cartome.org

1 May 2002


Source: http://www.us.net/signal/CurrentIssue/May02/persistent-may.html

 

May 2002 � SIGNAL Magazine 2002

Persistent Surveillance Comes Into View

Building a warfighting network is only the beginning; its capabilities mandate a new way of doing business.

By Robert K. Ackerman

The next step in network-centric warfare will be the creation of networked sensing suites that tailor their observations to the adversary's rate of activity. These various sensors will concentrate on observing changes rather than on observing scenery.

The goal is to combine the broad spectrum of current and future sensors into an effective intelligence tool that is geared to the activity of an adversary. The amassed information will be input into a ubiquitous high-speed Internet protocol network that employs advanced fiber optics and satellite laser links to connect network-centric warriors throughout the battlespace.

This approach also involves a paradigm shift in how data is entered into the network. Instead of analysts processing raw data into information for input into the network, the raw data will be placed on the network for empowered users to exploit for their own particular requirements. The decision on what is important will move from the entity that captures the data to the person who uses it.

All of these characteristics fall under the principle of persistent surveillance, which is the focal point of the U.S. Defense Department's thrust to fully exploit network-centric warfare. John P. Stenbit, assistant secretary of defense for command, control, communications and intelligence, sees it as the next key step in advancing the digitized battlespace.

Stenbit emphasizes that persistent surveillance is not the same as constant surveillance. "The ability to observe everything continuously, all the time, would be awful," he declares. "Who wants to see all the world all the time with enough resolution to find every person? What would we do with it?"

The defining concept of persistent surveillance is time. Someone observing a scene need not re-examine it any faster than the period of time in which changes take place. "We spend a lot of time taking pictures of radars and other electronics," he notes. "But, as long as things don't change, you don't have to take another picture because you know everything that is there."

It is when dealing with targets that do change that the observation interval must be coordinated with the frequency of those changes. "If what you're interested in has time associated with it, unless you change the rapidity with which you come back and look, you are going to miss something. You will have a picture in your mind that is no longer true," he states.

He likens this pitfall to having a traffic reporter in a helicopter describe urban traffic patterns all day long based on a single observation of empty streets at 4:00 in the morning. That report would be hopelessly inaccurate throughout the day as traffic piled up. On the other hand, a reporter who observed traffic every hour would be considerably more accurate. And, a reporter who observed this traffic all the time would be the most accurate of all.

The approach is to maintain rates of surveillance that fall within the rate of change for the target. "Persistence in our context is to match the frequency of revisit with the time stability of the object that you are looking at--the speed with which things change," Stenbit clarifies. He adds that this is crucial because the military increasingly is interested in targets that move or are fleeting. Their value changes with time, such as a missile launch site that is of less interest when empty.

One issue is how do we meld persistence with other things to cause a symbiotic relationship," Stenbit offers. "If you know that your surveillance will be there at the right time, you can create environments that make the observation more interesting." This might include lofting flares to create a better picture, for example, or causing a disturbance to trigger a reaction that reveals more information for the sensor. Aircraft flights that trick an enemy into switching on their air defense radars are an example of this approach. Combining it with persistent surveillance can produce the maximum return available.

"Persistence provides more flexibility to plan combined operations," he says. "Because you are more likely to be able to plan to be there at the right time, you are more likely to be able to have other things happen within the time scale you are worried about to make the surveillance more interesting."

Unmanned aerial vehicles (UAVs) have proved to be key enablers for persistence, Stenbit observes. And, this has been achieved without placing people in jeopardy. Improved communications capabilities, especially in bandwidth, have opened the door for this application.

Stenbit points out that the concept of persistent surveillance represents a significant difference between the needs of the military and those of the intelligence community. Where the intelligence community tends to lean toward "every so often," the military generally is interested in "whatever 'now' is defined as," he posits.

The same difference applies to many existing intelligence disciplines. Signals intelligence, or SIGINT, tends to occupy the "now" part of the observation spectrum, Stenbit notes. In effect, the military has been engaging in persistent surveillance with SIGINT for a long time, as detectors and receivers constantly monitor the ether for revealing signals. What has changed, he explains, is everyone's perception of time.

The persistent surveillance approach also addresses the growing need for specific battlefield intelligence, particularly through innovative approaches. Stenbit charges that the United States has "under-invested" in new ways to obtain secrets. Over the past 15 years, the country has allocated funds to improve the ways it has performed surveillance and collected intelligence. Collection now occurs faster and more often, with more and better images amid increased bandwidth. However, it has not substantially changed its activities during this time. And, this is both a drawback to the United States and an asset to a potential adversary.

