{"id":536,"date":"2009-01-29T22:42:29","date_gmt":"2009-01-30T06:42:29","guid":{"rendered":"http:\/\/cubist.cs.washington.edu\/Security\/?p=536"},"modified":"2009-01-29T22:52:06","modified_gmt":"2009-01-30T06:52:06","slug":"security-review-fingerprint-scanners","status":"publish","type":"post","link":"https:\/\/secblog.cs.washington.edu\/Security\/2009\/01\/29\/security-review-fingerprint-scanners\/","title":{"rendered":"Security Review: Fingerprint Scanners"},"content":{"rendered":"<p>Fingerprint identification is the oldest biometric method. Everybody has a set of unique fingerprints, formed by the ridges and valleys on the skin. Fingerprints have been used in many fields, such as crime scene investigation and criminal database to identify people. In terms of technology, biometrics are used as an authentication method, alone or in combination of other techniques, such as password or another biometric forms.<\/p>\n<p>A fingerprint scanner collects, prints, and creates images that can then be analyzed and compared to images already on record. Optical and capacitance are the two major scanners used on the market to collect and analyze fingerprints. An optical scanner works like a digital camera where it collects data on the light reflected off of one&#8217;s fingerprints. The lightness\/darkness of the reflection is created by ridges and valleys on the skin. A capacitance scanner also creates an image of one&#8217;s print; however, it uses electrical current and conductivity to mark the light\/darkness area of the print.<\/p>\n<p>After an image collection, a fingerprint scanner system compares specific features of the print (or minutiae), such as angle of certain ridge\/valley, location of circles, etc. To get a match, the system doesn&#8217;t need to find every pattern in both the sample and the record, it just needs enough matches, and threshold varies.<br \/>\n<!--more--><br \/>\nGOALS\/ASSETS<\/p>\n<ul>\n<li>\nThings you&#8217;re trying to protect with your fingerprint. When someone employs fingerprint biometrics as a form to protect\/restrict access, then whatever the biometrics is used on is an asset. Although scanners are quite cheap these days and come with most tech products (laptop), it is still not widely implemented.<\/li>\n<li>\nYour identity. Since your own fingerprints are unique, using it as an authentication method would mean that your unique identity is exposed and is at risk.<\/li>\n<li>\nThis is very similar to the previous point: the record of prints. Since sample prints are compared to record prints to provide authentication, then this database is very valuable as well. We do not want people to tamper with this record set, such as adding, deleting, or modifying existing prints, because then authorization is flawed and would fail (or falsely succeeded, depend on how you see it).<\/li>\n<\/ul>\n<p>ADVERSARIES AND THREATS<\/p>\n<ul>\n<li>\nPeople who&#8217;s trying to steal your the stuff you&#8217;re protecting with your fingerprints. If it&#8217;s that important to use biometrics as a mean of authentication, then someone will want to get a hold of it, whether it&#8217;s your embarrassing childhood pictures or national security documents.<\/li>\n<li>\nIdentity thieves. Electronic file that store fingerprints data are vulnerable to data theft, just like passwords stored in the database. The security of protected information might be compromised if fingerprints data is stolen. Fingerprint is &#8220;something you are&#8221; and uniquely yours. As biometrics become a norm in authentication and protection, it is more valuable to these folks to get a hold of your prints to be &#8220;you,&#8221; because it may grant them access to many more private data.<\/li>\n<li>\nAn additional threat is that of the non-visible mark that human fingers leave on the scanner that it touches. This can easily be brought visible and get stolen using some special tapes.<\/li>\n<\/ul>\n<p>WEAKNESSES<\/p>\n<ul>\n<li>To be considered as a match, only certain numbers of features need to be present in both the sample and record. This threshold could be arbitrary. Assuming that this threshold of matches could be changed, adversaries could alter it, either making it unreasonably high that even the prints&#8217; owner couldn&#8217;t get an authentication, or making it low so that everyone can be authenticated with any prints.<\/li>\n<li>\nInaccuracy of the scanners. Even though everybody has their unique set of fingerprints, but the replication (aka prints) could vary by angle, size, smear, etc. Assuming that the record prints are collected via the same medium (the same type of scanner) into the database, we have couple of areas where this inaccuracy could occur: a) during record collection, meaning when the prints enters the database the first time, b) during sample collection at the time when someone requests authentication with an fingerprint. Also, since only a certain amount of features needed to be matched, adversaries rely on the inaccuracy of scanner\/prints and this threshold to fake authentication.<\/li>\n<li>\nFingerprints cannot be replaced. Once a fingerprint is compromised, then there is no way to create another one (unless you count your other 19 digits). Other forms authentication allows changes and replacement once compromised. We only have 10 fingers (and 10 toes). After that, you&#8217;re kinda out of ideas.<\/li>\n<\/ul>\n<p>DEFENSES<\/p>\n<ul>\n<li>\nUse biometrics in combination with other types of authentication. Although fingerprints are unique, however, as we have explored earlier, it has plenty of weaknesses. To make up for that, we could employ defense in breadth by using it as part of an broader and more elaborate authentication scheme, such as password or secret question, or the more extreme DNA sample.