CA2159801A1

CA2159801A1 - Method and apparatus for fingerprinting and authenticating magnetic media

Info

Publication number: CA2159801A1
Application number: CA002159801A
Authority: CA
Inventors: Ronald Scott Indeck; Marcel Wettstein Muller; George Lawrence Engel; Alan Lee Hege
Original assignee: Individual
Current assignee: Washington University in St Louis WUSTL
Priority date: 1993-04-09
Filing date: 1994-04-08
Publication date: 1994-10-27
Also published as: AU6630194A; AU680901B2; ATE421124T1; EP0693201B1; EP0693201A4; JPH09500225A; US5428683A; US5365586A; WO1994024638A1; EP0693201A1; BR9405859A; DE69435183D1

Abstract

A method and apparatus is disclosed for determining the remanent noise in a magnetic medium (102) by DC saturation of a region thereof and measurement of the remaining DC magnetization. A conventional magnetic recording transducer may be used to determine the remanent noise. Upon determination, the remanent noise may then be digitized and recorded on the same magnetic medium to thereby "fingerprint" the magnetic medium. This "fingerprint" may then be later used to verify and authenticate (114) the magnetic medium as being an original. In such manner, any magnetic medium, or any object having an associated magnetic medium, may be "fingerprinted"
including credit cards, computer programs, compact discs, videotapes, cassette tapes, etc.

Description

094/~U~8 S9 8 01 PCT~S94/03917 METHOD AND APPARATUS FOR
FINGERPRINTING AND AUTHENTICATING MAGNETIC MEDIA

Cross Reference to Related Application This application is a continuation-in-part of application Serial No. 08/046,040 filed April 9, 1993.
Backqround and Summary of the Invention The sources of noise in a readback signal from a magnetic recording medium have been investigated and identified. One of those sources includes the irregu-larities and defects in the microstructure of the magnet-ic medium itself. For many years, the noise generated 10 from this source has been thought, as with the noise generated from other identified sources, to be random and subject only to statistical analysis for its determina-tion. The inventors herein have recently demonstrated that this noise component is instead deterministic, i.e.
15 is permanent and repeatable, depending entirely on the head-medium position and on the magnetic history of the medium. As confirmed by experiments conducted by the inventors herein, when the medium has had no signal writ-W094/~U~8 ~9 PCT~S94/03917 -ten on it and has been exposed only to DC fields, the observed readback signals are almost identical. The magnetic contribution to the readback signal under these conditions results from spatial variations in the 5 medium's magnetization: ~gn~tic domains, ripple, local fluctuations of the aniso~-~y field and saturation mag-netization. These local properties, in turn, are affect-ed by the morrholo~y and magnetic properties of the indi-vidual grains which make up the ~9m~ n and which do not lO change after deposition. Hence, the noise from a nomi-nally uniformly ~gne-tized region measured at a fixed position on a ~gnetic medium is reproducible. As shown by the inventors herein, a magnetic medium may be DC
saturated and its output then measured to determine its 15 r~nent state or r~n~nt noise. The inventors have confirmed that this rem~n~nt noise is a function of the magnetic mi~lox~l~cture by ~omrA~ing the r~-~nent noise after a positive DC saturation with the r~-n~nt noise after a negative DC saturation. It has been found that 20 these wave forms are virtual "mirror images" of each other thereby demonstrating a close correlation. Simi-larly, other methodologies were used to confirm that the re~qnent noise was deterministic, repeatable, and related to the physical microstructure of the magnetic medium

