IBOC Interference Calculations

The procedure outlined here is intended to estimate the impact on the nighttime groundwave coverage of a Canadian AM station from a U.S. AM IBOC station located on a 1st adjacent channel. Although it is not the only situation that may cause interference, 1st adjacent IBOC is the dominant source of nighttime problems, due to the strength of nighttime skywave and the fact that most of the digital power resides in the 1st adjacent channels. Nearly all of the power that gets dumped into the adjacent channels is in the primary digital sidebands, which are located between 10 and 15 kHz from the carrier frequency of the IBOC station (for more details about the AM IBOC signal and its power spectrum, see my AM IBOC Parameters page).

The power in one primary digital sideband is -16 dBc, i.e., 16 dB less than the AM carrier power. Therefore, if we know the received field strength of the analog AM signal, we can easily determine the field strength of a primary digital sideband by applying the 16 dB factor.  In terms of field strength, the ratio between the two is 6.3 (i.e., 20log(6.3) = 16).

Now we must also consider the nature of the resulting co-channel interfering signal. Since it is composed of many closely spaced carriers (OFDM) carrying complex (64QAM) digital modulation, it is very much like wideband noise; indeed, it produces a "rushing" or "hissing" noise when received by a conventional AM receiver. This is quite different from interference from another AM station. Fortunately, the IBOC digital signal is very similar to the signal used in the DRM system, which has been studied in detail in Europe. The ITU has determined that if an analog AM signal is replaced by a DRM digital signal, the average power must be reduced by 6-7 dB in order to provide equivalent protection to co-channel stations. Putting it another way, a digital emission with a given average power will cause 6-7 dB more interference power at the audio output of an AM receiver than an AM signal having the same (carrier) power, all other factors being equal.

What this means is that if the IBOC digital sidebands were treated as separate signals (which they should be, but the FCC refuses to recognize this "inconvenient truth"), the protection rules would have to be adapted accordingly (as they were for DRM in Europe and other parts of the world). In the Americas, the current standard for protection from AM co-channel interference at the protected contours specifies a D/U ratio (desired to undesired field strength ratio) of 26 dB. Presumably, this is the D/U ratio that corresponds to the point where the interference just becomes noticeable, but not annoying enough to bother most listeners. For interference from a digital signal like DRM or an IBOC primary sideband, however, this D/U ratio would have to be increased to around 32 or 33 dB.

Okay, so let's assume that the digital interference would become noticeable at a reception point where the co-channel D/U ratio is about 32 dB. It's probably safe to say, then, that if the D/U ratio were reduced by 10 dB to 22 dB, the interference would be extremely annoying, perhaps inducing most listeners to tune out. And, if the D/U ratio were reduced by a further 10 dB to a mere 12 dB, reception would likely be completely destroyed.

Not that we've laid the groundwork, here's the procedure for evaluating the interference:

  1. Using the standard procedures for estimating skywave field strength, as outlined in the Canada-U.S. bilateral agreements on AM broadcasting, calculate the nighttime field strength of the interfering 1st adjacent AM IBOC station at the location of the desired AM station (as is usual in such calculations, this field strength will be assumed to be the same throughout the coverage area of the desired station). The field strength calculated should be for 10% time, as specified in the agreements for interference calculations. The 10% field strength is normally abbreviated as F(10).

  2. Determine the field strength of the primary IBOC sideband that is co-channel to the desired station by reducing the field strength calculated in Step 1 by 16 dB (i.e., divide it by 6.3).

