Everything over IP transitions to IP over everything

Most likely, you are familiar with the world of IP communications. In a nutshell, IP telecommunications breaks telephone calls, TV signals, fax transmissions, computer data, video streams — any kind of telecommunications — into packets of data and then transmits those packets over telecommunications systems. At the receiving end, those packets are decoded and reassembled and then translated back into the original format: voices, images, facsimiles, or that great American novel you’re working on.

There are a number of reasons why IP-based telecommunications have grown in popularity. Perhaps the biggest reason: You don’t need a separate video cable, phone line or DSL line for data. Instead, it’s all packets, flowing simultaneously over the network. This cuts costs, makes the networks easier to maintain and minimizes unique types of equipment for different purposes.

The transition to IP telecommunications has become so prevalent that a new buzzword has emerged: everything over IP, or EOIP. The Defense Department and its telecommunications arm, the Defense Information Systems Agency, embraces EOIP and has a formal migration plan that will have all DOD telecommunications moved to IP by the second half of the next decade — only a few years away.

Convergence is a related buzzword describing how former circuit-based communications, such as telephone systems (and their switching centers) and Frequency-Division Multiplexed (FDM) communications are moving to an integrated, IP-based, EOIP architecture.

Many believe that EOIP is the Holy Grail of integrated telecommunications. Yet, there is a second side to this architecture that deserves examination: not EOIP but its obverse: IPOE or IP over everything.

While EOIP encourages the development of native IP-based applications for telephones, TVs, radios, and computer programs that users interact with, IPOE encourages native IP-based — or at least IP-compliant — telecommunications transport. For example, traditional copper telephone wiring degrades IP packets too much to be useful for IP communications over more than 300 feet. Radio signals are subject to fade, echo and unintentional interference, so they have to be specially designed for IP communications. Even fiber-optic transmissions have special repeaters every 30 miles or so to amplify and clean up the data packet before sending it on.

But satellite communications have presented a special problem. Most modern satellite communications systems rely on geosynchronous satellites, placed in orbit at 22,250 miles above the equator. At this altitude, the satellite orbits the earth in exactly 24 hours, appearing to be fixed in the sky. Because these satellites don’t move relative to an observer on Earth, the antenna on them don't need to move, making them a relatively inexpensive receiver system.

The problem is that even at the speed of light, 186,000 miles per second, the signal transmitted from the ground has to travel up to the satellite — 22,250 miles or more, depending on where the transmitter is — and then back down, or about a quarter-second round trip. A reply signal must retrace those steps, resulting in the pause people commonly experience when being interviewed on TV over a satellite link. And of course, there are other factors in the communications chain that can compound signal delays, sometimes up to a second or more, each way.

This delay wreaks havoc with IP communications and has been one of the main stumbling blocks for greater adoption of IP over satellite. There are two main approaches for solving the inefficiencies imposed by this delay: acceleration and compression.

Acceleration, also known as TCP spoofing, was first demonstrated commercially by the U.S.-based Comsat Corp. in the 1990s. Spoofing strips out the sender’s acknowledgment packets from the IP data stream before the data is sent to the satellite, and sends acknowledgments back to the sender, as if the distant receiver really had gotten the data. In this way, the sender doesn’t have to wait the half second or more for the acknowledgment and feeds data much faster. Without this spoofing technique, an IP data circuit has a limit of approximately 800 kilobytes/sec. However, by using spoofing, data transfer rates are only limited by the capabilities of the satellite, typically in the megabytes-per-second range.

Compression, on the other hand, replaces a series of bits in the data stream by smaller, prearranged characters. For example, if both the sender and the receiver agreed that the word “satellite” could be replaced in a sentence by the symbols “!&C”, we could send the word satellite in about one-third of the time with no loss of intelligibility. There are many data compression algorithms in use today, and DOD will be adopting data compression standards for the new Joint IP Modem, scheduled to be deployed worldwide in the next few years.

In addition to compression and acceleration, we are able to use IP-over-satcom to significantly reduce the amount of satellite power required to interconnect multiple sites. Instead of having a separate circuit from each sender to each receiver, IP-over-satcom allows DOD to create an on-demand IP-based network in the sky. Future DOD users will simply turn on their satellite terminals — much as how we turn on our cell phones — and the terminals will register themselves automatically and securely into the DOD satcom network. Because the terminals all talk to one another, we’ll have greater redundancy and be able to manage the network with greater assurance and reliability than before. Because these IP-based terminals will be highly automated, changing the parameters of the terminals will be accomplished in minutes instead of hours.

Satellite communications technologies are about to move forward in ways undreamed of only a few years ago. IP over Everything will mean the final communications convergence, leading to seamlessly integrated data moving all over the globe. That’s great news for DOD and everyone depending on satellites for communication.