2~ itself. Remanent noise arising from the permanent micro-structure ~h~ h~ ts identifiable features characteristic of that permanent mi~lo~ cture after pract~nAlly any magnetic history. See Spatfal Nofse PAenomena of ~ongf-t~fn~7 Magnetfc ~ecord~ng Media by Hoinville, Tn~e~k and 30 Ml~ ~, IEEE Transactions on Magnetics Volume 28, No. 6, November 1992, the disclosure of which is in~ ulated herein by reference.
There is a long felt need in the art for a method and apparatus to identify or fingerprint various kinds of 35 documents as well as the wide variety of prerecorded magnetic media presently being marketed and/or distribut-~0 94/2~8 .~ ~ r L PCT~S94/03917 i ' ~J ' ~' ed in the United States and throughout the world. Exam-ples of these magnetic ~ include those produced and sold in the entert~inm~nt industry including magneto-optic discs and tapes, cassette tapes, reel to reel 5 tapes, videotapes, etc. Still another ma;or market in ro~etic media is the tr~?nAous volume of computer pro-grams routinely sold and/or distributed on floppy dis-kettes. Magnetic media are also used for other purposes for which it is i~ oL~ant to be able to identify and 10 authenticate originals including videotapes, ~-~æ~-tte tapes, and other prerecordings on magnetic ~-A;~ of tele-phone conversations, video recordings of criminal activi-ties, and other such investigative and Aoc~ ntary uses.
Still another example of a need in the art for authenti-15 cation and verification of magnetic ~^A; ~ lies in the ~gnetiC data card field. Examples of magnetic data cards include the well known credit card as well as ATM
cards, debit cards, security or ID cards, mass transit cards, and even airline tickets or other vo~lche~s which 20 have magnetic stripes thereon for the magnetic recording of data. As well known in the art, virtually every mag-netic data card has a n ~gnetic stripe of prerecorded r~g~etic data which is used to record the customer's account number or some other such identifying data. Tre-25 mendous sums of money are lost Ann~ y through LG1Y~1Y
and other fral~Alllent copying and use sch~c which could be virtually ~~ n~ted if an apparatus and methoAolo~y could be implemented for reliably authenticating and verifying the identity of a ~g~-tic data card prior to 30 its being a~ ~ved for its associated transaction. Still other examples extend to paper doc~?nts and the like for w which there has been some specific efforts of which the inventors herein are aware.
As related in an article entitled Novel Applica-35 tions of Cr~ptography in Digital Cn~mm77n;cations by Omura, IEEE Communications Maqazine, May 1990, a t~hn;que is W094/2~8 ~s ~ - ~ PCT~S94/03917 -disclosed for creating counterfeit-proof objects. As related therein, the basic idea is to measure some unique "fingerprint" of the paper and to sign (encrypt) it using the secret key of the manufacturer of, for example, a 5 stock certificate. The fingerprint is obt~ne~ by moving a narrow intense light beam along a line on the paper and measuring the light intensity that passes through the paper. The light intensity function determined by the unique random pattern of paper fibers along the line then 10 forms the fingerprint of the particular piece of paper.
This fingerprint is then digitized and encLy~ed by the secret encryption function. The encrypted fingerprint is then separately printed onto the paper in digital form such as a bar code. At a later date, the authenticity of 15 the stock certificate may be verified by using a non-secret public decryption function to de~ly~ the encrypt-ed data on the paper and reconstruct the intensity func-tion, or fingerprint, that was recorded thereon. Next, the actual intensity function of the stock certificate is 20 measured. If this newly measured intensity function agrees with the intensity function reconstructed from the de~ly~ed data, the document may be ~.r.l ~ed authentic.
This scheme takes advantage of a well know secrecy ~yS~..
referred to as a public key crypto~ys~-,. This ~y~
25 employs a trap door one way function. A user chooC~ a secret key (the trap door) and after applying the trap door one way function to the data, the procedure deter-mines an algorithm used for decoding which is made pub-licly known. The trap door one way function is also used 30 to produce the en~ly~ed m~ge. Then every other user can understand the original message by applying the algo-rithm to the ~ly~ogram. In this system no one else can produce a publicly readable message attributable to the original user's trap door as only the user has knowledge 35 of that algorithm. This prevents the simplistic forgery ~0 9412K~8 fl ~ Q ~1 PCT~S94/03917 attempt of changing the pre-recorded fingerprint to agree with a forged document's fingerprint.
Still another example of an attempt in the prior art to fingerprint or counterfeit-proof objects is shown 5 in U.S. Patent 4,806,740. As shown therein, an object, such as a stock certificate, has deposited thereon a stripe of magnetic medium having a variable density re-sulting from the non-uniformity of the paper, the process of depositing the magnetic medium on the document, and 10 the dispersion of magnetic particles within the medium.
The density variations are r~nAs~ly created as the mag-netic medium is applied, which affords a unique doc~l~ent as these density variations are fixed and repeatable to identify the document. A second ~gnetic stripe is also 15 applied to the document, but this magnetic stripe is ~ol..~Lised of a medium that is tightly specified and high-ly controlled in accordance with well known st~nA~ds in the recording art to be part of a magnetic read/write system. In operation, the non-uniform ~gnetic stripe is 20 erased, recorded by a stAnAA~d recording comprised of a A~ DC signal or a lineA~ AC signal or a l~neA~ bias ~ig~Al. After recording, another head senses the magnet-ic characteristic of the recorded magnetic stripe which is translated into a digital, ~Ch; n~ readable format, 25 and then separately recorded on the second ~gn~tic stripe in a simple write function. For authentication, the stock certificate is passed under another set of heads ~Th~ ~h first reads the digitally recorded m~chi ne reAAAhle representation of the sensed noise signal and 30 then a ~conA set of heads reads the variable density magnetic stripe by first erasing it, recording the same st~nAA~d noise function, and then S~n~; ng the output of the prerecorded noise function as it is "distorted" by the variable density magnetic stripe. If it matches the 35 recorded representation thereof, then the document is declared to be authentic and original. Thus, with the W094/~8 ~ Sg~O ~ PCT~S94/03917 -method of the '740 patent, a pair of magnetic stripes must be applied to the document and a specified signal (~nom;n~ted as noise) must be recorded, measured, and then its output digitally recorded. Furthermore, one of 5 the magnetic stripes must be applied in other than re-cording industry stAn~d and in a random manner to en-sure the rA~o~n~ of the output thereof. These steps make the '740 patent method difficult and in~o~v~l~ient to implement.
Yet another example of a prior art attempt to ut~ e a ~gnetic fingerprint of a magnetic medium for authenticating credit cards, documents, and the like is found in Pease et al U.S. Patent 4,985,614 issued on June 15, 1991. This '614 patent is actually quite s~l~r in 15 ~o~c~pt to the '740 patent discussed above in that it focuses on the macroscopic, hereinafter denoted "macro"
variations in a r-gn~tic medium, and their effect on an ~nh~nc~n~ signal recorded thereon in one embo~m~nt or st~nfl~ng alone in a ~ç~on~ embodiment. With either em-20 bodiment, these "macro" variations are determined byreA~i~g a chosen length of approximately 2.6 1nch~ of a magnetic stripe between 3 and 9 times (5 in the preferred embodiment) and then correlating the collected data points to "average out" the effects of head noise, elec-25 trical noise, and any other non-medium noise. This cor-relation results in a "representative profile" which represents the vari~nc~ which would be in~llc~ by these macro effects to a ~gn~l if it were recorded on this 2.6 inch portion of r~gnetic stripe. If these variations are 30 not significant enough to produce a reliable correlation, indicating a lack of significant macroscopic nonlln~for-mities in the medium, the medium is discarded. This is an indication that the medium has been manufactured too closely to tolerance, or otherwise does not have enough 35 macro level variation which might be present due to a manufacturer's watermark or the like, to induce reliably ~0 9412U~8 r ~ ~l PCT~S94103917 r detectable and repeatable variations to a recorded sig-nal. The '614 patent also suggests that macro level noise may be ~nhAnc~ by locally altering the apparent magnetic characteristics of the stripe such-as by pl~c; ng 5 magnetic symbols on the substrate underlying the ~gnetic region, or by e~hoc~ing selected regions of the ~g~etic material so as to physically move some amount of the material. As the noise levels measured have significant effects on the peaks of a recorded ~nh~n~;ng signal, a lO simple peak detect and hold circuit is taught as suffi-cient to collect the data, and a simple "comparison" of the pre-recorded "representative profile" with the pres-ently sensed data points is taught as sufficient to de-termine if the medium is authentic. Therefore, not only 15 does the '614 patent focus on the use of macro level noise, its device and methodology disclosed for imple-menting a macro level noise detector is believed to be incapable of reliably creating a microstructure noise level fingerprint and validating its existence at a later 20 time in order to authenticate an original.
In order to solve these and other problems in the prior art, the inventors herein have developed a method and apparatus for ut~ ;ng the unique, det~rr;n~tic, remanent noise characteristic of the magnetic medium it-25 self due to its magnetic microstructure to fingerprintnot only ~oc~l~ents~ but other objects and more i"~ol~ant-ly, the magnetic medium itself so that it can be identi-fied and authenticated. This inventive t~chn;que relies upon the discovery that the microscopic structure of the 30 magnetic medium itself is a per~nent random arrangement of microfeatures and therefore deterministic. In other words, once fabricated, the recording medium's physical microstructure r~;n~ fixed for all conventional record-ing processes. In particulate media, the position and 35 orientation of each particle does not change within the binder for any application of magnetic field; in thin W094/~K~8 ~ PCT~S94/03917 -2~s9~

