THE UNIVERSITY OF MICHIGAN OFFICE OF RESEARCH ADMINISTRATION ANN ARBOR ON THE PERFORMANCE OF FM COMMUNICATIONS IN THE PRESENCE OF RF NOISE Technical Memorandum No. 91 2899-60-T Cooley Electfr'oics Laboratory Departrment of Eleotrical Engineering By: G. A. Hellw'arth Approved by: H. W. Farris A CEL publication is given a memorandum designation due to reservations in one or more of the following respects: 1. The study reported was not exhaustive. 2. The results presented concern one phase of a continuing study. 3. The study reported was judged to have insufficient scope. Project 2899 TASK ORDER NO. EDG-7 CONTRACT NO. DA-36-039 sc-78283 SIGNAL CORPS, DEPARTMENT OF THE ARMY DEPARTMENT OF ARMY PROJECT NO. 3A99-06-001-01 November 1961

TABLE OF CONTENTS Page LIST OF ILLUSTRATIONS iii ABSTRACT iv 1. INTRODUCTION 1 2. EXPERIMENTAL PROCEDURE 2 3. BASIC EXPERIMENTAL RESULTS 3 4. INPUT VS. OUTPUT S/N 5 4. 1 Broad-Band Noise Performance 5 4. 2 Performance Against Intentional Narrow-Band Noise 6 5. CONCLUSIONS 8 REFERENCES 9 DISTRIBUTION LIST 10 LIST OF ILLUSTRATIONS Figure Page 1 Block diagram of experimental test system. 2 2 FM receiver output signal and noise vs. input FM carrier-to-noise power ratio, the noise power taken in a fixed 6.8 kc bandwidth. 4 3 Receiver output vs. input signal-to-noise ratios. The input noise power is taken in a fixed 6.8 kc bandwidth. 6 4 Receiver output vs. input signal-to-noise ratios. The input noise power is the total power entering the receiver; the input signal power is the peak envelope power. 7 iii

ABSTRACT A critical view is taken of some common notions of input signal-to-noise ratio "thresholds" below which frequency-modulation radio links become inoperative but above which highly satisfactory communications result. Experimentally derived curves are shown giving output vs. input signalto-noise ratios for three FM receiver bandwidths and a signal employing sine wave modulation. When the experimental FM curves are compared with theoretical curves for AM, DSB, and SSB, it becomes clear that the "threshold" concept is extremely misleading. FM systems can enjoy a competitive output signal-to-noise ratio for low input S/N against uniform RF noise, while against narrow-band RF noise, wide-band FM is superior to the other systems for practical output S/N. iv

1. INTRODUCTION This technical note is intended to help clear up some popular confusion with regard to the noise performance of frequency-modulation communication receivers. The notion of an input signal-to-noise ratio "threshold" has been propagated in the belief that below this threshold FM reception is hopelessly lost in the noise whereas above the threshold all reception is "hi-fi" and clear (Ref. 5, p. 279, and Ref. 8). Of course, no such threshold exists in fact; the FM output S/N is smoothly related to the input S/N in spite of the nonlinearity of the receiver, and in the case of a low input S/N the output S/N may be not only quite useful but larger than the input S/N. Perhaps the phenomenon known as the "capture effect" has caused some of the confusion. In the first place, so-called capture results not from the receiver limiters or frequency detector, but rather from the definition of frequency measurement itself which the receiver merely tries to carry out. Secondly, the capture effect refers to the specific case of two constant-amplitude input signals, of which the desired FM signal is one. In this case the average output frequency equals the average frequency of the larger amplitude signal (Ref. 6). If the deviation ratio1 is large, then the FM receiver output indicates the (slowly changing) frequency modulation of the larger amplitude signal while showing evidence of the smaller amplitude signal only during those instants in time when the difference between the instantaneous frequencies of the two signals is less than the receiver audio bandwidth. Finally, if either near-unity deviation ratios or amplitude-modulated signals are being considered, it is best to abandon notions of the capture effect and take a fresh look at the specific situation. Bandlimited Gaussian noise is not a constant-amplitude signal. Because of the instantaneous amplitude fluctuations of the noise, capture of the FM carrier could be considered to take place only during the occurrence of noise peaks exceeding the FM carrier amplitude. In such a situation the RF noise would always produce some peaks large enough to produce interference and yet, because the noise envelope also goes to zero occasionally, the noise does not cause capture all the time. The percentage of time that the noise peaks exceed the FM carrier amplitude is a smooth function of the signal-to-noise ratio, the funcThe deviation ratio shall be considered to be one-half the ratio of the FM receiver IF to audio bandwidths.

