Resolving digital-analog, ESMR-SMR interference

Mutual interference between digital and analog transmitters and receivers has posed problems for ESMR and SMR system operators. For some, it’s a little secret_dirty or not_that they’d rather not discuss.

Sometimes radio system operators get in touch with MRT to talk about a problem in hopes that it might be publicized. The idea is that maybe people who are causing them trouble might change their ways when the glaring light of publicity shines on them. Sometimes we publish a story about the problem, although it is not always possible to conclude whether it helped or not.

Last year, Sal Dragotta of Viking Communications in Milwaukee called to talk about interference that his analog specialized mobile radio (SMR) system was receiving. The results of his own investigation indicated that the interference came from adjacent-channel digital SMR transmitters. Since then, others have spoken about interference received by digital SMR systems from adjacent-channel analog transmitters.

Maybe it should not be surprising, but various involved companies that might have something to say about mutual interference between analog and digital systems are not thrilled about designating someone to speak on the record on the subject. This includes representatives of SMR and enhanced SMR (ESMR) companies, paging companies, two-way radio and paging equipment manufacturers, and broadcast stations.

Consider one variation of the problem. The SMR frequency band originally was configured for analog service. Transmitter emission masks, receiver filter responses, site separation distances and other system characteristics were set to keep mutual interference among analog signals to a minimum. Along came digital transmitters with emissions that “fill out” the channel to its edges with higher average power levels. Analog receiver filters, sharp enough to block the weaker energy at channel edges from analog transmitters, pass enough of the digital signal to interfere with reception.

Another variation: Digital base stations often are distributed on low towers or buildings throughout a coverage area. In contrast, analog base stations usually are placed at high elevation, sometimes removed from the population center. When analog mobiles that use channels adjacent to the digital system travel near the digital base stations, their transmitter signals can desensitize the digital receivers. The reverse does not happen as often because digital mobiles may not travel near the analog base stations, especially those at mountaintop sites.

One more: Consider the operation of digital paging and broadcast transmitters near SMR base station receiving sites. Even though the transmitters might meet emission mask requirements, enough energy at adjacent frequencies might be emitted to cause on-channel interference with nearby receivers.

These were the subjects of a for-the-record discussion, an “Analog-Digital Summit,” that the magazine sponsored in September 1997, in Santa Clara, CA. Manufacturers of two-way radio, paging and spread-spectrum equipment declined invitations. So did an ESMR system operator. An SMR system operator, and a TV broadcast station licensee, with plans to install a digital TV transmitter, accepted invitations, but their representatives did not make it to the meeting because of last-minute conflicting obligations with a customer and a supervisor.

Attending the meeting was Madjid A. Belkerdid, Ph.D., a professor at the University of Central Florida, Orlando, FL. Belkerdid’s area of expertise is spread-spectrum communications and RF systems. He is a member of the editorial advisory board of Intertec Publishing’s RF Design magazine, which conducted a conference in Santa Clara at about the same time. Also attending was John Kuivinen, P.E., a regulatory and compliance engineer with Radio Communications Systems, Vista, CA. He has placed extensive RF and wireless applications before the FCC and Underwriters Laboratories, including Part 15 certification and home and medical alert systems. A third participant was Darrell L. Ash, senior vice president and chief technology officer for RF Monolithics, Dallas, a filter manufacturer.

The first part of the discussion involved the areas of greatest expertise for Belkerdid and Kuivinen, cellular and PCS. Although the magazine covers some cellular and PCS infrastructure topics, especially base stations, the following information has more to do with SMR interference because that was the original idea for the meeting.

MRT: “Some apparent difficulties have been reported between analog and digital SMR systems . . . interference within a given band of spectrum where some operators are using transmitters with digital and some are analog. What’s been reported is that the analog operators are receiving interference from the digital transmitters.”

Belkerdid: “It’s not harmonics of the carrier . . . it’s the digital pulses of the data itself that increase the bandwidth, and it interferes with the analog channel.

“It also depends on the situation because of the near-far problem. . . . If the analog [mobile] happens to be far away from his tower and closer to the digital [transmitter], he could be three or four frequency channels far away and you can still get to him.”

Kuivinen: “Most of the regulations that we have on the books now, as Part 90 and Part 21 common carrier, relate to fixed-frequency, single-channel radios that were basically tube model radios that were in common use in the mid-’70s to the late ’70s. Those radios generated very little in the way of nonharmonically related spurious energy. They are relatively spectrally pure, particularly since each one was tweaked to its particular frequency by a qualified service technician, and they didn’t have broad coverage of any kind.

“. . . Solid-state radios came on, a little less spectrally pure . . . but people could work around them because they had some known amount of energy. But with this new spread-spectrum technology and pulse radios, you can end up literally with dc to daylight being polluted with RF noise.”

Suppressing emissions outside of an authorized channel bandwidth is key to controlling interference. Suppression, measured in decibels (dB), is what Kuivinen spoke of next.

Kuivinen: “I don’t see any way out of the problem unless the FCC says basically that you have to clean up these radios, and 80dB just isn’t good enough for nonharmonically related energy, to go to 100dB or something of that nature as a minimum for sideband energy greater than, say, five bandwidths or three bandwidths located away from the carrier frequency that you have to be down 100dB. I don’t see any way around that other than something like surface acoustic wave (SAW) filter technology or a very exotic and sophisticated filter technique.”(For more information on SAW filters, refer to the article on page 32 of this issue.)