film media, the microcrystalline orientations and grain boundaries of the film remain stationary during the re-cord and reproduce processes. It is the maqnetization within each of these fixed microfeatures that can be 5 rotated or modified which forms the basis of the magnetic recording process. If a region of a magnetic medium is saturated in one direction by a large applied field, the r~nent magnetization depends strongly on the micro-structure of the medium. This rem~nt state is deter-10 ministic for any point on the recording surface. Eachparticle or grain in the medium is hundreds to thousands of Angstroms in ~im~n~ion. Due to their small size, a small region of the magnetic surface will contain a very large number of these physical entities. While the fab-15 rication process normally includes efforts to align theseparticles, there is always some dispersion of individual orientations and positions. The actual deviations will be unique to a region of the medium's surface making this orientation a signature or a "fingerprint" of that medi-20 um. To reproduce this distribution, intentionally ornot, is practically impossible since this would entail a precise manipulation of the orientation of numerous par-ticles at the submicrometer level. Thus, the orientation of a large set of particles on a specific portion of a 25 recording surface can uniquely identify that medium. In experiments, the inventors have found that the rP~nent noise from a length of between about 30 micrometers and 4300 micrometers provides enough data to "fingerprint" a magnetic medium. This may be contrasted with the 66,040 30 micrometers (2.6 inchps) of length required in the '614 patent in order to fingerprint a magnetic medium with macro noise.
In essence, the present invention is elegantly simple and adapted for implementation by conventional 35 recording heads as are commonly found and used in virtu-ally every read or read/write device presently utilized 094/~ S98ol PCT~594/03917 by the public at large. Such examples include credit card readers, magneto-optic disc players, cassette play-ers, VCRs and personal computers. Furthermore, a card reader may be coupled with virtually any device or pro-5 cess, and the card reader used as a "gatekeeper" to per-mit input or access only by those who can present a valid p~s~c~d for authentication. In its simplest ~mrl~?~ta-tion, a conventional recording head need merely DC satu-rate a specified portion of a ~etic medium, and then l0 ~read" or "play back" the remanent noise which 1~ ~i n~, For co11~e11ience, the fingerprint may be obt~ine~ from the region between two recorded magnetic transitions already in place on the medium. This rem~n~t noise, which is an analog signal, may then be digitized and recorded, in the 15 medium itself or elsewhere, in m~hi~ readable format perhaps using a trap door function. Thusly, the ~ tic medium has become "labeled" with its fingerprint. Veri-fication or authentication of that magnetic medium is simply achieved by reversing this process except that in 20 the more security sensitive applications the digitally recorded fingerprint must be decrypted using the publicly known key. Should the measured r~m~ent noise match the remanent noise as recorded, the magnetic medium is au-thenticated.
There are many variations in ut~ Ation of the inventors' method and apparatus which ~xp~n~ its universe of applications. For example, some applications need not re~uire the use of a trap door function such as, for example, when the encoded objects are not publicly dis-30 tributed and instead are being identified solely for the user's purposes. One such example would be for use with inventory items.
Still another application involves the "copy protection" of mass distributed application software.
35 Over the years, many schem~ have been tried and almost uniformly abandoned for copy protecting publicly distrib-W094/~4~8 2 ~ 5 9 8 0 ~ PCT~S94/03917 -.

uted diskettes of prerecorded software. This has hap-pened for many reasons including the problem that almost all of the copy protection schemes previously implemented interfere with the r--nni~g of the software on the user's 5 computer. With the present invention, a copy protection scheme may be implemented which does not interfere with the rllnn~ ng of the software and instead merely provides a pr~-onA~tion to rllnni ng of what is otherwise normally written code. In its implementation, a software diskette 10 may first instruct the computer in which it is inserted to read a fingerprint of a specified portion of the dis-kette and comr~re it with a prerecorded version of the same fingerprint. If the fingerprints match, then the software may permit the computer to further read and 15 implement the application software stored thereon. How-ever, if the fingerprint detected by the computer does not match that which is stored in the software, then the software itself may inhibit further r~ ng of the pro-gram and ~levellt its implementation. This would abso-20 lutely prevent a user from making a copy of a program foruse by s~- ?one else. This scheme may also be slightly modified as ~i~qll~sed in the detailed description of the preferred embodiment to permit a user to make a single archive or backup copy such that the finye~lint nomr~ri-25 son permits the first non-mat~-h~g fingerprint copy to be run but then ~.ev~llts any other non-matqhi ng finger-printed copies to run. This implementation is easily achieved and "copy protects" application software reli-ably, inexpensively, and requires only minor hardware 30 changes to the massive number of computers already in consumers' hands.
Still another significant application of the present invention involves authenticating credit cards using the single magnetic stripe already implemented on 35 most major credit cards. Again, this may be contrasted with the '614 patent which suggests that a second stripe -094/2~8 $95l PCT~S94/03917 be added because of the re~uired 2.6 inches of stripe length which must be dedicated to the macro fingerprint t~-hn~ques. The same method would be used as explained above to measure and record the "fingerprint" of the 5 particular magnetic stripe cont~;neA on a particular credit card and then a credit card reader would require that same fingerprint to be matched every time it is used to verify its authenticity. While there are already a large number of credit cards in circulation, these cards 10 are routinely subject to expiration such that there is a conff nuAl repl~c^ment of these cards in the public's hands. Thus, over time the installed base of credit cards could be readily transformed to those which have been "fingerprinted".
In a variation to this application, the present invention may be coupled with a data base or processor, such as in so-called Smart Cards. These credit card-like devices actually contain, in addition to perhaps the st~n~d credit card r~gn~tic stripe, an on-board elec-20 tronic memory and/or microprocessor. ~his memory or mi~Lo~locessor may contain all sorts of information in-cluding money substitute data. For example, at present a large number of these smart cards are in use in Europe as pre-paid telerhon~ cards which are pre-loaded with a 25 monetary amount which is charged against by a pay phone.
The cards are used until their pre-loaded monetary e~uiv-alent has been depleted and then they are discarded.
While various security methodologies have been developed to protect against fraud, these are subject to breach.
30 The present invention is uniquely suited as a security ~ch~ ~ for smart cards as it depends solely on the mag-netic mi~o~Ll~cture of the particular magnetic medium.
In use, the r-gn~tic fingerprint could be stored on the magnetic stripe, in the smart card memory (on board the 35 card), or in a central computer. When coupled with a trap door function, no fraudulent card could be created W094/~U~8 2 1 S 9 8 0 ~ PCT~S94/03917 ~

