Pervasive Sensing Holds the Key To Network Centric Operations

By HUNTER C. KEETER
Associate Editor

The promising idea of network centric warfare presents some daunting challenges, including the requirement for fetching a few bytes of good quality information from oceans of data to support decision-making. Military commanders increasingly require in-depth knowledge of the enemy’s intentions to stay ahead of them in operations that are ever more fast-paced and intense.

Network centric warfare refers to joining the many parts of a military force through shared command, control and communications systems. So joined, these parts become a more efficient and effective whole. Information is the linchpin to networked operations.

In July 8 testimony before the Senate Armed Services Committee, Adm. Vern Clark, chief of naval operations, advocated “persistent knowledge” of the battlespace as a key enabler to future military success.

But information from persistent sensing capabilities is not enough to solve the data fusion and information management challenges of network centric warfare. Some officials argue that it is the pervasiveness — the depth — of that sensing capability that is the discriminating factor.

Pervasive sensing means flooding an area with sensors — from the sea bed to space — that an adversary cannot detect, but that operate well behind their lines gathering information with an up-close-and-personal perspective.

According to intelligence expert Wayne Perras, of the Navy Warfare Development Command in Newport, R.I., effective information comes from inside an adversary’s operations and thought processes.

“There is tension between the terms persistence and pervasive,” Perras told Sea Power. “It is good to be persistent; it is necessary, but it is not sufficient. You must be pervasive. The only reason to gather [intelligence, surveillance and reconnaissance] is to deeply penetrate the enemy. For example, space-based radar is great. It is on all the time; it covers the area, but does it provide all the data a commander needs in terms of deep penetration? Can the radar data show what an enemy desires to do?”

Technology will provide part of the answer, according to Dr. Dennis M. Bushnell, chief scientist at NASA’s Langley Research Center, near Norfolk, Va.

“The future is clear. Ever-smaller [sensors] will offer better sensitivity and capabilities across multiple physical spectra,” Bushnell told Sea Power. “What is emerging is a global sensor grid.”

The grid’s sensors could communicate with one another over wireless data links, forming a self-calibrating, self-correcting web. The sensor nodes could be of almost any sort — measuring physical phenomena from radio frequency output and electro-optical and infrared spectra, to chemical and biological agents, and seismic variations. These nodes could be emplaced by humans, such as Special Forces troops, or dispersed from manned and unmanned aircraft.

Advances in microtechnology — the miniaturization of electronics and power sources — are at the heart of future pervasive sensor networking. Perhaps not surprisingly, much of the leading-edge work being done in micro electro-mechanical systems is happening in the private sector.

One firm, Dust Networks, of Berkeley, Calif., plans its first release this month in a new line of tiny sensor products aimed at supporting building infrastructure automation and monitoring, as well as potential defense applications.

In 2001, Dust Networks founder, Dr. Kris Pister, was part of a University of California at Berkeley team that worked with the Defense Advanced Research Projects Agency (DARPA) and others on a project called “Smart Dust” to develop a “robust, self-configuring, self-organizing wireless sensor network” for future battlefields.

Smart Dust was envisioned in the form of tiny micro electro-mechanical systems, like grains of sand, each with the ability to collect measurements and transmit data over a wireless network.

Testing at U.S. Marine Corps Base Twentynine Palms, Calif., in March 2001 included the successful delivery from an unmanned air vehicle of prototype Smart Dust “motes,” battery-powered circuit cards that were surrogates for the smart dust particles.

Other projects envision pervasive sensing networks that are capable of offensive action. While DARPA officials declined requests for interviews on the subject of networked sensing, the agency and several contractors are developing a sensing and electronic warfare system called Wolfpack.

BAE Systems is one of the firms developing “wolves,” unattended radio frequency sensor nodes at the heart of the Wolfpack network. The initial version would include as many as 10 wolves that are deployed in an area to detect, track and even jam radio emissions — in the 30 MHz to 20 GHz range.

Designed for a 60-day mission life using commercially available electronics and battery technology, the Wolfpack nodes could process signals received from threat systems — such as a Scud missile transporter, erector or launcher — and distribute geo-location information over a wireless network. That geo-location information could be used to program aim points for precision guided-munitions, accurate to within a 10-meter circular error probable.

Don Snelgrove, BAE’s director of Wolfpack programs, told Sea Power that as technologies advance — especially with regard to batteries and other power sources — networks of unattended sensors could be placed on land or beneath the ocean to act as persistent conduits for signals intelligence.

“The advantage that you get by being up close and small is that you are surreptitious. No one knows you are there, while you have the ability to listen in on all sorts of signals,” Snelgrove said.

Future iterations of the unattended sensor networking architecture envisioned for Wolfpack may monitor other types of phenomena, such as electro-optical-infrared, acoustic, seismic, chemical and biological traces.

Though almost daily advances push back the limits of sensing capabilities, important challenges remain. For example, how costly are these networks going to be to set up and to maintain? What are the rates of attrition and replacement for the sensor nodes on these networks? There are not yet enough economic analyses to answer these questions, according to Perras.

Dr. Toshio Hori, of Japan’s National Institute of Advanced Industrial Science and Technology, Digital Human Research Center, wrote on his institution’s website that several technological breakthroughs also are just out of reach at present. These include, “small and cost-effective sensor modules; high speed, low latency and reliable network infrastructures … and sensor information processing technologies.”

These breakthroughs would enable the development of smaller and more capable sensor units, as well as the links that network many sensor nodes spread throughout an area.

Another challenge concerns the role of humans in information processing — the critical task of converting sensor data into actionable knowledge. Bushnell noted that the weak link in the chain is the human being.

“The abilities of a human being to make sense out of all this information are far slower than those of the machines,” Bushnell said. “The question is going to be, increasingly, how much can we stand the latency of human beings in the network?”