tion possessing nonlinearities but not abrupt thresholds. Thus it is a nonlinear relationship between the output and input signal-to-noise ratios which is of interest and which will be experimentally determined. The experimental results will be noted and compared with the theoretical predictions and results of others. 2. EXPERIMENTAL PROCEDURE The experimental test equipment consisted simply of an FM signal source, a radio-frequency noise source, a good quality FM receiver, a wideband voltmeter, and a narrowband selective voltmeter. The input signal and noise levels were set separately by obtaining equal rms voltages at the output of the 455-kc fixed gain IF amplifier of the receiver; then calibrated attenuators were used to obtain the desired input ratio. The output noise voltage was measured in the absence of signal modulation (for convenience ), while the output signal was measured by a selective voltmeter (HP 300A wave analyzer). Figure 1 shows a BALLANTINE 320 TRUE RMS BOONTON NOISE 20 2 C VTVM OUT GENERATOR - IN __ OUT MULTIBANDWIDTH ADDER ANT FM AUDIO IN RECEIVER OUT 60 MC IF N AMPLIFIER ATTN -IN NOISE SOURCE SIGNAL WITH AGC. HP 300A OUT WAVE ANALYZER Fig. 1. Block diagram of experimental test system. 2The change in output noise when the signal modulation is removed is small and not important to these measurements. - 2 -

block diagram of the test system. These tests all involve (1) 1-kc sine-wave signal modulation, (2) near maximum modulation depth, (3) audio bandwidth of 3. 4 kc (100 cps to 3. 5 kc), and (4) white Gaussian radio-frequency noise interference, linearly added to the receiver input. For frequencymodulation signals, maximum modulation depth is assumed to be a peak deviation equal to one half the receiver bandwidth. However, when the input noise level is high, the maximum audio output signal is normally obtained with the input deviation reduced to about 70 or 80 percent of maximum. This results from the slight attenuation occurring at the band edges of the IF amplifier frequency response. If less than maximum signal deviation is normally used, the output signal level is decreased, and thus the output S/N is decreased. The use of any sine wave frequency within the 3. 4 kc audio bandwidth would yield the same results; 1 kc was arbitrarily chosen. The use of other audio bandwidths affects the results since this changes the deviation ratio. The output noise spectrum is approximately uniform for low input S/N and approximately proportional to frequency for large input S/N (Refs. 3, 4, and 7). 3. BASIC EXPERIMENTAL RESULTS Figure 2 shows both the output signal and the output noise as functions of the input S/N for the three IF bandwidth positions of the FM receiver. The interfering noise signal is assumed to have a wide bandwidth and uniform distribution, while noise power included in the S/N ratio is only that noise in a fixed 6.8-kc bandwidth. This method of S/N measurement is conventional and is used in all the discussions in the references. The method meaningfully relates to the communication range problem in which the interfering signal is uniform wideband atmospheric or receiver noise. At high input S/N, the receiver output signal is a maximum corresponding to the actual deviation of the input carrier. The noise is reduced by only a certain amount, as indicated by the theoretically derived asymptote shown in Fig. 2 in accord with the observations of Crosby and others (Refs. 1, 2, I, and 9). The asymptote has unity slope so that at high input S/N the output S/N is proportional to the input S/N and the deviation ratio. At low input S/N, the receiver output noise reaches its maximum value, a function of the receiver IF and audio bandwidths, while the signal modulation is suppressed. Theoretical calculations show that the audio output signal should be reduced linearly with the amplitude of the RF input -3

_0 -BR' 200 KC o0-0-O X-10 / 50 KC u, | BR 200 KC / ^ OUTPUT -- ~ ~ —20-'0- / / SIGNAL 50 KC / / (I KC SINE u-0-0 O S \~~ 10 KC / MODULATION) -40 - OUTPUT a' |// VA NOISE - 50 0\ o 0 7 -70 0 N i3. -80 - \ - 80 -20 -0 0 10 20 30 40 50 FM RECEIVER INPUT S/n (DB) n IN 6.8 KC &W Fig. 2. FM receiver output signal and noise vs. input FM carrier-to-noise power ratio, the noise power taken in a fixed 6.8 kc bandwidth. signal for a given (large) amount of input noise (Ref. 7). However, the experimental evidence of Fig. 2 points more toward a square-law effect; that is, the level of the signal component of the output will change as the square of the input carrier voltage. The experimental results of Fig. 2 also show an interesting phenomenon in that the maximum obtainable output signal is nearly independent of the receiver bandwidth -4