` 5 ~ ~ 12 without ~ s to the trap door function and every trans-action could be quickly pre-authorized at a local card reader, without phon; ng a central clearing authority. In an extension to all credit card applications, the finger-5 print data may be stored along with each transaction sothat a complete record or trail is created which traces a particular card's history. Thus, the present ~m~1 y used ~h~ - where a number of fraudulent cards are creat-ed with a correct but stolen account number could either 10 be thwarted or effectively prosecuted.
Another level of security incorporates random placement of the fingerprint position. This might be a function of the card's number. For example, the card number modulo "P" might point the read electronics to a 15 particular data bit around which the fingerprint will be found.
Still another significant category of applications involves utilizing the present invention in its gatekeep-er function. Any ~yst-~ , process, ~h;n~, location, or 20 other function to which ~cc~ss is desired to be restrict-ed to only those who are authorized, the present inven-tion provides a unique and reliable solution. In its simplest implementation, a passcard may be created with a magnetic stripe which is fingerprinted in accordance with 25 the present invention. Although examples will be dis-~llsce~ in terms of utilizing a passcard, it should be undel~ood that any magnetic medium can be s~ rly used in accordance with the t~ch;ngs herein. As such, all other such examples and implementations are int~n~ to 30 be included within the present invention and shall be understood to be included within the term "passcard".
This passcard may then become a personal ID card which may be used not only to control access, but also identify the particular person ~r~ing the service, function, 35 etc. by storing the particular magnetic fingerprint of the card being used. Numerous examples may be readily ~ 094/2~8 1S~801 PCT~S94/03917 considered. For example, access to a computer network through a remote terminal may be controlled utilizing a ~ p~A~d of the present invention. ~his would be imple-mented through the use of a diskette which may be readily - 5 inserted in any floppy disk drive which could authenti-cate the fingerprint on the diskette. Alternatively, an inexpensive card reader, adapted to read a passcard, could be utilized as well. Many other applications would utilize the modified card reader. For e~mple, a bank lO teller may be assigned a passcard which could then be used to track all of the transactions entered by the teller and thereby more reliably guard against teller fraud. The myriad of identification cards utilized by businesses, health plans, universities, hospitals, and l5 other organizations or facilities could readily adopt and use a p~c~d to more securely identify and preauthorize the users of its services, facilities, etc. Not only would existing uses be readily ~?nahle to replAc~ent with the p~CX~d of the present invention, but other new 20 services and systems could be implemented because of the high degree of security provided by the present inven-tion. This may include home shopping and pay-per-view video. This may well lead to the creation of national data bases, national ID cards, and other more universal 25 implementations of credit cards or passcards. This is especially true if a ~yx~l, utilizes not only the ~^gnet-ia fing~L~llnt of a particular passcard, but also utiliz-es a s~con~ry security check such as a picture ID, h~l~ n fingerprint, hologram (presently imprinted on credit 30 cards), or other such methodology which would thereby render the passcard system virtually impregnable. With such security, individuals may be more wi~l; ng to turn over such detailed personal financial and health informa-tion as would make these systems feasible.
While the principal advantages and features of the invention have been described above, and a number of W094/~8 ~S9~ PCT~S94/03917 -examples given, a greater underst~n~; ng of the invention may be att~;n~ by referring to the drawings and the descr~ption of the preferred embodiment which follow.
Brief Description of the Drawings Figure 1 is a magnified representative depiction of the microscopic structure of a region of magnetic medium;
Figure 2 is a magnified depiction of several tracks of a magnetic medium having microscopic structure 10 representatively shown thereon;
Figure 3 is a partial view of a track of ~gnetiC
media having its fingerprint recorded thereon in ~h; ne r~ ble code;
Figure 4 depicts three conventional recording lS heads and a magnetic medium traveling ther~~ e~;
Figure S is a view of a credit card having finger-print data ~nco~e~ thereon for r~;ng by a credit card reader;
Figure 6 depicts a personal computer with a com-20 puter diskette for insertion in a floppy disk drivethereof;
Figure 7 is a perspective view of a magneto-optic disc player with a magneto-optic disc in its tray;
Figure 8 is a cassette player depicting a cassette 25 tape for play therein;
Figure 9 is a perspective view of a VCR with a tape ready for insertion;
Figure 10 is a block diagram of a magnetic finger-print verification circuit;
Figure 11 is a block diagram of the magnetic trigger circuit shown in Figure 10;
Figure 12 is a schematic diagram of an implementa-tion of the present invention ut;~ ng a PC;
Figure 13 is a schematic diagram of the memory 35 ut;l;~ in the implementation of Figure 12;

~ 094/2~8 PCT~S94/03917 ~I Sg Figure 14 is a schematic diagram of the trigger circuits utilized in the implementation of Figure 12;
- Figure 15 is a schematic diagram of the preamp circuits utilized in the implementation of Figure 12;
- 5 Figure 16 is a block diagram of a ~gnetic inger-print verification circuit set up for implementation in an IC;
Figure 17 is a schematic diagram of a correlation circuit ut~ ng single bit data streams;
No Figure 18;
Figure 19 is a schematic diagram of an active differentiator;
Figure 20 is a schematic diagram of the threshold generator;
No Figure 21;
Figure 22 is a schematic diagram of the ADC refer-ence generator;
Figure 23 is a schematic diagram of a gain cir-cuit;
Figure 24 is a plot from a read of a magnetic credit card stripe;
Figure 25 is an enlarged view of the encircled portion of the waveform in Figure 24; and Figure 26 is a waveform giving the correlation of 25 two fingerprints.
Detailed Descri~tion of the Preferred Embodiment As shown in Figure 1, a region of magnetic medium 20 is built up with a plurality of miCrOClyx~Al l in~.
structures 22 in a random pattern. This micro~lyx~alline 30 structure 22 is comprised of particles or grains varying from hundreds to thousands of An~lol.ls in diameter. The view of Figure 1 is greatly enlarged and magnified in order to depict this physical microstructure. As shown ~ in Figure 2, this microcrystalline structure extends 35 throughout the magnetic medium even though the magnetic W094/~4~8 2~5~ PCT~S94/03917 medium 24 shown in Figure 2 may be itself comprised of tracks 26, 28, 30 as well known in the art. Although shown schemat~Ally as separate regions, the fingerprint can be obt~~ from any portion of the medium 24.
Referring now to Figures 3 and 4, a plurality of conven~;on~l recording heads 32, 34, 36 are shown mounted in a head transport 37 with a traveling magnetic medium 38 controllably driven past recording heads 32, 34, 36 all as is well known in the art. These recording heads 10 32-36 may be any magnetic transducer or magneto-optic transducer head, as known in the art. Recording heads 32-36 are all co~n~cted to electronic circuitry 39, as well known in the art, to control and read their input and output and to further process signals for playback or 15 other use. Although only three heads 32, 34, 36 are being shown in Figure 4, it will be well unde~ood to those of ordinary skill in the art that a plurality of recording heads of any number may just as easily be pro-vided and, as taught herein, may be required in order to 20 effect the purposes of the present invention. As shown in Figure 3, the magnetic "fin~el~lint" at a specified region 40 of a thin film magnetic medium or tape 42, shown representationally in Figure 3 as a thin film tape, may be recorded at a second position 44 on said thin film 25 magnetic medium or tape 42 in a digitized, m~Ch~ n~ read-able code 46 or the like.
As their preferred embodiment, the inventors have utilized a methodology for re~; ng or deterr; n; ng the r~n~nt microstructural noise characteristic of the 30 region 40 of the magnetic medium which is being "finger-printed". Preferably, this region 40 is on the order of several tens to hundreds of micrometers. This region is then DC saturated and then subjected to a "read" step for deter~; ni ng the r~m~nent noise produced thereby.
While this is the preferred embodiment, it should be understood that the fingerprint is always there, ~0 9412~8 1 S9 8l PCT~S94/03917 whether the medium has been recorded over or not. There-fore, it is not strictly n~ sary that the specified portion of medium contA; n; ng the fingerprint be DC satu-rated, or DC saturated in the same polarity in order to 5 obtain the fingerprint. Instead, it is only important that the re-~n~nt noise be determ; n~A in a r-nn~r which facilitates its being correlated successfully with the earlier deter~;~e~ r~m~n~nt noise.
If this information is obt~;ne~ in a "single shot"
lO measurement, then the results will obviously include both electronics noise as well as the rc~ent noise attribut-able to the particles' orientation. As this "noise" or n re~n~nt noise" is electronically determined as an ana-log signal, this information may then be digitized and 15 recorded with about a hundred to two hundred digital bits of inormation as may be representationally shown as code 46 in Figure 3. In experiments, the inventors have made multiple measurements and averaged their results in order to e~ n~te the elec~lol1ics noise present in the mea-20 sured wave form. However, there was observed a highcorrelation coefficient when the two sets of data, i.e.
single shot and averaged, were compared thereby demon-strating that a single shot reading could readily be used in comparison to an averaged set of data in co~cial 25 application. The norm~ ed cross correlation coefficient r is used where N

