Virtualization and cloud implementation are critical components of information technology planning, acquisition and management going forward. Cloud implementations are important to security, efficiency, effectiveness, cost savings and more pervasive information sharing, particularly among enterprises. Cloud architectures also are extremely important for more effective use of mobile technologies. Mobility increasingly is important, particularly for the military, which needs a full range of information technology services while on the move. Yet increased movement to the cloud, along with traditional uses of spectrum, are putting unprecedented demands on every part of the spectrum.
The federal government certainly has embraced the cloud. Its “Cloud First” policy has been in place since 2010. The General Services Administration put FedRAMP in place to facilitate acquisition of cloud services against a clear set of standards and criteria. These have evolved over time; just last month, FedRAMP published updated security standards. The Defense Information Systems Agency has put government-provided cloud services in place for the U.S. Defense Department. A set of private, hybrid and public cloud implementations are appearing throughout the federal government, with some of this occurring under the umbrella of federal data center consolidation efforts. Cloud technologies certainly can facilitate data center consolidation in some cases.
Parallel cloud initiatives are taking place throughout industry, at the state and local levels and among coalition partners. NATO, for example, is pursuing a cloud strategy aggressively. The information technology industry has lined up globally to provide cloud products and services. While some experts had expected cloud implementation to proceed more rapidly, progress is being made and cloud offerings are affecting every part of our lives. There is no doubt that, through the cloud, everyone has access to capabilities that were not available before.
As everyone moves to the cloud, more applications and data are being placed in central storage for remote use or retrieval. Infrastructure, application and data architects, as well as decision makers and users, need to keep in mind that network access is needed whenever applications are executed remotely and/or data is retrieved. Bandwidth on demand is needed if these services are to be provided effectively. This translates to the need for ubiquitous bandwidth, whether a user is mobile or at a fixed location.
Much of this bandwidth puts demands on spectrum, and governments are trying to determine which segments of the spectrum need to be reserved for government use—at the federal, state, local and tribal levels—and which segments can be used to support commercial demands, which are growing at an incredible rate. International considerations in the spectrum are extremely important and complicate national decisions in every country.
From satellite to cellular to carrier trunking to first responder/emergency management radio to military radios, new waveforms are emerging, devices are becoming more capable, and use/demand is growing at a frightening rate. The potential for unacceptable data transfer rates and interference is tremendous. And the problem only will worsen as innovators find better ways to deliver content from the cloud, including streaming video. Remember the LightSquared controversy—LightSquared wanted to offer wide-bandwidth Internet access to everyone, including rural users, which was a worthy objective. The problem was adjacency to GPS that caused interference, which created a safety hazard. These kinds of contentious issues are common. Most are just not as visible as the LightSquared issue.
Spectrum managers and decision makers are working to understand future requirements and balance the many demands for a finite spectrum. This effort is extremely complicated and a very tough job. We need to help them by considering bandwidth demands, current and future, in every initiative we pursue. Particularly as we develop new and innovative ways to leverage the cloud, we must estimate increases in corresponding bandwidth demand and communicate that to spectrum managers and decision makers. Spectrum managers cannot just be left with linear requirements for growth—it is not going to work that way.
Back in the Cold War, we were very concerned about Warsaw Pact jamming. I always said then that the Warsaw Pact didn’t need to spend much effort on jamming because we did such a good job of stepping on ourselves. As long as we have used radios, we have had interference. That was in a simpler time, when sources of conflict in the spectrum were much more limited. We need to ensure we don’t build a cloud-based infrastructure we cannot fully use because we didn’t get spectrum right.