provide a standards-compliant (IEEE 802.11) wireless LAN access network that linked the command and control (C2) computers, hand-held computers, and so forth, into this all-encompassing network.
A number of technical issues were uncovered during the first ELB ACTD, but the operational results appeared promising despite the technical problems. Thus, a second ELB ACTD has been scheduled that will fix some of the obvious technical problems with the communications network and allow further operational experiments. This new experiment will also introduce newer C2 systems such as the integrated Marine multiagent command and control system, laptop servers down to the company level, multiple battlespace views, and interfaces to other (joint) systems such as Army tactical operations centers and the naval fires network.
The schedule features a number of shake-down experiments and tests that include FST-1 in June 2000 (focusing on communications and networking), FST-2 in conjunction with Millennium Dragon in September 2000, FST-3 in February 2001 (focusing on rehearsal), and the actual experiment, conducted in conjunction with Kernel Blitz (X) in June 2001.
The committee believes that the ELB ACTD provides a striking mixture of good news and bad news. The good news is mainly operational—ELB has provided the first glimpse of how an OMFTS information infrastructure might work in practice. The bad news is entirely technical—the ELB network is an early prototype that is many years, and many tens of millions of dollars, away from a real production version of a tactical data network.
First, the good news:
The RF links worked. The ACTD established link ranges in excess of 100 nautical miles with a 10-W power amplifier, had no trouble with Doppler shifts in air-to-air linking, and could deliver reasonable data rates (from 625 Kbps at 110 nautical miles up to 10 Mbps at 30 nautical miles).
The multihop network showed promise. Network connections were established between El Centro, California, and ships at sea, and realistic applications were able to work across this network. The basic division of the network into a backbone and a set of access networks worked well.
The experiment was the Marines’ first real glimpse of how an OMFTS information infrastructure will work in a network-centric future. As such, it gave very valuable insights into the basic operational concept and demonstrated that, to a very large extent, the concept worked.
And now the bad news:
The network suffered from a number of serious technical defects. It was often unstable and will likely have serious issues with scalability to larger numbers of network nodes (radios).
The observed decrease in effective bandwidth as a function of range is a consequence of the decreasing signal-to-noise ratio as a function of range. Simply put, unless something can be done to maintain the effective system bandwidth, the system will only be able to support 6.25 percent as many users at a range of 110 nautical miles as can be supported at a range of 30 nautical miles.
Applications and protocols designed for a highly stable campus LAN (e.g., an Ethernet) may tend to “commit suicide” in a tactical RF network. Congestion and changes in connectivity are much more severe than those encountered in most commercial environments, and a typical application program behaves very poorly under such circumstances.