~ (x1 - x)(Y1 ~ Y) r =
N _ N
i--l ( Xi x ) ~ l(Y~ -- Y ) as explained by the inventors in their earlier pub~ e~
35 article mentioned above.
In order to recover or measure the "fingerprint"
or re~n~nt noise, the process is similarly repeated and, when comparing two single shot wave forms, a smaller 59~~ ~
W094/2~8 l ~ PCT~S94/03917 .

correlation therebetween was experienced. However, the correlation experienced with two single shot wave forms was significant and clearly demonstrated this method's feasibility for ~o~ercial application as well.
As shown in Figure 24, the portion of the signal used for fingerprinting is very small with respect to the rest of the recorded signal. As shown in Figure 25, the encircled portion or fingerprint from Figure 24 may be amplified to show in greater detail the waveform. In 10 Figure 26, a correlation using the present invention produces a definable "peak" which verifies the existence of the fingerprint in the medium.
As shown in Figure 5, a practical implementation for the subject invention includes a magnetic data card 15 48 which has a magnetic stripe 50 thereon with magnetic stripe 50 being encoded with a code 52 representative of a fingerprint of a region 54 of magnetic stripe 50.
Thus, as the magnetic data card 48 is "swiped" through a card reader 56, the card reader 56 may read the code 52 20 to determine the stored fingerprint data, read the fin-gerprint at region 54 of the magnetic stripe 50, compare them for a match, and if they match then authenticate magnetic data card 48 as a genuine card which has not been altered and which may be approved. Alternatively, 25 the fingerprint need not be stored on the card but may instead be stored centrally, as in a data base elsewhere.
As shown in Figure 10, a schematic block diagram for a magnetic fingerprint prototype includes a read head 100 for reading the magnetic medium 102 which may be on a 30 credit card or passcard 104 as previously described. A
magnetic trigger circuit 106 (including the gain circuit shown in Figure 23) pulses on a logic element 108 which activates an analog to digital converter 110 (including a reference voltage generator shown in Figure 22) to con-35 vert the output from read head 100, V,, into a stream ofdigital data which is stored in a memory 112. A

~ 094/2K~8 æl $9~1 PCT~S94/03917 microcontroller 114 then processes the data and compares it with the original fingerprint in order to authenticate the credit card or passcard 104. The magnetic trigger circuit 106 is shown in greater detail in Figure 11. It 5 includes a preamp 116 (shown in greater detail in Figure 15) which amplifies the output from read head 100 to pro-duce, through a set of analog comparators (see Figure 14) with thresholds produced by threshold generators (see Figure 20), a positive pulse output 118 and a negative 10 pulse output 120, as shown by the t;m;ng graph in the lower half of Figure 11. The logic 108 may be implement-ed as shown in Figure 12 by connection to an IBM PC
through connector 122. A memory element 124 is shown in greater detail in Figure 13, trigger circuits 126 are 15 more completely shown in Figure 14, and preamp circuits 128 are shown in Figure 15. A block diagram 130 for a magnetic fingerprint device is shown in Figure 16 which is arranged for implementation in a custom integrated circuit.
In many of the implementations of the present invention, the speed at which the magnetic medium is propelled past a recording head, card reader, or the like is fixed both when the magnetic fingerprint is first measured and, later, when the magnetic fingerprint is 25 read in order to verify or authenticate its validity.
Some examples of this fixed reading speed include the speed at which a floppy disk is rotated in a computer, the speed at which a VCR tape is played, the speed at which an audio cassette tape is played, the motorized 30 reading of an ATM card, etc. This is even true for cer-tain credit card applications where motorized card read-ers could be implemented. However, there are also other applications for which a controlled speed cannot be ex-~ pected. Perhaps the most prevalent situation known today 35 is the widely used credit card readers which require aretail clerk to "swipe" the card through a reading track.

