” he says. 

Imagery Sectors Embracing the Cloud

Cloud computing “solves a number of problems, says Brian O’Toole, CTO of GeoEye. “On the business side, the cloud model will enable global distribution of content and access to on-demand processing resources. It is going to enable the delivery of managed services. It is also going to enable the delivery of new information products to new customers that were difficult to serve efficiently in the past.” Secure cloud computing also can have an impact in the military arena. “Cloud computing technologies and services will provide a global infrastructure that can handle the data volumes for processing and delivery of relevant information products to a much broader user base. This in turn will drive demand for more content and related value added services. Cloud environments will not only increase accessibility but will also support the last-mile delivery to the warfighter and business professional,” he says.

GeoEye has used the cloud to distribute images, O’Toole says. “We have used in the past public clouds like Amazon back when we were able to get a shot at the President Obama’s inauguration. We had to add surge capability. We had a million downloads in a day of that image. We basically deployed our Web site that people were using to download that image out into Amazon’s service. Because of our Web site software, we were able to deploy out to that environment and support the surge,” he says. The imagery satellite operator now is “looking at a number of cloud environments for different uses and applications. Cloud environments can generally be defined in three categories. There is the public cloud which are services like Amazon’s EC2, which are multi-tenet on-demand environments. Data centers are a type of cloud environment that act more like an outsourced [information technology] department, providing processing, storage and network resources. In the middle is the emergence of private clouds, which brings the benefits of a public cloud environment but are more dedicated and secure. We are looking at all three since they each have their strengths and weaknesses.”

In terms of the impact for the satellite industry, O’Toole says, “It is going to drive new demand through improved accessibility. In the past, the size of the data and the processing pipelines have provided challenges in enabling rapid production and access of imagery products to end users. Most satellite systems have dedicated and proprietary ground systems which produce a set of specific imagery products. Value-added processing was done off-line and delivered within days, weeks or months. In the near future, we will be able to process out of the ground systems on-demand. We are doing some things right now with one of our customers, where we can download data in less than a few hours and have it online into a product that can be used by thousands of users. That was unheard of a few years ago. Even with the Obama inauguration, we had a picture out to the media within hours. This is radically different for this market.”

McCarthy says he expects the NIGA to start taking a more progressive approach towards using cloud computing. “I think the real benefit is that cloud computing allows for the analysis of a huge volume of data that previously had not been able to be analyzed,” he says. “Years ago, predecessor agencies would have taken pictures and manually looked at those pictures in order to answer questions. That is not difficult when the number of satellites is low and the amount of data being collected by each one is relatively small. Modern satellites are more numerous and generate more data. Now we are pursuing the ability to analyze remotely sensed data by computer. Eventually anyone in the remote sensing industry, for instance someone analyzing crop health, may have the ability to analyze the entirety of the United States for a certain problem because all that data has been gathered at a high resolution. Another application of cloud computing in the geospatial realm would be doing weather modelling at a finer, more local scale to get more accurate weather predictions. These large volumes of data may yield more predictive power when analyzed this way.”

More sophisticated satellites will fuel even more demand for cloud-computing applications, McCarthy says. “As we get more sophisticated techniques for getting that data down to the ground, the data stream will go from megabytes per day to gigabytes per day and eventually to terabytes a day. Three or four satellites gathering terabytes of information per day will soon develop petabytes of information in an archive. A petabyte is 1,000 terabytes, so to ask questions on that volume of data, cloud computing is going to be absolutely necessary. You will not be able to answer some questions without about being able to divide that processing volume amongst many computers. … I would suspect satellite companies would be investing in their own cloud computing capabilities at ground stations so data can be brought down and processed into information so more sophisticated products might quickly be delivered to customers. Alternatively they might contract with firms like Amazon or Google for processing. There have been attempts at cloud computing for massive data already, for instance the SETI At Home project, where radio telescope data is processed by sending little packets of data to thousands of home computers across the world for analysis. While this is not satellite data, it is very similar and is one of the first great applications of cloud computing. The satellite industry could learn from this example and instead of processing their data on home computers, process it on their own clusters of computers or those that they rent from vendors. Cloud computing presents many more opportunities for satellite vendors and users to get much more out of the data than they have before,” he says. 

Future Cloud-Based Opportunities

Takahashi believes there is a strong opportunity for satellite companies in the video space. “Satellite technology is typically the lowest-cost medium for broadcast applications. Therefore, in the traditional video broadcast model, I do not think satellite technology is under threat. However, the traditional broadcast model itself continues to be challenged by other video delivery platforms such as online, mobile, etc., which analysts are beginning to define as the Syndicated Video Economy. I think there are tremendous opportunities for satellite companies in this new video model. For example, there are a number of companies offering encoding as a service, such as On2 Flix Cloud, Encoding.com and HD Cloud. These public cloud services allow Web publishers, content owners, and advertisers to easily and cost effectively encode and transcode video in multiple formats to deliver video content across any platform. Although this content may never touch a satellite, it is still the video market and as it evolves so can satellite companies.”

The concept of cloud computing is clearly gaining credence in the information technology world as companies look for more efficient ways of distributing and disseminating information. “Cloud computing will dramatically change telecom and [information technology planning for companies,” Takahashi says. Gartner recently released a report that said one in five companies will get rid of all [information technology] assets and move them into the cloud by 2012. If it gets anywhere close to Gartner’s predictions, [information technology] spending and planning will be very different for many companies. I suspect most of these will be SMBs that would rather forego the upfront capex and time of developing or maintaining their own IT systems,” he says. The pace of change also could be quite fast. “The Holy Grail of cloud computing is a hybrid environment that combines private and public cloud resources, enabling users to send bursts of traffic, or workload, into a hosted cloud and only paying for what they use. In order to get there, there are many concerns over IT chargeback, scalability, performance, security and other issues that need to be addressed. I think cloud computing companies will solve some of these problems over the next 12 months,” he says.

In terms of the trends emerging, O’Toole says, “The cloud will enable the industry to build and offer new services to a whole new set of customers through subscription-based business models. For example, they may want to subscribe to imagery content for an area of interest and have that streamed directly into their business environment. A commercial example is what NetFlix is doing with on-demand services. They have always had a subscription model, where you order on-line and get a DVD in the mail. Now, they are starting to shift where you can get movies on demand through your cable box and over the Internet. The service is providing on-demand access to content. I think the next 10 years will be exciting as we see new information products emerge with the added benefit of flexible access and delivery through cloud enabled solutions.”

Mark Holmes is Via Satellite’s Associate Editor.