  3. Determine the coverage contour for the desired station at which the interference would be just noticeable (D/U ratio of 32 dB) by increasing the field strength calculated in Step 2. Note that we can jump directly from Step 1 to this step by simply increasing the Step 1 field strength by 16 dB. From there, we can guesstimate the contour at which the interference would become very annoying by reducing this field strength by 10 dB (i.e., dividing by 3.16, to get a D/U ratio of 22 dB), and the contour at which reception is essentially destroyed by reducing it by 20 dB (i.e., dividing by 10, to get a D/U ratio of 12 dB).
It is also important, of course, to put this new interference in context by comparing it with the existing analog interference levels. Nighttime coverage of AM stations is normally expressed in terms of the usable field strength contour, abbreviated Eu (in the U.S., it is often referred to as the nighttime interference-free, or NIF contour). In a nutshell, the basis of this contour is an estimation of the total received field strength of co-channel stations at night, arrived at by summing them by means of the root sum square (RSS) method.  In the RSS method, the co-channel signal strengths are ranked in descending order, and the calculation proceeds until the next one on the list is less than 50% of the previous calculation (the "50% exclusion principle"). Then the Eu is determined from the resulting RSS sum by increasing it by 26 dB , the co-channel protection ratio (i.e., multiplying by 20). For example, suppose that the calculated co-channel field strengths for a given station are (in descending order) 0.450 mV/m, 0.300 mV/m, 0.250 mV/m, and so on. Since the second value is more than 50% of the first, we must include it in the calculation.  The RSS sum of the first two values is 0.541 mV/m. The third-highest field strength is less than 50% of this, so we don't need to include it, and the calculation is complete. The calculated Eu is therefore 20 x 0.541 = 10.8 mV/m, so the station in question is deemed to provide adequate nighttime service (barring local noise level issues and such) in areas where it provides a field strength of 10.8 mV/m or more.

Now, if a new station is proposed to be established on that particular channel, it must meet a basic requirement: it has to protect the established station by generating a field strength at that station's location that does not increase its RSS value.  What this means in practical terms is that the field strength must meet two criteria: it must be less than 50% of the existing RSS value, and it must also be less than the smallest current contributor to the existing RSS calculation. In other words, the new station must not become a factor in the station's RSS calculation. So, in the example discussed above, a new station cannot create a field strength of more than 0.541 / 2 = 0.2705 mV/m at that location.

The fundamental problem with AM IBOC is that digital sidebands are not required to conform to this allocation principle. This is a massive loophole that will cause many nighttime woes on the AM band.

Some real world examples should clarify things further.


CFRB-1010

CFRB is a 50 kW Class A station in Toronto, with a DA pattern that sends most of the power to the northeast, towards downtown Toronto. Although Class A stations are in principle protected against co-channel interference to their 0.5 mV/m contours (with a D/U ratio of 26 dB), CFRB accepts higher levels of interference than this from co-channel station WINS in New York City. The calculated F(10) from WINS in the Toronto area is 0.137 mV/m. This means that the nominal 26 dB D/U ratio is reached at CFRB's 2.8 mV/m contour rather than out at the 0.5 mV/m contour.  CFRB also receives enough signal from co-channel CBR in Calgary to figure in their RSS value, so their calculated Eu is actually about 3.2 mV/m. Despite this apparent tolerance for interference, however, things could get a lot worse for CFRB when nighttime IBOC comes.

At the present time, the only significant source of such interference is WMVP-1000 in Chicago, but it's a biggie. WMVP is also a 50 kW Class A, only 710 km from Toronto, and its nighttime DA has a major lobe pointed directly at CFRB. By my calculation, WMVP creates an F(10) of 3.74 mV/m in the Toronto area. The corresponding field strength of its primary digital sideband (located between 1010 and 1015) is thus 3.74/6.3 = 0.594 mV/m. Note that this is a much stronger co-channel interference signal than WINS, even without factoring in the greater interference-causing potential of digital versus analog. According to the procedure outlined above, we can estimate that this interference will become noticeable at the 3.74 x 6.3 = 23.6 mV/m contour of CFRB. The interference then would likely become very annoying at around the 23.6 / 3.16 = 7.5 mV/m contour, and at the 23.6 / 10 = 2.4 mV/m contour, reception is essentially kaput. So, at roughly the same contour where interference from WINS and CBR would be currently just noticeable, IBOC interference from WMVP will probably completely destroy reception. WINS will no longer be a factor in limiting CFRB's coverage. If the WMVP digital sideband were treated as a co-channel station on 1010, then it would be the only station required in the RSS calculation, even without factoring in the additional protection needed when the signal is digital. The other co-channel signals don't even come close!

