SUPRESSION OF LORANC NAVIGATION SIGNAL IN DIGITAL CAVE

SUPRESSION OF LORAN-C NAVIGATION SIGNAL IN DIGITAL CAVE RADIOS (AN EXPERIMENTAL APPROACH) BCRA Cave Technology Symposium Mr. Antonio Muñoz Group of Technologies in hostile Environments (GTE) University of Zaragoza (Spain)

Group of Technologies in hostile Environments (GTE) http: //gte. unizar. es 2

Group of Technologies in hostile Environments (GTE) http: //gte. unizar. es 3

Outline Loran-C Digital Radios (SDR) Supression algorithm Experimental results Conclusions Group of Technologies in hostile Environments (GTE) http: //gte. unizar. es 4

Loran-C: introduction Radionavigation signal Position is computed by carefully estimating pulse arrival times Operated in chains identified by its GRI (Group Repetition Interval) Stations power ranging from 11 k. W up to 1 MW Signal properties – Uses 20 k. Hz Bandwidth – Estimated dynamic range: 100 d. B Group of Technologies in hostile Environments (GTE) http: //gte. unizar. es 5

Loran-C: why eliminate it? User perspective – – – Introduces a great deal of distortion in voice communication Disturbing noise in absence of communication Contributes to battery discharge Designers perspective – Input dynamic range – Dificulties AGC control – Band has no other source of noise Loran-C perspective – Valuable position determination – Helps to evaluate the electrode coupling/wiring problems Group of Technologies in hostile Environments (GTE) http: //gte. unizar. es 6

Loran-C: signal description (I) Built arround a deterministic pulse envelope Each chain has a master and several slaves – Master: 9 equally spaced pulses (1 ms) except for the last (2 ms) – Slave: 8 equally spaced pulses (1 ms) Pulses have also phase coding to enhance chain identification Group of Technologies in hostile Environments (GTE) http: //gte. unizar. es 7

Loran-C: signal description (II) Master burst 6731 Lessay Slave bursts Loran-C Group of Technologies in hostile Environments (GTE) http: //gte. unizar. es 8

Loran-C: some math Known facts – – GRI times are between 50 and 100 ms Chains have from 3 to 5 stations Pulse amplitude is reduced to ~0. 1% @ 400 us Received Loran-C signal >> received voice link signal Pulse is “active” 400 us / 1 ms (40 %) In each GRI (50 ms) there are 25 (9 + 8) pulses – Channel is “used” 20% of the time (25 * 0. 4 / 50) in worst case Pulses are like gaps in audio signal – To reconstruct it perfectly BWmax < 1250 Hz (FS = 1/400 us) Group of Technologies in hostile Environments (GTE) http: //gte. unizar. es 9

SDR: topology SSB Modulation uses Weaver scheme – Simplementation – Suitable for SDR systems – Frequencies choosen to avoid in band auto generated interferences (PWM armonics) IF 88450 Hz Group of Technologies in hostile Environments (GTE) 1500 Hz http: //gte. unizar. es 10

SDR: signals (I) Group of Technologies in hostile Environments (GTE) http: //gte. unizar. es 11

SDR: signals (II) Group of Technologies in hostile Environments (GTE) http: //gte. unizar. es 12

Supression Algorithm Where goes the algorithm? – RF Signal information is complete(!), processing requirements HIGH (Power consumption) – IF Some signal information is lost, processing req. MEDIUM – BB Most of signal information is gone, processing req. LOW Separate mixing structure? Does it have to be accurate? How Loran-C removal can be done? Group of Technologies in hostile Environments (GTE) http: //gte. unizar. es 13

Algorithm: implementation No separate mixing structure Implemented in IF/BB (48/24 ksps) Must be simple (low resources) No Loran-C synchronisation needed Uses signal power to implement signal detection Group of Technologies in hostile Environments (GTE) http: //gte. unizar. es 14

Algorithm: signal processing (I) Wideband signal is mixed in RF with 88. 450 Hz, “folding” the spectrum arround Fmix = 88540 Hz Group of Technologies in hostile Environments (GTE) http: //gte. unizar. es 15

Algorithm: signal processing (II) Phase & Quadrature signals are decimated to get desired IF/BB frequency DCF 77 Audio information Group of Technologies in hostile Environments (GTE) http: //gte. unizar. es 16

Algorithm: signal processing (III) High pass filter IQ signals (to remove audio information) Compute power of the filtered IQ signals and make two averages, one fast and one slow Group of Technologies in hostile Environments (GTE) http: //gte. unizar. es 17

Algorithm: signal processing (IV) If fast average is n times greater than slow average, then blank signal starting at the point which fast average was greater than slow average Detail (zoomed version) Group of Technologies in hostile Environments (GTE) http: //gte. unizar. es 18

Algorithm: resources Are very dependant of sampling speed Case of BB sampled at 24 ksps: – – Delay lines for I & Q channels (64 samples) Compute IQ Power (2 multiplications + 1 add) Delay line for computed power (32 samples) Fast and slow power averages (2 adds + 2 substractions) Group of Technologies in hostile Environments (GTE) http: //gte. unizar. es 19

Experimental Results SNR is enhanced, but improvement has a strong dependance with relative signal amplitudes (Loran-C vs. Voice) Proposed algorithm introduces little distortion while eliminates some of the annoying noise. As the human hearing has logarithmic behaviour with perceived power, algorithm has to be very accurate to completely eliminate Loran-C perception. Audio samples Group of Technologies in hostile Environments (GTE) http: //gte. unizar. es 20

Conclusions This is our first approach to Loran-C supression Pulse correlation combined with confort noise would greatly enhance the results Use of other bands Group of Technologies in hostile Environments (GTE) http: //gte. unizar. es 21

Thank you!! Antonio Muñoz anmunoz@unizar. es Group of Technologies in hostile Environments (GTE) http: //gte. unizar. es 22

SUPRESSION OF LORAN-C NAVIGATION SIGNAL IN DIGITAL CAVE RADIOS (AN EXPERIMENTAL APPROACH) BCRA Cave Technology Symposium Mr. Antonio Muñoz Group of Technologies in hostile Environments (GTE) University of Zaragoza (Spain)