IM/EMI Issues, a conflict of public interest
Police, fire and emergency medical services agencies depend on radio communications to protect people and property. Any type of interference to the operation
March 1, 1999
Police, fire and emergency medical services agencies depend on radio communications to protect people and property. Any type of interference to the operation of dedicated public safety frequencies diminishes an agency’s ability to provide service, and it endangers the public. A situation we encountered in Northwest Oregon in 1998 demonstrated that commercial mobile radio services (CMRS), particularly enhanced specialized mobile radio (EMSR), through frequency use, site placement and output power, can seriously interfere with nearby public service communications.
Operational overview Washington County includes the Tualatin River Valley and mountainous terrain on the western outskirts of the Portland metropolitan area in Northwest Oregon. The Washington County Consolidated Communications Agency (WCCCA) provides emergency dispatch radio communications for 21 public safety agencies in Washington County, as shown in Figure 1, below right. Washington County covers 842 square miles, with a population of 350,000. In 1991, voters approved a levy to construct a new 9-1-1 system that included a Motorola Smartnet trunking system. Because of limited resources and limited available frequencies, a 10-channel trunked simulcast system was chosen. This system would give WCCCA maximum area coverage with a minimum number of frequencies. WCCCA constructed four simulcast sites with 10 frequencies each. Given the large area that had to be covered, four sites were strategically chosen and, as with traditional radio systems, the repeater sites were installed on mountaintops for maximum coverage. (See Figure 2 on page 18.) With this design, WCCCA can provide radio service for more than 2,000 users, covering about 63% of Washington County. This system gives WCCCA good mobile coverage and fair portable coverage on the valley floor, which contains the bulk of the population.
The one big limitation of WCCCA’s system is good portable radio coverage inside certain buildings. To overcome this building penetration problem, WCCCA has installed bi-directional amplifiers in many buildings, such as the county jail. Some high-traffic buildings, such as the Washington Square shopping mall, had to be excluded from in-building coverage. Due to the extreme expense and difficulty in establishing who would be the responsible party, no bi-directional amplifiers were ever installed, so presently there is marginal public safety coverage in this commercial area.
Defining interference There are several types of radio interference. Co-channel interference occurs between similar frequency channels used in the same geographic area. Adjacent-channel interference can be caused by a transmitter operating on a channel bordering either side of the channel in question. The transmitter sidebands mix with the carrier being received on the desired channel, creating noise or “splatter.” Interference can also be caused by electromagnetic interruptions. Harmful interference has a specific connotation. As defined by the Communications Standard Dictionary, it is
“…any emission, radiation or induction that endangers the functioning or seriously degrades obstructs, or repeatedly interrupts a communication service, such as a radio navigation service, a search and rescue communications, or a weather service. It is assumed that these services are operating in accordance with approved standards, regulations and procedures. Harmful interference causes circuit outages and message losses, as opposed to interference that is merely a nuisance or annoyance that can be overcome by appropriate measures. In order to be harmful interference, it must seriously degrade the performance of the communications, radar, or other electrical or electronic system.”
Interference with public safety communications from properly operated commercial radio services has traditionally not been a problem in our operating area. As with our own system, land mobile repeater sites were located on mountaintops and end users were on the valley floor. When the 800MHz (824MHz-849MHz and 869MHz-894MHz bands) cellular service was introduced, with cell sites located on the valley floor, there were still few problems. Public safety 800MHz bands (806MHz-821MHz and 851MHz-866MHz) enjoyed adequate separation from cellular, which was also grouped in one bandpass, while public safety was grouped in another. This situation changed with the introduction of 800MHz ESMR into the valley.
Enter ESMR The FCC band plan for the 1998 800MHz SMR auctions divided the band into three pools: a public safety, industrial/land transportation and business pool, a general category pool, and an SMR category pool. The pools are interspersed from 810.25MHz to 816MHz and from 855.25MHz to 861MHz. In the FCC’s 800MHz auction that closed in December 1997, McLean, VA-based Nextel Communications was the high bidder for the Portland-Salem, OR, economic area, which at that time encompassed an estimated 2,310,060 pops. The operator began service in 1998, placing 800MHz (806MHz-821MHz and 851MHz-866MHz bands) cell sites on the valley floor, among public safety users. (See Figure 3 on page 20.) As shown in Figure 4 on page 20, the closest commercial and public safety frequencies come to each other is four 25kHz channels away, so there is no possibility of adjacent-channel interference. However, because the band plan for 800MHz intermixes commercial frequencies with public safety frequencies, there is no chance of any group bandpass filters, as in the case of cellular. A problem was brewing, as we soon found out.