If KDKA-1020 should decide to run IBOC, CFRB's plight will become even worse. KDKA is a 50 kW non-directional Class A that is only 330 km from Toronto, and it would drop its lower primary sideband on top of CFRB, between 1005 and 1010, thus filling in the wall of noise on 1010.


CJAD-800

CJAD in Montreal is a Class B with a 10 kW nighttime DA that has a major lobe directed northwest towards the heart of Montreal. The major source of IBOC interference for this station will be WGY-810 in Schenectady NY, at a distance of only 275 km. WGY is a 50 kW non-directional class A that produces an F(10) in the Montreal area of about 2.33 mV/m. IBOC interference is thus estimated to become noticeable around CJAD's 14.7 mV/m contour, and very annoying at around 4.6 mV/m. Since CJAD's calculated usable field strength contour is about 8.0 mV/m (the RSS value is determined by just one station, CKLW), the IBOC interference will not be as overwhelming as in the previous example. Nevertheless, it will become the dominant source of interference, and is likely to have a significant impact on CJAD's nighttime coverage (the field strengths of CKLW and the WGY digital sideband will be approximately the same, but the latter will be more important due to the greater impact of digital interference).


CFFR-660

CFFR in Calgary, AB is a 50 kW Class B with a directional night pattern towards the NNE. Unfortunately, KBOI-670, a class B in Boise ID 833 km away, also has a 50 kW nighttime DA pattern towards the north, aimed squarely at CFFR. And, of course, they run IBOC. The calculated F(10) for KBOI in the Calgary area is a whopping 4.16 mV/m, so the primary digital sideband on 660 would weigh in at 0.660 mV/m. This indicates that the IBOC interference would be audible at CFFR's 26 mV/m contour, and very annoying at their 8.3 mV/m contour. Reception outside the 2.6 mV/m contour would be an exercise in futility.

Currently, the dominant source of co-channel interference to CFFR would be from KTNN in Window Rock, AZ, about 1700 km from Calgary. KTNN is a 50 kW Class B with a west-facing cardioid night pattern that throws considerable power to the north, towards CFFR. The calculated F(10) for KTNN in the Calgary area is 0.422 mV/m, indicating that just-noticeable interference would occur at CFFR's 8.4 mV/m contour.  However, there is a second station, KAPS, that produces enough field strength (0.305 mV/m) to be included in the RSS calculation.  The RSS value comes to 0.520 mV/m, and the Eu works out to be about 10.4 mV/m. So, here again, we have a situation where new IBOC interference becomes the dominant source of co-channel interference, producing a higher field strength than the existing co-channel stations, even without considering the higher protection ratio needed for digital interference. This will cause very annoying interference at the same contour where interference was just noticeable before the advent of IBOC.  CFFR's coverage will suffer accordingly.


CINF-690

CINF is a 50 kW Class A station in Montreal with a very broad, moderately directional pattern that is directed towards the metro area.  As a Class A, it enjoys a high degree of protection from co-channel stations. The calculated Eu at this time is 2.2 mV/m (the major source of interference is a station in Cuba; otherwise, the Eu would likely be close to 0.5 mV/m). The major current IBOC threat to CINF is from WLW-700 in Cincinnati, OH. WLW is a 50 kW Class A non-directional station located 1101 km from the CINF site. The calculated F(10) from WLW in the Montreal area is 1.40 mV/m, which translates into a received co-channel field strength on 690 of 0.222 mV/m. This will become the only contributor to the RSS calculation for CINF. Based upon a protection level of 32 dB for digital signals, this will produce an Eu value of 8.9 mV/m, a dramatic increase from the current value.

As bad as this may seem, it could get much worse, if WRKO-680 decides to run IBOC. WRKO in Boston is only 379 km from CINF, and their 50 kW nighttime signal has a major lobe to the north, putting a whopping 2.44 mV/m received signal into the Montreal area (note that if adjacent channel stations are included in the CINF RSS calculation, this would likely be the largest contributor, indicating an Eu of the order of 2.6 mV/m). If they ran IBOC, the co-channel interfering signal on 690 in Montreal would thus be 0.388 mV/m, leading to a calculated Eu of 17.9 mV/m (including the contribution from WLW)!


