The concept of bandwidth as network throughput and occupied spectrum are separate but related concepts. In this article, we discuss these two concepts, and explore how reducing network bandwidth could also reduce spectrum bandwidth. Why is this important? Because this would increase the efficient use of a finite resource—the electromagnetic spectrum.
For many readers, the most common use of the term “bandwidth” is to quantify computer network throughput. The amount of data that can be sent across a network is described as that network's bandwidth. The ability to quickly move information across the internet results in the difference between being able to watch a movie uninterrupted and watching a movie buffer.
Turning the clock back to the 1980s, the first home computer networks connected at less than 100 bits per second (bps). Full motion video was inconceivable at that speed. Sending a single still image larger than a few pixels wide took minutes or hours. However, consumer networks using telephone modems quickly evolved to 300, 1200, 2400, and eventually 56,000 bps (56 Kbps). This was fast enough for the average internet user to have a satisfying experience. Amazingly, in 2020, there are still millions of people accessing the internet over low speed telephone modems. As a result, there are applications specifically built to operate at this low bandwidth. “High speed” is an evolving concept. In 2020, high speed consumer networks, including home networks, are generally on the order of 100 million (100 Mbps) to 1 billion (1 Gbps). There is no upper limit on the horizon, only the limits of current technology capabilities.
Within the electromagnetic spectrum, the term “bandwidth” quantifies the amount of spectrum needed to transmit a specific signal. For example, in the early days of police walkie-talkies, the transmission used 25 kilohertz bandwidth of the spectrum. Over time, technology evolved and a single channel was reduced from 25 kHz to 12.5 kHz bandwidth, and then again to 6.25 kHz. Using only 25% of the original bandwidth — a much more efficient use of spectrum.
Current technology limits the amount of data that can be sent across a given amount of spectrum. When network throughput requirements increase faster than the ability to move data cross spectrum, more spectrum is needed. As a result, increasing network throughput is a fundamental force driving spectrum requirements, as users seek more throughput than currently available in their allocated spectrum bandwidth.
Having the highest resolutions may seem standard to anyone with a high speed network. However, when the network connection isn’t high speed, such as when reaching back from a forward deployed location, or being accessed from a ship at sea, network limitations severely constrain the speed, and therefore the user experience. Most modern data applications are not built for the type of low speed access found in some deployed locations.
The only way to get more speed is to use more spectrum. In other words, once the limits of technology are reached, the only way to increase network bandwidth is by increasing spectrum bandwidth. Increasing spectrum bandwidth may not be feasible for shipboard and deployed users. The developers of military network applications should consider that the user may be operating in contested spectrum or a low network bandwidth environment.
Unnecessarily high bandwidth applications are wasteful. One novel approach might include testing an application in extremely low bandwidth conditions like those found aboard ships and in deployed environments. It would be wise to consider all alternatives to high bandwidth applications in order to be good stewards of a finite resource — the electromagnetic spectrum.
Tom Kidd is the director for DON Strategic Spectrum Policy in the Office of the Chief Technology Officer, Department of the Navy Chief Information Officer.