W094/~8 PCT~S94/03917 ~
?~S9~
` 20 For those situations wherein the speed of the magnetic medium is variable, or different from when the finger-print was first determined, the inventors have developed a number of methodologies to ensure that data correspond-5 ing to the magnetic fingerprint are captured and used inthe verification process. This helps e~ n~te improper rejections of valid credit cards, passcards, and the like. For clarity, these various methodologies will now be explained with respect to a credit card having a mag-lO netic stripe. However, it should be understood thatthese methodologies are e~ually applicable to any appli-cation wherein the speed of the magnetic medium is vari-able or potentially variable from that of the original fingerprinting step.
With the magnetic fingerprint whose block diagram is shown in Figure 10, a series of data points, perhaps 150, are taken between the two trigger pulses recorded in the magnetic medium. As shown with the magnetic trigger circuit of Figure 11, the signal "POS PULSE" becom~c ac-20 tive when the "center" of a sufficiently large positive going pulse is detected. Similarly, the signal "NEG
PULSE" is active when a sufficiently large negative going pulse is detected. The derivative of the incomi ng signal is taken using an active differentiator (shown in greater 25 detail in Figure 19) in order to locate the center of the pulse. By locating the center of the pulse, the distance between the center of the pulses which define the finger-print area is thus fixed and represents a distance which is reliably ascert~n~hle every time the credit card is 30 swiped through a credit card reader. A level sensitive detection is also performed but this is only to help guard against false trigger events. The A to D converter samples at a fixed rate such as F~1. Thus,-the spacing between samples is delta xl where delta xl is equal to Vl ~
(velocity of the credit card) divided by F~1 (the sampling rate). If the sampling frequency F,l is a few hundred 094/2~8 S9s~l PCT~S94/03917 kHz, then delta x1 is on the order of one micron. The number of samples taken, P (perhaps 150), is counted and may be recorded on the credit card as the fingerprint.
In order to improve the reliability, several r~A~;~gs of 5 the magnetic fingerprint may be made and then averaged in order to eliminate the effects of head noise, electronic circuit noise, and any other noise other than the magnet-ic microstructure noise of the magnetic medium. This completes the process of determ; n; ng the fingerprint for 10 a credit card.
When the card is in use, and its fingerprint is desired to be authenticated, it is necessary to sample the same region of the magnetic medium of the card which was sampled at the time of its fingerprinting. Further-15 more, the distance between sample points must be the sameas it was when the card was fingerprinted and, in our example, this distance is delta xl. Although it is de-sired for this to be exactly the same, the inventors have found that some variance, up to a few percent, may be 20 acceptable. The sampling interval is determined both by the sampling frequency of the A to D converter and the velocity at which the credit card moves past the read head. For ~nllA 1 card rPA~; ng applications, the vast majority presently in use, retail store keepers are vir-25 tually assured to swipe the cards at different ratesthrough the card readers. One solution for obtA; n; ng samples at the same intervals as the original sample interval is to greatly oversample. In other words, a sampling rate F~2 is chosen which is much greater than F~1.
30 This creates a new sampling interval delta x2 which pro-duces many more samples, perhaps 100 times more, than were taken in the original fingerprinting process. In other words, instead of 150 samples, 15,000 samples may be taken. This requires that only every Mth sample be 35 utilized in order to provide the same data set. M may be chosen as the ratio of Q/P where P is the number of sam-ples taken between the two triggering pulses at the time W094/~8 PCT~S94/03917 -. ~, . ..
215980i of fingerprinting and Q is the number of samples taken at the point of sale using an oversampling frequency F~2.
Since the distance between the trigger events does not change, and the ratio of F,1 to F6z is known, the velocity 5 of the card as it is swiped at the retail store location compared to its velocity at the time of fingerprinting may be readily determ;n~A. With this oversampling tech-nique, it may be readily shown mathematically that over-sampling by a factor of approximately 100 will result in 10 an effective sampling interval at the point of sale which will be within 2% of the original sampling interval for velocities which are as much as five times greater than the velocity used at the time that the fingerprint was determined. If necessary, an even greater oversampling 15 rate would accommodate even larger velocity ratios.
The work by the inventors in prototyping the present invention indicate that a digital word size as small as three bits for ~nco~;ng the value of the finger-print samples may be sufficient to obtain acceptable 20 results using the proposed correlation analysis tech-nique. This small word size suggests that a custom inte-grated circuit (IC) may be designed to conduct this cor-relation. It is well known in the art (for example see the literature on sigma-delta conversion) that n-bit 25 words at a rate r1 can be constructed from a one-bit data stream provided that the data rate of the one-bit data stream is significantly higher than the rate r1. In its simplest form, the word length can be increased by one bit for every factor of four in the oversampling rate.
30 Moreover, it is highly likely that in the case of the proposed correlation t~chn;que it might not be necessary to construct the n-bit data word. If that is the case, performing the correlations on a one-bit data stream is trivial. For the case of one-bit data, and signals with 35 a zero mean (such as with magnetic medium noise), the 094/~8 Sg~l PCT~S94/03917 correlation coefficient r is given by the following ex-pression -~ = ~ XL Y~
i=0 Therefore, this correlation analysis can be performed by using a simple up/down counter 132 as shown in Figure 17.
As shown therein, the read head 100 has its output ampli-fied by a preamplifier 116 which feeds a high sampling 10 rate analog comparator 134 for the xl input into logic gate 136. The original fingerprint data, in a one-bit data stream, is input in synchronism such that the logic gate 136 output goes positive when the bits match and negative when they don't. The output of the up/down 15 counter 132 is compared against some threshold which, if in ~xc~cs of the threshold indicates a match with the original fingerprint.
Just as it is possible to take a one-bit data stream at a high rate and construct an n-bit word at a 20 lower rate, the inverse is also feasible. In other words, at the time of creation, the magnetic medium to be fingerprinted could be sampled with a four-bit A to D
converter, for example. The fingerprint data can then be stored in four-bit words. At the point of authentica-25 tion, the four-bit data samples could be translated into a one-bit data stream at the higher rate, such as is implemented in recent model CD players. This four-bit data stream is then correlated with the one-bit data stream emanating from the analog comparator 134.
Still another methodology to ensure the capture of data samples correspon~;ng to those used during the orig-inal fingerprinting process involves, essentially, mea-suring the velocity of the credit card and adjusting the sampling rate to match that velocity. More exactly, two 35 transitions or other fiduciary mark may be placed on the W094/~8 ~ ~l"~ PCT~S94/03917 -~S9~