CKAC-730

CKAC is another 50 kW Class A station in Montreal, with a similar pattern and coverage. Their current Eu is 2.1 (only Latin American stations contribute to the RSS calculation). The IBOC threat here is WGN-720 in Chicago, at a distance of 1195 km. The F(10) from WGN at CKAC is 1.27 mV/m, which translates into a received co-channel field strength on 730 of 0.202 mV/m. This will push the Eu up to 8.1 mV/m, a substantial increase. Here again, the IBOC signal will become the sole contributor in the RSS calculation.


CIGM-790

CIGM in Sudbury, ON, is a 50 kW Class A with a north-directed directional pattern. Their current Eu is 4.4. They will receive substantial IBOC interference from WBBM-780 in Chicago, only 750 km away. The F(10) from WBBM at CIGM is 2.29 mV/m, which translates into a received co-channel field strength on 790 of 0.363 mV/m. This will push the Eu up to 14.5 mV/m, a huge increase. Here again, the IBOC signal will become the sole contributor in the RSS calculation.

CHFA-680

CHFA in Edmonton, AB, is a 10 kW Class B station with a directional pattern to the north, and a current Eu value of 6.5 (due to co-channel KNBR). As was the case with CFFR-660, the problem here is KBOI-670. The F(10) from KBOI (1127 km away) at CHFA is 2.60 mV/m, which translates into a received co-channel field strength on 680 of 0.413 mV/m. This will become the only contributor in the RSS calculation and push the Eu up to 16.5 mV/m, a substantial increase.

CKSB-1050

CKSB is a 10 KW Class B station in St. Boniface, MB, with a cloverleaf-shaped night pattern that protects several co-channel stations. The current RSS value for CKSB is 0.273 mV/m, and the Eu is 5.5 mV/m. However, this will change considerably after the IBOC station WHO-1040 begins nighttime digital operations. WHO is a 50 kW non-directional Class A located 948 km from CKSB. The F(10) from WHO at CKSB is 2.26 mV/m, which translates into a received co-channel field strength on 1050 of 0.359 mV/m. When the appropriate adjustment for digital interference is made, this will become the only contributor in the RSS calculation and push the Eu for CKSB up to 14.4 mV/m.


The Bottom Line

Yes, I kid you not - in all of the cases cited above, plus many others not described here (including many domestic situations in the U.S., of course), a station on an adjacent channel will now become the dominant source of co-channel interference. In all but one of the cases, the IBOC interference would become the sole determinant of the station's RSS value. In all cases, the usable field strength contour (NIF) will decrease dramatically. And this only takes into account stations currently using IBOC - future expansion of IBOC operations can only make matters worse. The case of CFRB is the most shocking. Class A clear channel stations are the elite of the AM broadcasting world, with the best protection from nighttime interference; yet, in one fell swoop, this elite station will acquire a NIF contour that is more typical of a Class C graveyarder! The following table sums up this sorry mess.
Station CFRB-1010 CJAD-800 CFFR-660 CINF-690 CKAC-730 CIGM-790 CHFA-680 CKSB-1050
Current Overall RSS Value, mV/m 0.159 0.402 0.520 0.111 0.105 0.222 0.323 0.273
Current Eu (NIF) Contour, mV/m 3.2 8.0 10.4 2.2 2.1 4.4 6.5 5.5
New RSS Contribution Due to IBOC, mV/m 0.594 0.370 0.660 0.222 0.202 0.363 0.413 0.359
Adjusted IBOC RSS Contribution*, mV/m 1.188 0.740 1.320 0.444 0.404 0.726 0.826 0.718
New Overall RSS Value, mV/m 1.188 0.842 1.320 0.444 0.404 0.726 0.826 0.718
New Eu (NIF) Contour, mV/m 23.6 16.8 26.4 8.9 8.1 14.5 16.5 14.4
* Field strength of the primary digital sideband increased by 6 dB to allow for increased protection level required from digital co-channel interference

Allowing this to happen flies in the face of established allocation rules and international agreements. And yet, this bizarre interference generating system has been blessed by the FCC! What were they thinking? It almost defies belief...

Last Update: 24 August 2007
© 2007, by Barry McLarnon