"Every time we use a technique to do something effective to somebody, we depreciate the value of that technique," he says. That adversary, after a while, can begin to associate U.S. activities and take measures accordingly. For example, a foe that learns of a U.S. capability to monitor a certain type of communications can cease using that medium and communicate in a more secure manner, especially if it is less technologically sophisticated. This effectively negates a key U.S. capability, which is why Stenbit emphasizes the need for new methods of intelligence collection.

"The construct of being able to continually do things differently from what they expect is an investment in our ability to surprise them," he emphasizes.

During the 1980s Iran-Iraq War, the United States provided Iraq with various forms of assistance, including intelligence. Iraq thus learned some of the U.S. capabilities in this arena. When the U.S.-led coalition faced Iraq in the Gulf War, Iraq had changed many of its methodologies to counter U.S. capabilities it had learned from the earlier assistance. It had not learned everything about U.S. capabilities, and this contributed to its military's downfall. However, Iraq has continued to incorporate lessons it has learned since then, along with lessons from Bosnia, Kosovo and Afghanistan.

Persistent surveillance is one way for the United States to recapitalize its intelligence collection repertoire, Stenbit offers. "The game over the past few years has been how fast do we look, versus how fast do they move, versus how fast do we plan."

He cites as an example the time cycle for the U.S. Air Force's air tasking orders (ATOs). Upon learning that this cycle was 72 hours, adversaries would move their forces every 60 hours. When that time eventually was reduced to 40 hours, foes would move their forces every 30 hours. Reducing this time further would, for the most part, simply impel an adversary to move forces more often. The solution, Stenbit notes, is to change the way an air war is fought.

"You change by flying aircraft around in circles in the ATO, in which case you are ready to attack within five minutes," he says. "Now, it is not an ATO anymore--it's an ATO with ad hoc adjuncts.

"The time dimension changes as events go on, and you must continue to surprise people," Stenbit concludes.

All of these aspects feed into the current transformation to network-centric warfare. Stenbit's view of the network-centric future calls for advanced fiber-optic networks on the ground and laser communications in space to achieve a ubiquitous wideband Internet protocol (IP)-based network.

He continues that his goal is to incorporate dense wave-division multiplex fiber optics into the battlefield network. This would introduce an optically based ground network that features up to 100 different fiber optic colors, and each major command or agency would have its own color. This IP-based worldwide network would move information at about 10 gigabits per second.

Moving information out of the ground network would be the next element. This would involve satellite laser communications comprising three types of links--satellite to and from groundstation/gateway, satellite to satellite and aircraft to satellite.

The aircraft links would permit transmitting imagery via satellite into the network. Satellite-to-ground laser links might not always be feasible, depending on conditions, so the satellite-to-satellite laser hookups could serve to move information at high transfer rates to another military communications satellite such as the Advanced EHF (extremely high frequency) satellite (SIGNAL, February, page 25) for conventional transmission to groundstation gateways.

Building this network is only part of the challenge. Stenbit points out that having the advanced capabilities enabled by this ubiquitous wideband network will require more sophisticated workstations than currently planned. "Once you have all of these things, how do they collaborate? You need collaboration tools for huge collaboration," he predicts.

Other emerging issues include distributed storage and registration, automatic metadata tagging and IP browsing. "The technological issues needed to make it as useful as it could be--as opposed to just more useful than it is today--will involve amazing technologies that I don't have the answers to," he declares.

This network will enable a broader use of different sensor data, which in turn lessens the degree of persistence necessary for effective surveillance. "If you have a ubiquitous network, you could have whatever images you have of whatever sort," Stenbit posits. Gun camera imagery, for example, is not thought of in other terms. Yet, it can be a valuable source of information if the data can be moved to an image interpreter.

"In a networked world, the actual requirement for the optimum amount of persistence goes down," Stenbit asserts. "[This is] because you are able to share the information over a broader context, and you may not need the special information as often. When you do need it, you will be able to know it faster.

"You need slow sources so that you can detect when the timing is changing [so that] you can allocate fast sources," he allows. "But, you cannot do any of that unless you are in the networked world. What you want is an adaptive system."

Achieving the ultimate adaptive system will require both technological and cultural adjustments. Stenbit describes this ultimate network as one that is trusted, reliable and ubiquitous. The cultural change must come in the form of users posting information on the network before they process it.

This is a key issue, Stenbit offers. "You have to unlock the 'knowledge is power' business," he says. One user's view of information may be different from another's, and early processing of the information would tilt it toward the first view. So, the data would have to be entered in an unprocessed form so that other users could apply it as needed. "The big cultural issue is 'host before you process,'" he states. "The reason to get to the networked world is to break away the determination of what's important from the person who has the information into the hands of the person who uses the information."