card which are a fixed distance D apart. The time it takes for the card to be pulled from the first transition to the second transition defines the velocity that the card is being pulled through the reader. The sampling 5 rate may then be ad;usted to match that velocity which relies on the assumption that the velocity that the card will travel for the second interval will match that of the first interval. As these dist~nce~ are only a few hundred microns, this assumption is reliable. This tech-10 nique does not require any oversampling which thereforemi ni ri ~es the amount of memory required to store the data points collected during the verification or card read step. In implementing this approach, a phase locked loop may be utilized where the input frequency is divided by a 15 factor M and the feedback loop is divided by a factor N
such that N over M times F,~fe~ equals FL ( sampling fre-quency). The factors M and N may be chosen dep~n~ing upon the other parameters of the system, such as the sample size, expected sampling frequency, distance D,0 etc. As noted, this t~hn~ue provides the advantage of ting oversampling which reduces the required memo-ry. It does suffer from a disadvantage in that a veloci-ty measurement must be made over a very small physical region, several hundred microns, which may present accu-5 racy problems. Also, this does increase the amount of magnetic stripe which must be dedicated to the finger-printing t~hn~que of the present invention although even with this doubling in length, still a minute fraction of the magnetic stripe of a credit card is being used.
As mentioned previously, credit card readers which are motorized, or which otherwise standardize the veloci-ty at which the card is pulled through the reader could be used to ~ nAte this sampling and mat~hing problem.
Furthermore, it is possible to combine several of the 35 methodologies to thereby form a hybrid methodology which might provide the best results. For example, the credit 094/2~8 ~1 ~ PCT~S94/03917 card reader might be designed to match the sampling in-terval which, as explAine~ above, ideally el~m~nAtes the requirement for oversampling. However, oversampling may also be used in conjunction with interval matching to 5 ensure that variations may be corrected for.
As shown in Figure 6, a computer 58 has a floppy disk drive 60 for reading loppy diskettes 62, all as is well known in the art. As still another implementation of the inventors' device, the software recorded on floppy lO diskette 62 may first require that the floppy disk drive 60 read a designated region of the magnetic medium com-prising floppy diskette 62 to determine its fingerprint, compare that fingerprint with the fingerprint stored in the program resident in floppy diskette 62, and if they 15 match permit computer 58 to run the application program stored on floppy diskette 62. If the program stored on floppy diskette 62 is not the original floppy diskette, then the application program will not be run as the mea-sured fingerprint will not match the fingerprint stored 20 in data cont~ine~ in the floppy diskette 62. In a varia-tion of this implementation, a single, archival, copy of floppy diskette 62 may be permitted should the program resident in floppy diskette 62 allow it to run if the measured fingerprint is either a match with the finger-25 print stored, or is its first non-match. The non-match fingerprint would then also be stored in the application software resident in floppy diskette 62 such that the program would recognize the original floppy diskette 62 and a ~-on~ floppy diskette as used by a user to create 30 this backup, archival, copy.
In still another implementation of the present - invention, a magneto-optic disc player 6~ has a disc 66 placed in its tray 68 and ready for play upon retraction of tray 68. However, disc 66 may have its fingerprint 35 stored to match the disc medium. Should magneto-optic disc player 64 have the appropriate circuitry for precon-f W094/2~8 ~ PCT~S94tO3917 ~
2~s98a~

ditioning play of disc 66 with a comparison of the mea-sured and recorded fingerprints, unauthorized copying of disc 66 may be prevented. Similarly, any taped copies made from disc 66 would necessarily have the incorrect 5 fingerprint stored therein and its subsequent play would also be prohibited should the tape playback unit have an a~L~liate circuit for preconditioning play based on mat~h;ng stored and measured fingerprints.
As shown in Figure 8, still another implementation lO of the invention includes measuring and storing the ap-propriate fingerprint on a cassette or digital tape 70, with a cassette tape player 72 having the necessary cir-cuitry for measuring and comparing the recorded and mea-sured fingerprints to pre-condition play, as explained l5 above.
In still another implementation of the present invention, as shown in Figure 9 a VCR 74 has a VCR tape 76 ready for insertion therein. Using the present inven-tion, a fingerprint can readily be measured and encoded 20 onto the VCR tape for comparison by appropriate circuitry cont~neA within VCR 74. Thus, if appropriately config-ured, the VCR would not play back a tape unless it was authentic or original.
There are various changes and modifications which 25 may be made to the invention as would be apparent to those skilled in the art. However, these changes or modifications are included in the te~ch;ng of the disclo-sure, and it is intended that the invention be limited only by the scope of the claims appended hereto.

Claims

What Is Claimed Is:

1. A device for fingerprinting an object for the later determination of its identity, said object includ-ing at least in part, a magnetic medium portion, said device having means for determining a remanent noise for said medium portion, said remanent noise comprising the fingerprint for said object, and means for recording said remanent noise for later comparison with the later deter-mination of said remanent noise to thereby identify said object.

2. The device of Claim 1 wherein said recording means has means for recording said fingerprint on said object.

3. The device of Claim 2 wherein said recording means includes means for recording said remanent noise in said magnetic medium portion in machine readable form.

4. The device of Claim 3 further comprising means for encrypting said remanent noise prior to its being recorded by said recording means.

5. The device of Claim 4 wherein said object is a magnetic data card, and said magnetic medium portion is a magnetic stripe on said magnetic data card.

6. The device of Claim 5 wherein said remanent noise is encrypted with a trap door function.

7. The device of Claim 5 wherein said object is a computer diskette, and said magnetic medium portion com-prises a portion of said computer diskette.

8. The device of Claim 7 wherein said remanent noise is encrypted with a trap door function.

9. The device of Claim 5 wherein said object is a magneto-optic disc, and said magnetic medium portion comprises a portion of said magneto-optic disc.

10. The device of Claim 9 wherein said remanent noise is encrypted with a trap door function.

11. The device of Claim 5 wherein said object is a magnetic recording tape, and said magnetic medium portion comprises a portion of said magnetic recording tape.

12. The device of Claim 11 wherein said remanent noise is encrypted with a trap door function.

13. The device of Claim 2 wherein said magnetic medium portion comprises a label affixed to said object.

14. An object having its fingerprint recorded for the later verification of its identity, said object hav-ing a magnetic medium portion, said fingerprint compris-ing a remanent noise for said magnetic medium portion.

15. The object of Claim 14 wherein said fingerprint is recorded on said object.

16. The object of Claim 15 wherein said remanent noise is recorded in said magnetic medium portion in ma-chine readable form.

17. The object of Claim 16 wherein said remanent noise is encrypted prior to its being recorded on said object.

18. The object of Claim 14 wherein said object is a magnetic data card, and said magnetic medium portion is a magnetic stripe on said magnetic data card.

19. The object of Claim 18 wherein said remanent noise is encrypted with a trap door function.

20. The object of Claim 14 wherein said object is a computer diskette, and said magnetic medium portion com-prises a portion of said computer diskette.

21. The object of Claim 20 wherein said remanent noise is encrypted with a trap door function.

22. The object of Claim 14 wherein said object is a magneto-optic disc, and said magnetic medium portion comprises a portion of said magneto-optic disc.

23. The object of Claim 22 wherein said remanent noise is encrypted with a trap door function.

24. The object of Claim 14 wherein said object is a magnetic recording tape, and said magnetic medium portion comprises a portion of said magnetic recording tape.

25. The object of Claim 24 wherein said remanent noise is encrypted with a trap door function.

26. The object of Claim 14 wherein said magnetic medium portion comprises a label affixed to said object.

27. The device of Claim 1 wherein said determining means includes means for DC saturating a region of said magnetic medium and means for reading a residual voltage over said region, said residual voltage comprising said remanent noise.

28. The device of Claim 27 wherein said region comprises a linear portion extending between about 30 micrometers and about 4300 micrometers.

29. The device of Claim 27 wherein said determining means further comprises means for averaging a plurality of residual voltage readings, said average comprising said fingerprint.

