Work is underway to establish the first interplanetary laser communication link. The $300 million NASA experiment, if successful, will connect robotic spacecraft at Mars with scientists back on Earth via a beam of light traveling some 300 million kilometers.

For scientists eager to download bandwidth-intensive imagery and other data collected by planetary orbiters, probes and landers, the laser communications would offer a dramatic breakthrough in the amounts of information spacecraft can reliably transmit back to Earth.

The 5-watt laser NASA plans to test at Mars by the end of the decade is expected to transmit data at rates nearly 10 times faster than any existing interplanetary radio communications link. The difference, NASA officials said, will be comparable to moving from a dial-up modem to a broadband Internet connection.

But the new technology is not without its challenges and NASA says it could be decades before lasers are ready to take over as the primary means of communicating with spacecraft.

The U.S. military has plans to field a constellation of optical communications relay satellites in Earth orbit starting around 2012. Those satellites are intended to help the Pentagon deal with a bandwidth crunch that has been heightened in part by a growing fleet of unmanned aerial vehicles that are transmitting data-rich imagery. NASA faces a bandwidth crunch of its own in deep space as more powerful spacecraft and instruments become reality. Highly reliable data links with fast transmission rates also are deemed critical to the human planetary expeditions NASA hopes to undertake.

NASA is tackling some of the technical challenges facing interplanetary optical communications with the Mars Laser Communication Demonstration (MLCD) now in development at NASA's Goddard Space Flight Center, the Jet Propulsion Laboratory and MIT's Lincoln Laboratory.

The MLCD payload is slated to fly aboard the Mars Telecommunications Orbiter, a $500 million telecommunications relay satellite NASA intends to launch in 2009. The experiment would begin in 2010 when the spacecraft arrives at Mars and run for at least a year.

Rick Fitzgerald, the MLCD project manager at Goddard, said the Mars Telecommunications Orbiter's laser payload will be the first time that such a device has ever flown in deep space. He said one of the trickiest parts of the experiment will be accurately targeting the laser beam.

"The tough part is pointing," Fitzgerald said. "Being 2.3 astronomical units away from Earth" -- about 344 million kilometers -- "even the tiniest pointing error can send you way off into space. A big part of the challenge is being able to point accurately toward the Earth."

Unlike radio frequency signals that wash over the entire Earth, Fitzgerald and his colleagues will be shooting for a much smaller target - the southwestern corner of the United States.

A ground-based receiving station is planned for the Palomar Observatory north of San Diego, Calif., with a secondary site possibly in New Mexico or Arizona.

Fitzgerald said the team plans to retrofit Palomar's 5-meter Hale Telescope with an optical receiver to pick up laser transmissions from Mars. NASA also is looking for a secondary site to test whether laser transmissions can be received using an array of several much smaller optical telescopes. NASA also plans to place an uplink beacon atop Table Mountain near JPL to help the laser payload find its target.

Unlike the radio frequency signals NASA has used for decades to communicate with spacecraft, laser transmissions cannot penetrate heavy cloud cover. Steve Townes, the MLCD deputy project manager at the Jet Propulsion Laboratory in Pasadena, Calif., said the team considered positioning receivers on high-altitude balloons or on Earth-orbiting spacecraft, but budget and schedule constraints drove the decision to put the receivers on the ground.

An operational laser communications system, Townes said, likely would require receivers positioned above the atmosphere or building network of ground receivers at different location around the globe to help reduce weather limitations.

Once the Mars Telecommunications Orbiter is in place around Mars, the MLCD team plans to begin their experiment by transmitting test data back to Earth. The team expects to be able receive data at a rate of 1 million bits per second when Mars is at its furthest point from Earth and reception is occurring during the day. When Mars is at its closest approach and reception is taking place at night, however, the team expects to receive at a rate of 10 million bits per second and perhaps as high as 30 million bits under certain optimal conditions.

NASA's Mars Odyssey orbiter, in contrast, transmits data at about 128,000 bits per second, or about twice as fast as a dial-up connection but a tenth the speed of the typical broadband Internet connection.

If the MLCD proves its mettle as a communications link, it could be called into service to help transmit science data from NASA assets on the ground such as the Mars Science Lander, a nuclear-powered rover expected to arrive at Mars the same year as the telecommunications orbiter. Fitzgerald said the MLCD team is working out an agreement with the Mars Telecommunications Orbiter team to transmit science data back to Earth in parallel with the orbiter's radio frequency bands.

Ramon De Paula, the NASA program executive for the 2009 Mars Telecommunications Orbiter, said MLCD represents another step in the evolution of deep space communications. The Mars Reconnaissance Orbiter launching in August has better radio frequency transmission capabilities than Mars Odyssey, which relayed over 95 percent of the data collected by the twin Mars Exploration Rovers through the tried-and-true X-band frequency and further convinced NASA that a dedicated telecommunications satellite was a good idea.

The 2009 Mars Telecommunications Orbiter is expected to be something of a communications marvel even without its experimental laser transmitter.

De Paula said the telecommunications orbiter will transmit in Ka-band, a frequency NASA pioneered in the early 1990s but one that has yet to come into its own as a primary downlink for spacecraft.

The spacecraft also will be equipped with an X-band transmitter, but De Paula said the intention is to use it primarily as a back up. A Ka-band system flew aboard NASA's experimental Deep Space 1 probe in 1998 and will get another workout aboard Mars Reconnaissance Orbiter.

De Paula said he does not expect to see the first fully optical spacecraft for many more years to come. "It will have taken 25 years to make Ka-band operational," he said. "It will probably take 20 to 25 years beyond 2009 to make optical a fully operational frequency."