When we get behind closed doors-only The first incident to bring an interference issue to the attention of WCCCA was a radio service complaint from one of the fire departments. This fire station, one of 33 in the county, is designated Fire Station 253. It serves one of the busiest areas of the county because it is next to a major freeway and next to the Washington Square shopping mall, where, as mentioned before, we have no in-building coverage.
The engine company is equipped with an 800MHz Motorola Spectra radio, an 800MHz Motorola VRM 600/DATA 911 mobile data terminal (MDT) and several 800MHz Motorola MTS2000 portable radios. The station is also equipped with an 800MHz base radio used for “tapping out” the station. (“Tapping out,” a carry-over term from the old days of alerting stations by Morse telegraph, is the emergency alerting system that rings the station bells and turns on the lights.)
The engine company first complained to WCCCA Technical Services about poor radio communications. Its main complaint was that it was difficult to hear what fire dispatch was saying. This created a life-threatening situation because Station 253 was not always getting tapped out. The base radio was not always responding to the alerting signal.
WCCCA’s technical staff began troubleshooting the problem. The usual procedures, such as checking antenna and receiver performance, revealed no evidence of a problem. Attention was then directed to the trunking system itself. At first, the WCCCA technicians thought there might be a simulcast phasing error. This, too, proved to be a dead end. The next area of investigation was determining the signal strength at, and around, Station 253.
The expected signal strength level should be about 265dBm at Station 253, This is based on calculations for ideal conditions (our closest trunking site is only 4.4 miles away), with an average of 58W ERP and with 9dB antenna gain and 38 antenna downtilt. As shown in Figure 5 on page 22, a Z-Technology R-505 field strength meter (range, 0dBuV to 1110dBuV; accuracy 62dB), a GPS receiver and a notebook computer were mounted in a jeep to log field strength readings of WCCCA’s control channel #1 (860.7375MHz). (See Figure 6 on page 22) Computer analysis and mapping of signals was performed with the STI-9000 Mobile Signal Analysis System from Survey Technologies. At first, the field strength reading around the station did not make a lot of sense. Some areas had the expected field strength level, and then there were areas where the signal dropped off to nothing. At first, WCCCA thought there was a signal hole in the simulcast system coverage. Because Station 253 has a relatively low elevation, WCCCA came to believe that this might be normal. However, because this involved public safety communications, WCCCA felt that it needed a more definitive explanation.
Two key issues led WCCCA to start suspecting an “outside” interference problem. First, it was discovered that the portable radios worked better inside the engine bay with the bay door closed. Second, the base radio got a more intelligible signal from fire dispatch using an inside antenna than when it was connected to the main outside antenna.
Noisy neighbors WCCCA’s investigation of other radio systems in the area discovered that a Nextel Communications site was located about 1/4 mile to the west of Station 253, as shown in Figure 7 on page 24. The site uses Motorola integrated digital enhanced network (IDEN) technology and 12 repeaters that constantly transmit data. The measured field strength at Station 253 from the Nextel site peaked out at about 225dBm. The signal level from WCCCA’s simulcast system averages about 269dBm at Station 253. During the signal strength tests as shown in Figure 6, the meter would not function properly in close proximity to the constantly keyed commercial repeaters.
During the testing, a Motorola MTS2000 portable radio was also tuned to monitor WCCCA’s control channel. When the field strength meter would cease to function, the portable radio would also cease to function. (The red area in Figure 6 indicates those areas.) That is why the portable radios worked better in the engine bay with the bay door closed. The metal bay door actually provided enough RF shielding to let the radio barely receive Fire Dispatch. With the bay door open, the radios lost all reception. When Engine 253 had to respond to an accident, the call was sent to the MDT, but as soon as the bay door was opened, mobile data experienced a high error rate, and the message was scrambled. Also, because of the RF overload from the commercial repeater signal, the mobile radio on Engine 253 was unable to receive a clear signal from Fire Dispatch. To verify the condition, Nextel agreed to a WCCCA request to shut down its site for a brief period. A portable radio was immediately able to receive Fire Dispatch loud and clear in the open air in front of the station. A receiver desense test conducted at the station indicated that the Spectra receiver was being desensed 20dB or more by the commercial site transmitters. We now faced a situation of harmful interference from a commercial source operating within FCC guidelines.