30. The device of Claim 1 wherein said determining means comprises a conventional magnetic recording head.

31. A device for authenticating an object having a fingerprint recorded, said object having a magnetic medi-um portion, said fingerprint comprising a remanent noise for said medium portion, said device including means for reading said recorded fingerprint, means for translating said remanent noise from said fingerprint, means for determining a remanent noise directly from said magnetic medium portion, and means for comparing said translated remanent noise with said determined remanent noise to determine whether they match, thereby authenticating said object.

32. The device of Claim 31 wherein said object has its fingerprint recorded thereon.

33. The device of Claim 31 wherein said fingerprint is recorded in said magnetic medium portion in machine readable form.

34. The device of Claim 33 wherein said fingerprint is encrypted, and said translation means includes means for de-crypting said fingerprint.

35. The device of Claim 34 wherein said object is a magnetic data card, said magnetic medium portion is a magnetic stripe on said magnetic data card.

36. The device of Claim 35 wherein said remanent noise is encrypted with a trap door function.

37. The device of Claim 31 wherein said object is a computer diskette, and said magnetic medium portion com-prises a portion of said computer diskette.

38. The device of Claim 37 wherein said remanent noise is encrypted with a trap door function.

39. The device of Claim 31 wherein said object is a magneto-optic disc, and said magnetic medium portion comprises a portion of said magneto-optic disc.

40. The device of Claim 39 wherein said remanent noise is encrypted with a trap door function.

41. The device of Claim 31 wherein said object is a magnetic recording tape, and said magnetic medium portion comprises a portion of said magnetic recording tape.

42. The device of Claim 40 wherein said remanent noise is encrypted with a trap door function.

43. The device of Claim 31 wherein said magnetic medium portion comprises a label affixed to said object.

44. A method for authenticating an object, said object including at least in part a magnetic medium por-tion, said object having a fingerprint recorded, said fingerprint comprising a previously determined remanent noise for said medium portion, said method comprising the steps of:
determining a remanent noise directly from said medium portion, reading said fingerprint, and comparing said determined remanent noise with said fingerprint.

45. The method of Claim 44 wherein said fingerprint comprises a machine readable representation of said pre-viously determined remanent noise, and the step of read-ing includes the step of reading said fingerprint with a machine.

46. The method of Claim 45 wherein said fingerprint comprises an encrypted representation of said previously determined remanent noise, and further comprising the step of de-crypting said fingerprint.

47. A method for fingerprinting an object for the later determination of its identity, said object includ-ing at least in part a magnetic medium portion, said method comprising the steps of determining a remanent noise for said medium portion, said remanent noise com-prising said fingerprint, and recording said fingerprint on said object for the later comparison thereof with the later determination of said remanent noise.

48. The method of Claim 47 further comprising the step of encrypting said fingerprint prior to recording.

49. A device for authenticating an object having a fingerprint recorded, said object having a magnetic medium portion, said recorded fingerprint comprising a signal corresponding to a first read of a remanent noise in said magnetic medium portion, means for determining said fingerprint by a second read of said remanent noise in said magnetic medium portion, and means for comparing said recorded fingerprint with said determined fingerprint including means for compensating for differences between said first read and said second read to determine whether said fingerprints match, thereby authenticating said object.

50. The device of Claim 49 wherein said compensating means includes means for compensating for differences between a speed at which each of said first read and said second read is performed.

51. The device of Claim 50 wherein said speed compensation means includes means for adjusting a sampling rate at which said second read is performed.

52. The device of Claim 51 wherein said adjusting means includes means for measuring the speed of said second read and comparing it with the speed of said first read.

53. The device of Claim 50 wherein said speed compensation means includes means for oversampling said remanent noise during at least said second read.

54. The device of Claim 53 wherein said speed compensation means includes means for adjusting a rate of said oversampling.

55. The device of Claim 50 wherein each of said recorded and determined fingerprints comprises a digital sample of said remanent noise, and wherein saidspeed compensation means includes means for creating for said second read a digital sample approximately 100 times larger than the digital sample of said first read.

56. A method for authenticating an object, said object including at least in part a magnetic medium portion having a fingerprint, said fingerprint comprising a remanent noise for said magnetic medium portion previously determined by a firstread thereof, said method comprising the steps of:
determining a remanent noise directly from said magnetic medium portion by a second read thereof, comparing said determined remanent noise with said previously determined remanent noise, and compensating for differences between said first read and said second read to determine whether said fingerprints match, thereby authenticating said object.

57. The method of Claim 56 wherein the step of compensating further includes the step of compensating for differences between a speed at which each of said first read and said second read is performed.

58. The method of Claim 57 wherein the step of speed compensating further includes the step of adjusting a sampling rate at which said second read is performed.

59. The method of Claim 58 wherein the step of adjusting includes the steps of measuring the speed of said second read and comparing it with the speedof said first read.

60. The method of Claim 57 wherein the step of speed compensating further includes the step of oversampling said remanent noise during at least said second read.

61. The method of Claim 60 wherein the step of speed compensating further includes the step of adjusting a rate of oversampling.

62. The method of Claim 57 wherein each of said previously determined remanent noise and said determined remanent noise comprises a digital sample of said remanent noise, and wherein the step of speed compensating further comprises the step of creating for said second read a digital sample approximately 100 times larger than the digital sample of said first read.

63. An object having its fingerprint recorded for the later verification of its identity, said object having a magnetic medium portion, said fingerprint comprising a remanent noise for said magnetic medium portion determined by a first read of said magnetic medium portion, and a speed of said read being recorded in association with said fingerprint for later use in compensating for differences between a speed of a subsequent read and the speed of said first read.

64. The object of Claim 63 further comprising a plurality of spaced timing marks on said magnetic medium portion so that a speed of said magnetic medium portion may be determined during a read.

65. A security card with a fingerprint for limiting access to a controlled access environment, said security card having a magnetic medium portion, said fingerprint comprising a remanent noise for said magnetic medium portion, said security card being validated to permit access only upon a match of a subsequentdetermination of said fingerprint with a previously determined fingerprint.

66. The security card of Claim 65 further comprising a data bit recorded on said magnetic medium portion, a value of said data bit corresponding to a placement of said fingerprint on said magnetic medium portion.

67. The security card of Claim 66 wherein said fingerprint is randomly located on said magnetic medium portion.

68. An object having its fingerprint recorded for the later verification of its identity, said object having a magnetic medium portion, said fingerprint comprising a remanent noise for said magnetic medium portion, and at least one data bit recorded on said magnetic medi-um portion, a value of said data bit corresponding to a location of said fingerprint.

69. The object of Claim 68 wherein said fingerprint is recorded on said magnetic medium portion, said data bit thereby pointing to a location on said magnetic medium portion.

70. The object of Claim 69 wherein said fingerprint is randomly located on said magnetic medium portion.