This is an old revision of the document!


GWEN

GWEN is the Ground Wave Emergency Network, commissioned as AN/URC-117. It was part of the early Minimum Essential Emergency Communications Network (MEECN). The primary purpose of GWEN was to relay messages between SAC sites, including emergency action messages.

The LF range used by GWEN, at around 150kHz, could propagate over long distances via ground-wave, where it would receive minimal interference from EMPs in the upper atmosphere. GWEN originated in late '70s with an experimental system built at Kirtland Air Force Base - this is local to me, but I have been unsuccessful in finding detailed information about this system. By the late '80s the GWEN system design was completed and it entered the deployment stage, but it reached only partial operating capability for a few years in the early '90s before being cancelled with fewer than a quarter of the planned sites constructed.

Relay Sites

GWEN consists of two types of sites: relay sites and input-output sites (if you consider receive-only devices to be a special type of input-output). Input-output sites were compact and located at various military installations and other facilities, or installed into aircraft. Input-output sites used the lower end of UHF to communicate the nearest relay site, typically over short range or line of sight. The relay site was then responsible for carrying the message throughout the system using a means more resistant to interference (particularly airburst EMP) than UHF.

GWEN relay sites were generally unmanned, but many relay sites were co-sited with other military facilities, particularly air force bases and army depots. Others were on purchased or leased property, although likely for ease of maintenance and proximity to input-output terminals they were usually not very remote.

To the right is an illustration of a typical GWEN site found in the Environmental Assessments for each site. The main feature of the GWEN site is the LF transmit antenna, which consists of a 299' tower with guy wires in a triangular layout and a much larger number of antenna elements running from the peak of the tower to the ground in cone about 600' across, putting the antenna elements at a 45 degree angle to the ground. Under the LF transmit antenna are copper ground plane elements running out to about the same 300' perimeter, this entire area is typically fenced in barbed wire to prevent disturbance of the ground by cattle or, perhaps more of a problem, wayward farmers. A small equipment area at the bottom of the LF transmit antenna had a single small hut containing the antenna tuner.

At a corner of the site, a short distance outside of the 600' across antenna area, there is a small fenced yard with two huts and a short wooden pole with the LF receive antenna (two magnetic loops) and the UHF antenna for communicating with input-output stations. One hut contains the radio equipment and the other a backup generator. This area is the most useful feature to identify GWEN sites as it has a very consistent appearance from above. The gravel road running from the equipment area to the base of the LF transmit antenna is also distinctive but is often not visible today.

It is clear from environmental assessments that GWEN relay sites required a telephone connection, but no document I have seen elaborates on the reason. GWEN relay sites could send various automated messages relating to maintenance needs and alarms, to conserve GWEN resources these may have been sent over telephone instead of GWEN itself.

Construction

The first nine GWEN sites were built by R&D Associates of Albuquerque as a prototype system. After these sites, the next set were built out to form the Thin Line Communications Capability (TLCC), which I am not sure of the exact nature of. The final operational capability was never constructed.

Protocol

GWEN incorporated an interesting architecture: GWEN sites included an HF radio which would be used for local link to the site, and then an LF radio with associated large ground-plane antenna to contact other nearby GWEN nodes. GWEN operated in digital mode using packet switching, allowing messages to be routed around inoperable or unreachable sites. I have not been able to find detailed information on the digital scheme besides that it used a fixed 2/3rd second transmit/receive cycle, with each 644-byte packet transmitted twice using MSK at 1,200 baud… or possibly 75 baud, depending on what document you infer from. 1,200 baud seems most likely since it would allow for a 644-byte packet plus padding during the 2/3rd second cycle, assuming that each character is a byte (not completely clear). One document (a Forecast International report) suggests that GWEN used a simple “flood” routing strategy where every packet was directed to every nearby node, but there is no mention of how loops are prevented. I suspect the document is simply incorrect on that point.

Wikipedia gives an interesting lead: 32 GWEN sites were reused for the National Differential GPS System, so an environmental impact assessment for that system lists their locations. The NDGPS system used long-wave as well for the data broadcast, allowing reuse of the GWEN transmitter and antenna. Of course, this is only 32 of the 58 sites which were supposedly constructed. Unfortunately, the 1994 decommissioning of the remaining sites may be too recent for FOIA exemptions on more complete records to have expired. I continue to occasionally dig for anything I can find about GWEN, but have not yet gotten to submitting requests under FOIA.

I recently made a major breakthrough on finding these sites that seems obvious in hindsight: all of the GWEN-turned-DGPS sites had FCC ASRs under FRN 0015880685, US Coast Guard Navigation Center. Searching the ASR database by this FRN turns up 35 registrations, almost all of which I've confirmed from imagery are GWEN sites - the remaining few either don't seem to exist on the ground at all (these need further research since it seems unlikely the Coast Guard is maintaining registrations for non-existent sites) or don't appear to be GWEN sites, suggesting they were built by the Coast Guard itself before they started to reuse GWEN.

I have made a KMZ file of all of the sites located by this method.

The process of the Coast Guard and Department of Transportation taking over GWEN sites as part of creating national inland coverage for the Differential GPS system is itself interesting. The DGPS correction signal was transmitted between 285 and 325 kHz, close enough to GWEN's operating frequency that the LF transmit antenna could be reused. To reduce the cost of national DGPS rollout, the Department of Transportation (which sponsored the effort) and the US Coast Guard (which was responsible for maintenance of the DGPS system) slowly collected the original equipment shelters from every GWEN site, refurbished them at a central Coast Guard shop, and then shipped them out for use at GWEN sites or newly built NDGPS stations. As a result, all former GWEN sites have original GWEN equipment huts, although they may not be original to that specific site! GWEN sites that were not selected as NDGPS sites will be missing their equipment huts, but you're unlikely to see that as documents suggest that the Air Force demolished all of them.

Throughout its short life, GWEN was hounded by public concerns around the health impacts of powerful LF transmitters. This concern appears to have been quite misguided, particularly considering the similarity of GWEN transmissions to pervasive and often higher-powered commercial AM stations, but the concern was sufficient to prompt a substantial National Research Council evaluation. This public upset has transformed, in the decades since GWEN's deactivation, into an outright conspiracy theory; casual internet searches related to GWEN turn up mostly blogs accusing modern cellular base stations of being clandestine GWEN nodes intended to harm human health or assist in some other evil deed.