Ash: “SAW filters are not going to help the radios [receivers] . . . Once the interference is in-band . . . the sideband beingin the band of the signal that you are trying to receive . . . so the solution would be to add filtering to the transmitter.”

Kuivinen: “That’s what I’m saying.”

Ash: “100dB; I don’t know if that’s the right number.”

Kuivinen: “I don’t either. I don’t even know if it’s technically achievable. I’m saying if you have numerous radios co-located within a kilometer of each other and you have a radio with the potential ability to interfere with every one of those radios within that kilometer . . . you have a lot of interfering signals.”

MRT: “You said, ‘If it’s possible?'”

Ash: It’s possible if you have an intermediate frequency. In other words, if you can divide the filtering into two bands, you can do it. But 100dB is very difficult at a single frequency.

” . . . If the transmitters are modulating at a lower frequency and then mixing up to a higher frequency to transmit . . . then you can use a SAW filter at the intermediate frequency and another SAW filter at the RF frequency before you’ve reached the high power levels.”

Based on the discussion to this point, the participants seem to have agreed that there may be possibilities for adding filters to transmitters that would reduce adjacent-channel interference from digital systems to analog systems. The following excerpts from the discussion cover the possibility of using predistortion to reduce potential interference.

Kuivinen: “In the case of some of our types of modulation schemes at Linear Corporation, we actually predistort the waveforms before we modulate the devices, in order to comply with FCC rules because we have limited bandwidths we have to work under, too.

“In the cases of these digital transmitting devices, I suspect that they’re just taking 9600-baud or whatever baud-rate data that they can jam out as fast as possible and then going straight in and modulating a crystal or a channel element of some kind, or a frequency synthesizer, and they do not worry about the bandwidth or the sideband rejection problem. They just come in, and they start blasting out, and they_I jokingly call it ‘tuned for maximum smoke’_get whatever power they can get out and as long as they follow some curve, they say, `Okay, we’re good enough.’

“As an RF design engineer, if I have a specification and it says I have to be 43dB down, I look at that . . . and if I put that one dollar more into the product and I go to maybe 45dB or 46dB down . . . but I’m not going to go to 80dB down if the FCC rules only say that I have to go to 43dB.

“Even on a system basis, I think it’s a requirement. If I know that the competition is not going to [exceed sideband suppression requirements], I can’t sell the product if it’s going to cost twice as much or three times as much as the competition’s products that just barely meet the minimum requirements.”

Belkerdid: “My guess is that [the manufacturers] are probably meeting [the specifications] to the dot, and in a digital receiver and in an analog receiver, you can pick up the near-far [interference].

MRT: “Meeting the RF emission mask is fine as far as the manufacturing of the product is concerned, but not causing harmful interference is a burden on the operator of that product.”

Kuivinen: “Historically, [the FCC] has always said ‘The last man on the mountain fixes all of the problems.’ [Meaning, the burden to make system changes to alleviate interference falls upon the owner of the latest system to be activated when the interference affects previously installed systems.] The problem is in the conversion process. It’s very difficult to prove who is doing what in the case of, particularly, sideband energy. Even the instrumentation that’s commonly available in the way of quality Hewlett-Packard spectrum analyzers and very high-tech measuring equipment, it’s very hard to find these kinds of low-level signal that are actually causing this interference and proving that it’s coming from that particular site. It’s not impossible, but it may not be practical.”

Wrap-up At this point, we’ll leave the discussion, and thank Madjid Belkerdid, John Kuivinen,and Darrell L. Ash for taking the time and for having the confidence to discuss these interference issues for the record. The transcript of the meeting runs to 107 pages, by the way. It included discussions of PCS and Part 15 uses of spread-spectrum modulation, sideband energy interference from TV and FM broadcast stations and paging transmitters, “cabinet radiation” and guard bands.

Now, remember Sal Dragotta, who placed the phone call that began the whole process? I telephoned him on Jan. 20, 1998, to ask whether and how he may have resolved his interference problem. He said he had traced the source of the interfering sideband energy to transmitters operated by Nextel Communications, McLean, VA.

“We were playing cat and mouse with Nextel for a while,” Dragotta said. “I recorded some high-definition spectrum analyzer pictures of the interference. I’ve had their engineers here. They’re not sure whether they’re in compliance.

“The problem for us was that they were using digital transmitters on channels adjacent to our analog channels. The first instance took us a half-day to find. Then we knew what to look for. We gave them a call, and the problem disappeared after a half-hour.

“We sent a letter to the FCC. We gave Nextel a list of our frequencies, and they’ve stayed away. Every few months, we hear a signal near ours.

“You can’t put digital signals next to analog channels. We looked at them with high-end equipment. The digital signals are 45W high and 25kHz wide. Analog receivers were never intended to have such signals near them.”

So the solution that appears to have been implemented is that Nextel is treating assigned frequencies adjacent to analog channels as guard bands, at least in Dragotta’s neighborhood. “Who knows what will happen as Nextel turns on more channels,” Dragotta said.

At this point, we’ll leave the subject for another day. We haven’t heard much about whether and how analog systems interfere with digital systems. Inasmuch as analog system growth mostly has been halted, analog systems rarely are the “last man on the mountain,” so maybe that’s why. Maybe, with the publication of this article, we’ll start hearing from more people about the nature of the problem and possible solutions.