with high-level, uniform input noise. That is, at low input S/N ratios there is an input signal deviation (normally about 70 percent of the receiver half-bandwidth) yielding a maximum output signal whose magnitude is about the same for the three bandwidths shown. This result has not been predicted by theory. 4. INPUT VS. OUTPUT S/N 4.1 Broad-Band Noise Performance If the decibel differences between the respective signal and noise curves of Fig. 2 are plotted against input S/N, one obtains the curves of Fig. 3, relating the input and output signal-to-noise ratios. In addition to the experimentally obtained FM curves, there are curves representing AM, SSB, and DSB systems where average power values are considered (for the AM input signal power just the carrier power is measured). The SSB and DSB curves are theoretical but easily obtained by ordinary receivers. The AM curve, experimentally obtained on a good quality AM receiver, matches the theoretical curve quite closely. It must be noted that the AM receiver's departure from linearity at less than 0 db S/N is the fault of the envelope detector and is not a necessary evil when using AM transmissions. The AM signal may be demodulated in the same manner as a DSB signal without the low S/N degradation. It should be remembered that the input noise power parameter of Fig. 3 (and Fig. 2) is really noise power per 6.8 kc of bandwidth regardless of the operating bandwidth of the receiver. With the curves plotted in this manner, the classical observations may be made: FM is better only at high input S/N, and the wider the FM bandwidth the better. FM is much poorer at low input S/N, and the wider the FM bandwidth the poorer the output S/N. This "improvement threshold" that is often mentioned for FM is located at the S/N where the FM output S/N reaches its highest value above AM. These "thresholds" are observed from Fig. 3 to be at input S/N of about 10 db for 10-kc FM, 18 db for 50-kc FM, and 27 db for 200-kc FM. Clearly, the FM output is usable at input S/N below these values if AM, DSB, or SSB are similarly usable at relatively low output S/N. Thus, in order to compare the efficiencies of the various systems, one must suggest a minimum output S/N ratio at which the required input S/N ratios are to be compared. -5

60 - 50- FM THEORETICAL ASYMPTOTES o k 40 - FM \0 30 z 20 a. 0 W 10 -20 - J I I i I I I -10 -5 0 5 10 15 20 25 30 RECEIVER INPUT S/n (DB); n IN 6.8 KC BW Fig. 3. Receiver output vs. input signal-to-noise ratios. The input noise power is taken in a fixed 6.8 kc bandwidth. 4.2 Performance Against Intentional Narrow-Band Noise The use of input noise-power-per-unit-bandwidth is valid for studying receiver operation against either broad-band interference or atmospheric and receiver noise. However, if one is to study receiver effectiveness in the presence of intentional narrow-band interference, he must reshuffle the curves of Fig. 3 so that the input noise parameter represents the total noise power entering the receiver within its bandwidth. In addition, let us change the input signal power from average to peak envelope power, another reasonable -6

60 - 30n FM, ~ 20,,,\,,,.~ 0 0 - 1 10 0 0 -0 -20 -20 -15 -10 -5 0 5 10 15 20 RECEIVER INPUT S /N (DB) S, = PEAK ENVELOPE POWER, N= TOTAL NOISE POWER Fig. 4. Receiver output vs. input signal-to-noise ratios. The input noise power is the total power entering the receiver; the input signal power is the peak envelope power. measurement in a practical situation with peak power limited transmitters. The result is Fig. 4, where a completely different set of conclusions must be drawn as to relative resistance to narrow-band, radio frequency noise interference. It is now seen that FM is always superior to AM, and wideband 200-kc FM is better than even SSB for output S/N above about -15 db (a practical lower limit for voice communications if one must assign limits). At the point where the 200-kc FM curve and the -7

SSB-DSB curves intersect, the FM signal is some 28 db below the so-called threshold and yet still as good as SSB. These results for the spot-interference case show FM communications to be quite useful in spite of the contrary feelings of many, who believe (Ref. 8) that within a few but unspecified number of decibels of the "improvement threshold" reception is reduced to a very low but unspecified level of usefulness. It should be emphasized that these tests were made using sine-wave modulation of the desired signal. Care should be exercised in extrapolating the test results to include other modulation waveforms such as pulse or speech signals. In general, the main concern is over the peak-to-average power ratio of the modulation wave since the modulation systems are by nature peak limited, whereas the output signal average power is of interest. Unclipped speech signals possess very high peak-to-average power ratios, thus resulting in a low average signal power. Extrapolating these results to include interference signals other than RF noise should not be attempted. One should consider, however, that although other jamming signals may be 5 to 10 db better against an ordinary receiver, counter-countermeasures often may be employed to reduce the effectiveness of even the best, most sophisticated interference signal to that of RF noise. 5. CONCLUSIONS In conclusion, these simple laboratory experiments have shown that FM "thresholds" should be treated as what they are: namely, an arbitrary point on a smooth curve above or below which the curve exists and may be used, depending on the situation. In addition, the curves indicate that FM transmission, both narrow-band and wide-band, possesses competitive performance capabilities for military voice communications use, especially considering some of the practical advantages of building and powering FM transmitters. -8