<\/li>\n<li>\nSecure scanner setting. As mentioned above, assuming that the number of features that are needed for a match could change, then protecting such setting is also very important. Limit who can change it and how it could be change would be a great way to prevent unauthorized changes of setting. The setting itself needs to be protected.<\/li>\n<li>\nFingerprint scanners should be able to test if the finger is a real finger, not an imitation of human finger (eg. tape or mold)<br \/>\nHave some form of encryption of the fingerprints so people cannot be identified. The encryption result should be random and not form into other existing fingerprints.<\/li>\n<\/ul>\n<p>EVALUATION<\/p>\n<p>We have only touched on a small set of assets, weaknesses and adversaries that associated with fingerprint scanner and authentication. This technology is fairly new and has not yet been widely adopted by all authentication methods mainly because it&#8217;s still in a stage of trial and getting approval of method. What makes biometrics different than all other types of authentication is it is considered as &#8220;what we are.&#8221; To make it work, users must invest personal asset (in this case their unique finger print) into the system, and thus risk of exposing it to the world, if not properly secured. This put users in a place where protecting their prints becomes another item on the list since prints are left wherever users touch. To a more extreme case, users might lose a finger or two, or even a whole hand, if adversaries must need fingerprints to gain access. Also, unlike passwords, fingerprints cannot be replaced. This limits the extensibility of how long fingerprints could be used.<\/p>\n<p>CONCLUSION<\/p>\n<p>Fingerprint scanner is an excellent means of identification. Instead of memorizing passwords, people can use something unique to themselves to protect their assets. But these devices may increase vulnerability to security attacks, and once the fingerprint gets into the wrong hand, it is really hard to replace it. So for now, it is better if we don\u2019t have everything we own protected by a fingerprint scanner. Fingerprint scanners cannot be thought as password replacements. Rather, they can be considered as an additional security items that help improve security. We are hoping that as technology improves, fingerprint scanner will become more reliable and it can be an integral part of human\u2019s life.<\/p>\n<p>Xia (My) Cam &amp; Devy Pranowo<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Fingerprint identification is the oldest biometric method. Everybody has a set of unique fingerprints, formed by the ridges and valleys on the skin. Fingerprints have been used in many fields, such as crime scene investigation and criminal database to identify people. In terms of technology, biometrics are used as an authentication method, alone or in combination of other techniques, such as password or another biometric forms.<\/p>\n<p>A fingerprint scanner collects, prints, and creates images that can then be analyzed and compared to images already on record. Optical and capacitance are the two major scanners used on the market to collect and analyze fingerprints. An optical scanner works like a digital camera where it collects data on the light reflected off of one&#8217;s fingerprints. The lightness\/darkness of the reflection is created by ridges and valleys on the skin. A capacitance scanner also creates an image of one&#8217;s print; however, it uses electrical current and conductivity to mark the light\/darkness area of the print.<\/p>\n<p>After an image collection, a fingerprint scanner system compares specific features of the print (or minutiae), such as angle of certain ridge\/valley, location of circles, etc. To get a match, the system doesn&#8217;t need to find every pattern in both the sample and the record, it just needs enough matches, and threshold varies. <a href=\"https:\/\/secblog.cs.washington.edu\/Security\/2009\/01\/29\/security-review-fingerprint-scanners\/\">Continue reading <span class=\"meta-nav\">&rarr;<\/span><\/a><\/p>\n","protected":false},"author":82,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[5],"tags":[],"class_list":["post-536","post","type-post","status-publish","format-standard","hentry","category-security-reviews"],"_links":{"self":[{"href":"https:\/\/secblog.cs.washington.edu\/Security\/wp-json\/wp\/v2\/posts\/536","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/secblog.cs.washington.edu\/Security\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/secblog.cs.washington.edu\/Security\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/secblog.cs.washington.edu\/Security\/wp-json\/wp\/v2\/users\/82"}],"replies":[{"embeddable":true,"href":"https:\/\/secblog.cs.washington.edu\/Security\/wp-json\/wp\/v2\/comments?post=536"}],"version-history":[{"count":11,"href":"https:\/\/secblog.cs.washington.edu\/Security\/wp-json\/wp\/v2\/posts\/536\/revisions"}],"predecessor-version":[{"id":547,"href":"https:\/\/secblog.cs.washington.edu\/Security\/wp-json\/wp\/v2\/posts\/536\/revisions\/547"}],"wp:attachment":[{"href":"https:\/\/secblog.cs.washington.edu\/Security\/wp-json\/wp\/v2\/media?parent=536"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/secblog.cs.washington.edu\/Security\/wp-json\/wp\/v2\/categories?post=536"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/secblog.cs.washington.edu\/Security\/wp-json\/wp\/v2\/tags?post=536"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}