RETRAM A network of passive radars to detect and track local meteors - first results

Sylvain AZARIAN Jean- Jacques MAINTOUX Frédéric RIBLE Jérémy MAINTOUX RETRAM A network of passive radars to detect and track local meteors - first results Sylvain AZARIAN - Jean- Jacques MAINTOUX contact@retram.org h<p:///

Outline of the presentation What is RETRAM Studied phenomenology Quick reminders on passive radar RETRAM setup for meteor detecyon & tracking Results and extensions NenuFAR and meteor detecyons

What is RETRAM? RETRAM started as an amateur experiment to detect and track meteors heads and trails using passive radar techniques, Project started in 2012, First RETRAM stayon in Orsay (South Paris). VHF Cross- pol yagis VOR + FM

Phenomenology A meteor entering the atmosphere (~60 km/s) loses some of his ma<er (ablayon) and heats surrounding air. T x R x This temperature increase along the trail becomes reflecyve for RF signals (plasma). RF transmissions coming from the ground are then sca<ered by both head and trail. RETRAM uses this phenomenon to detect the falling body. VOR signals sca<ered on meteor

Passive Radar theory quick reminders Passive Radar : «a class of radar systems that detect and track objects by processing reflections from non-cooperative sources of illumination in the environment, such as commercial broadcast and communications signals» Wikipedia

Passive detec?on : fundamentals of passive radar Energy sca<ered on body and reaching receiver 2b d rx d tx 2a P r = P. λ. g t t. g. σ. f ( 4π ) 3 ( d. d ) 2 tx r rx B 1 Where P t is the power at the transmi<er, g t and g r respecyvely antenna gains at transmi<er and receiver, the bistayc radar cross secyon σ d tx distance from object to transmi<er, Is object and frequency dependant d rx distance from object to receiver.

Bista?c Doppler Assuming the transmi<ed signal is wri<en as : 2a s( t) = A( t). e jω( t). t+θ( t) 2b d tx d rx Signal reflected on moving body is frequency shihed (Doppler effect) and becomes : B 1 s( t) = A( t). e jω( t). t+θ( t). e jd opp ( t). t D opp ( t) = 1 d λ t rx + d t Signal visual interpreta0on tricky because resul0ng Doppler shi9 is sum of two par0al deriva0ves Doppler Only target tracking can be done but needs many measurements and does not always work Experiments using VOR transmijers were tried and abandoned: too complex and requires a path hypothesis. tx

Signal delay Signal traveling from transmi<er to object and then from object to receiver is Yme delayed 2b d rx 0.0.0.0.A.B.C.D.E.F.G.H.I.J.K.L... 2a d tx A.B.C.D.E.F.G.H.I.J.K.L... rx( t) = tx( t τ ) = tx( t d d C tx + Receiver gets Yme delayed signal + direct signal. Delay esymayon gives an ellipse of possible locayon rx ) A.B.C.D.E.F.G.H.I.J.K.L... B 1 Possible location τ = d + d C tx rx = const

Signal collected by receiver Meteor trail Meteor head M 2 Finally the receiver collects : 3 Rx 1 4 Tx (1) Direct signal from the transmi<er (2) Signal sca<ered on meteor head (3) Signal sca<ered on meteor trail (4) Signal from environment (5) And noise rx( t) = k. tx( t τ1) 1 + opp α + n. tx( t τ n). e βn. tx( t τ m) + n jd 2+3 4 t m η 5

RETRAM setup for meteor detection & tracking

Experimental RETRAM receiver Antennas to collect direct path + reflected signal GPS Receiver Need to see all directions No directivity required 80 MHz sampling clock SDR Receiver 4 Channels Acquisition board RF Synthesizer Cavity filters BW = 2 MHz Helical filters BW = 4 MHz LNA LNA Acquisition chassis LNA LNA FM Frontend

Extrac?ng Doppler and delay : signal processing FM broadcast signal approx 100 KHz band with limited autocorrelayon (depends on content) Direct Path estimation to build reference signal Doppler/Distance replicas Crossed Yagis Dual-polarization 2 channels receiver Adaptive filter Cross correlation Reflected Path - Target Direct Path - Reference Threshold Detector Detection (distance, speed) rx( t) = k. tx( t τ1) + opp α + n. tx( t τ n). e βn. tx( t τ m) + n Adap?ve filtering to ajenuate unwanted signals jd t m η

Processing ambiguity At the end of the process we have a set of (distances, Doppler) esymayons Using one receiver and two transmi<ers does not solve the locayon problem Meteor Rx FM2 FM1 More receivers + transmitters required going to a network of stations Two possible posiyons (intersecyons of ellipses)