REFERENCES 1. H. S. Black, Modulation Theory, D. Van Nostrand, New York, 1953, pp. 218-233. 2. M. G. Crosby, "Frequency Modulation Noise Characteristics," Proc. I. R. E., Vol. 25, 1937, p. 472. 3. H. W. Fuller and D. Middleton, "Signals and Noise in an FM Receiver: I. Theoretical Discussion," Cruft Laboratory (Harvard) Technical Report No. 242, February 1957. 4. H. W. Fuller, "Signals and Noise in an FM Receiver: II. Experimental Discussion," Cruft Laboratory (Harvard) Technical Report No. 243, February 1957. 5. S. Goldman, Frequency Analysis, Modulation and Noise, McGraw-Hill, New York, 1948, pp. 270-280. 6. J. Grandlund, "Interference in Frequency-Modulation Reception," Research Laboratory of Electronics (M. I. T. ) Technical Report No. 42, 1949. 7. J. L. Lawson and G. E. Uhlenbeck, Threshold Signals, McGraw-Hill, New York, 1950, pp. 367-383. 8. H. Magnuski, "Jamming of Communication Systems Using FM, AM, and SSB Modulation," IRE Trans. on Military Electronics, Vol. MIL-5, January 1961, pp. 8-11. 9. M. H. Nichols and L. L. Rauch, Radio Telemetry, John Wiley and Sons, New York, 1956, pp. 48-57. -9

DISTRIBUTION LIST Copy No. 1-2 Commanding Officer, U. S. Army Signal Research and Development Laboratory, Fort Monmouth, New Jersey, ATTN: Senior Scientist, Electronic Warfare Division 3 Commanding General, U. S. Army Electronic Proving Ground, Fort Huachuca, Arizona, ATTN: Director, Electronic Warfare Department. 4 Chief, Research and Development Division, Office of the Chief Signal Officer, Department of the Army, Washington 25, D. C., ATTN: SIGEB 5 Commanding Officer, Signal Corps Electronic Research Unit, 9560th USASRU, P. 0. Box 205, Mountain View, California 6 U. S. Atomic Energy Commission, 1901 Constitution Avenue, N.W., Washington 25, D. C., ATTN: Chief Librarian 7 Director, Central Intelligence Agency, 2430 E.Street, N. W., Washington 25, D. C., ATTN: OCD 8 Signal Corps Liaison Officer, Lincoln Laboratory, Box 73, Lexington 73, Massachusetts, ATTN: Col. Clinton W. Janes 9-18 Commander, Armed Services Technical Information Agency, Arlington Hall Station, Arlington 12, Virginia 19 Commander, Air Research and Development Command, Andrews Air Force Base, Washington 25, D. C., ATTN: SCRC, Hq. 20 Directorate of Research and Development, USAF Washington 25, D. C., ATTN: Chief, Electronic Division 21-22 Hqs., Aeronautical System Division, Air Force Command, Wright-Patterson Air Force Base, Ohio, ATTN: WWAD 23 Hqs., Aeronautical System Division, Air Force Command, Wright-Patterson Air Force Base, Ohio, ATTN: WCLGL-7 24 Hqs., Aeronautical System Division, Air Force Command, Wright-Patterson Air Force Base, Ohio - For retransmittal to - Packard Bell Electronics P. 0. Box 337, Newbury Park, California 25 Commander, Air Force Cambridge Research Center, L. G. Hanscom Field, Bedford, Massachusetts, ATTN: CROTLE-2 26-27 Commander, Rome Air Development Center, Griffiss Air Force Base, New York, ATTN: RCSSLD - for retransmittal to Ohio State University Research Foundation 28 Commander, Air Proving Ground Center, ATTN: Adj/Technical Report Branch, Eglin Air Force Base, Florida 29 Commander, Special Weapons Center, Kirtland Air Force Base, Albuquerque, New Mexico 30 Chief, Bureau of Naval Weapons, Code RRR-E, Department of the Navy, Washington 25, D. C. 10