Mapping the effect To map out the exact area of harmful interference being generated from the commercial site, The test procedure was repeated. This time, the field strength meter was tuned to one of Nextel’s frequencies (861.7875MHz). Figure 8 on page 26 shows the antenna pattern from the commercial site. The blue area, which indicates an average signal level greater than 238dBm, corresponds roughly to the red area in Figure 6. This is the same area in which the portable radios would stop receiving.
To verify that WCCCA’s trunking system actually provided coverage in the area of interference from the commercial site, some extra measures had to be taken. With the assistance of a local Motorola engineer, the measuring equipment was modified, as shown in Figure 9 on page 26.
The field strength meter was surrounded with an aluminum shield. RF filtering was added to all power sources and to data cables that were connected to the meter. Dual-cavity filters were also added to the antenna input. The field strength meter and cavity filters were tuned to WCCCA’s control channel #1. Again, the entire area was driven to collect an entirely new set of data points, as shown in Figure 10 on page 26. The RF shielding and filter allowed the field strength meter to operate 90%-95% of the time in close proximity to the Nextel site. This verified that WCCCA’s system was actually providing signal in that area. Figure 11 on page 26 is a composite of maps from the three-part measurement procedure. The yellow area shows the area where public safety’s portable radios stopped receiving and where mobile radios received substantial interference and had difficulty receiving emergency broadcasts from Fire Dispatch.
Motorola review An engineering review by Motorola concluded that it was “evident that Nextel’s IDEN sites in Washington County are serious communications ‘holes’ in key urban areas for public safety communications.”
The report detemined that the initial commercial activation of the first six channels presented heightened interference, but the expansion to 12 channels created the RF environment in which WCCCA’s radios ceased to function. The report also noted that following the short-duration system shutdown that Nextel made at WCCCA’s request, Nextel used an HP spectrum analyzer to check the purity of its signals. The tests showed that the site was within specification and that spurious emissions or noise were below the noise floor of the analyzer (2118dBm). Measurements indicated that the 12-channel site produced 600W ERP composite total, using 28 downtilted omnidirectional antennas, also within specifications.
Bench tests by Motorola to duplicate the problem found that control channel reception was negligibly affected by high levels of interference (113dBm and more) as long as the interfering frequencies were spaced 1.5MHz or more from the control channel frequency. Spacing tighter than 1.5MHz caused the control channel to blank in a symmetrical bell curve. Within just a few hundred kilohertz of the control channel, a varying number of interfering frequencies could blank the receiver with as little as 224dBm.
The engineering report concluded that “although the interference-handling capability of Motorola radio receivers is good, the sheer brute strength of the continuous-duty IDEN site transmitters planted in the same immediate vicinity and on the same sub-band as local public safety communications systems will spell problems for communications reliability in those areas. The issue is further aggravated by installation of IDEN sites in the same heavily populated areas (where) public safety also needs heightened communications reliability.”
Moving toward resolution This was the only site on which WCCCA performed extensive tests. Several other ESMR sites are located in Washington County, and spot-checking has found the same harmful interference. As commercial sites are added to improve coverage, public safety coverage will deteriorate.
Since the testing phase, Nextel has cooperated by cutting its ERP at the site back by half, and reducing the number of active channels from 12 to eight. Some frequencies were also relocated. This has decreased the zone of interference from 1/4 mile to 1/8 mile at the local level.
In January, Motorola contacted WCCCA to arrange to send an engineering team assembled from its IDEN and RF design groups. After analyzing the problem in the lab, this team will follow up with real-world field tests.
A wakeup call to the the industry The commercial mobile radio service in this case, Nextel, cooperated with local public safety in working toward a resolution of the interference issue. However, the same technical conflict is probably taking place all across the country. The national office of APCO has taken an interest in the problem, and Oregon’s congressional delegation has expressed concern.
The FCC has looked at the problem, but concluded that because both parties in this case have valid FCC licenses and are operating within engineering specifications, there is nothing the agency can do. No one mentioned the band plan that contributed to the problem in the first place.
It’s analogous to a shipping channel that is used for critical cargo. The government allows some contractor to build an underwater structure in the middle of the channel, just below the surface, and then does not tell the shippers that it is there.
The FCC, manufacturers, commercial service providers and public safety agencies need to examine the consequences to public safety, particularly in populous areas, resulting from certain band plans, power levels and site locations. Improved frequency coordination and RF design are essential to dealing with interference.