First results and extensions

Trail detec?on Perseids 2013 0 50 Canal=1 Chartres n Bloc=105 Heure=05:41:35 2 1 25 20 0 50 Canal=2 Tour Eiffel n Bloc=105 Heure=05:41:35 2 1 25 20 100 100 15 15 db [km] 150 db [km] 150 10 10 200 200 250 5 250 5 Distance (km) north / south 30 20 10 0-10 -20-30 T1 300 1-500 0 T2 500 vb [m/s] -40-100 -80-60 -40-20 0 20 2 2 R 0 300-500 0 500 vb [m/s] System tuned to work with : First FM transmi<er in Chartres (T1) Second transmi<er on top of Eiffel Tower (T2) Receiver in Orsay 0 Distance (km) east / west

Trail evolu?on Perseids 2013 BistaYc distance to Chartres FM TX BistaYc distance to Eiffel Tower 1.6 x 105 Channel1 3.1 x 105 Channel2 1.55 3.05 1.5 3 1.45 2.95 Db [m] 1.4 Db [m] 2.9 1.35 2.85 1.3 2.8 1.25 2.75 1.2 30 40 50 60 70 80 90 100 110 120 Frames Time : 1 frame = 0,33s 2.7 30 40 50 60 70 80 90 100 110 120 Frames Time : 1 frame = 0,33s This plot shows the evoluyon of the distance / speed for one meteor trail during the 2013 Perseids Shower (12/08/2013 3h41 UTC)

What do we detect? Past trials were conducted during main known showers only, Did not detect head- echoes during this period (mostly confirmed by other observers), Detected several meteor trails (speed very low), Objects detected at more than 350 Km (bistayc d.) Need all day long system to detect less- frequent head- echoes (wavelength used here does not help to detect small objects)

First steps towards a networked system Rx 1 (Range,Doppler) extractor RETRAM receiver (Range,Doppler) extractor RETRAM receiver FM 1 Rx 2 (Range,Doppler) extractor RETRAM receiver Internet FM 2 Step 1 : extend current architecture to new locayons around Paris for data fusion validayon (2014) West : Rambouillet Amateur radio club F6KKR (with City funding support) East : Marne la Vallée Step 2 : Open collaborayon with scienysts and other groups Step 3 : develop a lower cost system for extended system coverage

NenuFAR and meteor Radio detec?on

NenuFAR and Meteor detec?on Supposes signal scajered on meteor and reflected to ground: 1. Frequency coverage? 2. Ability to collect direct path at low eleva?on? 3. Loca?on of object not known a priori need to receive and post beam form for focusing (or focus on expected radiant) Many antennas with beam forming capabili?es : improved direct path cancella?on on receiving array Capacity to detect smaller objects (weaker signals) Beam forming for DOA es?ma?on to limit ambiguity (replaces mul?ple sta?ons) Beam forming Meteor expected direcyon of arrival TXn TX1 Direct Path 1 1 NenuFar channel/ other antenna NenuFAR 1 NenuFar channel/ other antenna Direct path n

Thank you! Ques?ons? More details on

Backup slide Radioastronomie ObservaYons PrédicYble La fenêtre d observayon est adaptée à l objet. Le temps d enregistrement est adapté. Le Temps RT est maitrisé Non prédicyble Météores Temps d observayon long. Temps de RT très important pour des durées d observayon très courtes. DétecYon PrédicYble pour la plupart des cas Temps d observayon limité et adapté Signaux très variables Phénomène furyf et à forte migrayon => traitement adapté Sensibilité Elevée IntégraYon longue Faible IntégraYon faible / MigraYon Ouverture angulaire Etroite Zone couverte réduite souvent limitée par le RT RésoluYon angulaire Elevée La résoluyon angulaire contribue à l analyse Large Faible DirecYon d arrivée non prédicyble sauf pour certains essaims Pas nécessaire sauf si cela était important pour la localisayon Bande passante instantanée Elevée Contribue souvent à l observayon et à la capacité de détecyon Faible Liée à l illuminateur du système passif. Peu ou pas de système large bande. Principe Rayonnement direct Les rayonnements des différents objets sont observés et analysés Type Radar passif Nécessite un illuminateur connu et maitrisé. Radar Passif NA Objets éloignés Illuminateur maitrisé et adapté - CompaYble Bande NENUFAR - Possibilité de récepyon du trajet direct (référence temporel) - Bande passante compayble de la précision recherchée