Copy No. 31 Chief of Naval Operations, EW Systems Branch, OP-35, Department of the Navy, Washington 25, D. C. 32 Chief, Bureau of Ships, Code 691C, Department of the Navy, Washington 25, D. C. 33 Chief, Bureau of Ships, Code 684, Department of the Navy, Washington 25, D. C. 34 Chief, Bureau of Naval Weapons, Code RAAV-33, Department of the Navy, Washington 25, D. C. 35 Commander, Naval Ordnance Test Station, Inyokern, China Lake, California, ATTN: Test Director - Code 30 36 Director, Naval Research Laboratory, Countermeasures Branch, Code 5430, Washington 25, D. C. 37 Director, Naval Research Laboratory, Washington 25, D. C. ATTN: Code 2021 38 Director, Air University Library, Maxwell Air Force Base, Alabama, ATTN: CR-4987 39 Commanding Officer - Director, U. S. Naval Electronic Laboratory, San Diego 52, California 40 Office of the Chief of Ordnance, Department of the Army, Washington 25, D. C., ATTN: ORDTU 41 Chief, West Coast Office, U. S. Army Signal Research and Development Laboratory, Bldg. 6, 75 S. Grand Avenue, Pasadena 2, California 42 Commanding Officer, U. S. Naval Ordnance Laboratory, Silver Springs 19, Maryland 43-44 Chief, U. S. Army Security Agency, Arlington Hall Station, Arlington 12, Virginia, ATTN: IADEV 45 President, U. S. Army Defense Board, Headquarters, Fort Bliss, Texas 46 President, U. S. Army Airborne and Electronics Board, Fort Bragg, North Carolina 47 U. S. Army Antiaircraft Artillery and Guided Missile School, Fort Bliss, Texas 48 Commander, USAF Security Service, San Antonio, Texas, ATTN: CLR 49 Chief, Naval Research, Department of the Navy, Washington 25, D. C., ATTN: Code 931 50 Commanding Officer, U. S. Army Security Agency, Operations Center, Fort Huachuca, Arizona 51 President, U. S. Army Security Agency Board, Arlington Hall Station, Arlington 12, Virginia 52 The Research Analysis Corporation, 6935 Arlington Road, Bethesda 14, Maryland, ATTN: Library 53 The Johns Hopkins University, Radiation Laboratory, 1315 St. Paul Street, Baltimore 2, Maryland, ATTN: Librarian 11

Copy No. 54 Stanford Electronics Laboratories, Stanford University, Stanford, California, ATTN: Applied Electronics Laboratory Document Library 55 HRE - Singer, Inc., Science Park, State College, Pennsylvania, ATTN: R. A. Evans, Manager, Technical Information Center 56 ITT Laboratories, 500 Washington Avenue, Nutley 10, New Jersey, ATTN: Mr. L. A. DeRosa, Div. R-15 Lab. 57 Director, USAF Project Rand, via Air Force Liaison Office, The Rand Corporation, 1700 Main Street, Santa Monica, California 58 Stanford Electronics Laboratories, Stanford University, Stanford, California, ATTN: Dr. R. C. Cumming 59 Willow Run Laboratories, The University of Michigan, P. 0. Box 2008, Ann Arbor, Michigan, ATTN: Dr. Boyd 60 Stanford Research Institute, Menlo Park, California, ATTN: Dr. Cohn 61-62 Commanding Officer, U. S. Army Signal Missile Support Agency, White Sands Missile Range, New Mexico, ATTN: SIGWS-EW and SIGWS-FC 63 Commanding Officer, U. S. Army Signal Research and Development Laboratory, Fort Monmouth, New Jersey ATTN: U. S. Marine Corps Liaison Office, Code AO-4C 65 President, U. S. Army Signal Board, Fort Monmouth, New Jersey 66-74 Commanding Officer, U. S. Army Signal Research and Development Laboratory, Fort Monmouth, New Jersey 1 Copy - Dir of Research, DR 1 Copy - Technical Documents Center ADT 1 Copy - Ch, EW Special Devices Br SES 1 Copy - Ch, Advanced Techniques Br SEA 1 Copy - Ch, Jamming and Deception Br SEJ 1 Copy - File Unit #2, Mail and Records ADJ 3 Cpys - Ch, Security Division TS (for retransmittal to BJSM 75 Director, National Security Agency, Fort George G. Meade, Maryland, ATTN: TEC 76 Dr. B. F. Barton, Director, Cooley Electronics Laboratory, The University of Michigan, Ann Arbor, Michigan 77-99 Cooley Electronics Laboratory Project File, The University of Michigan, Ann Arbor, Michigan 100 Project File, The University of Michigan Office of Research Administration, Ann Arbor, Michigan Above distribution is effected by Countermeasures Division, Surveillance Department, USASRDL, Evans Area, Belmar, New Jersey. For further information contact Mr. I. O. Myers, Senior Scientist, Telephone 59-61252. 12