<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Archiving and Interchange DTD v1.0 20120330//EN" "JATS-archivearticle1.dtd">
<article xmlns:xlink="http://www.w3.org/1999/xlink">
  <front>
    <journal-meta />
    <article-meta>
      <contrib-group>
        <contrib contrib-type="author">
          <string-name>Information Security Department</string-name>
          <email>gm@cnpo.ru</email>
          <xref ref-type="aff" rid="aff0">0</xref>
        </contrib>
        <contrib contrib-type="author">
          <string-name>Department of Computing and IS</string-name>
          <email>mrvo788@gmail.com</email>
          <xref ref-type="aff" rid="aff1">1</xref>
        </contrib>
        <aff id="aff0">
          <label>0</label>
          <institution>Bauman Moscow State Technical University</institution>
          ,
          <addr-line>Moscow</addr-line>
          ,
          <country country="RU">Russia</country>
        </aff>
        <aff id="aff1">
          <label>1</label>
          <institution>University of Greenwich</institution>
          ,
          <addr-line>London</addr-line>
          ,
          <country country="UK">Great Britain</country>
        </aff>
      </contrib-group>
      <fpage>79</fpage>
      <lpage>82</lpage>
      <abstract>
        <p>-As the first protection line for a computer system, the authentication system is of critical importance in the information security area. Despite the steady development of information security mechanisms, the password is the most commonly used authentication tool. The key vulnerability of such a protection mechanism is selecting an insecure password. The period of 2014-2017 saw major Internet companies suffer a number of password database leaks, which followed by a study of real password security. It should be noted that password system protection has not advanced much over the past years; mainly, there has been a clear tendency for imposing stricter requirements on the password entry interface. This being the case, there is still a question, yet to be answered, which passwords can be considered secure and which cannot. The work offers examples of password system assessment and reviews leaked passwords for their security under the current requirements. Security was verified using metrics (password security indices). These metrics provided the basis for defining objective requirements for password system security.</p>
      </abstract>
      <kwd-group>
        <kwd>identification</kwd>
        <kwd>authentication</kwd>
        <kwd>password metric</kwd>
        <kwd>information security</kwd>
        <kwd>information protection</kwd>
        <kwd>password system</kwd>
      </kwd-group>
    </article-meta>
  </front>
  <body>
    <sec id="sec-1">
      <title>INTRODUCTION</title>
      <p>Even though the problems of password system security are
subject to permanent studies, this issue remains yet to be dealt
with in practice for subjective reasons. For example, some
software system developers treat the password system security
issue in different ways, while users often fail to meet in full the
authentication system security policy.</p>
      <p>Over the last two years, there have been a number of major
leaks from password databases of communication Internet
services (Twitter, Hotmail, Yandex, Google, Mail, Dropbox
etc.). This allowed password security to be investigated using
informal and formal indices.</p>
      <p>II.</p>
      <p>PASSWORD SECURITY STRENGTH CONCEPT
Let’s consider the password mining probability formula:
 =
 ∙ 
| |</p>
      <p>The above makes it possible to formulate simple substantive
criteria whereby a password is considered secure:









</p>
      <p>Password length must be at least 8 characters;
Characters of various cases must be taken into
account;</p>
    </sec>
    <sec id="sec-2">
      <title>Numeral must be used;</title>
    </sec>
    <sec id="sec-3">
      <title>Special characters must be used; Password must not be based on a word; Password must not include words relating to the password owner.</title>
      <p>
        Meanwhile, there is a discussion underway in literature as
to formal requirements for password systems [
        <xref ref-type="bibr" rid="ref1 ref10 ref11 ref2 ref3 ref4 ref5 ref7 ref8 ref9">1-5, 7-11</xref>
        ].
      </p>
      <p>III.</p>
    </sec>
    <sec id="sec-4">
      <title>REGULATORY REQUIREMENTS</title>
      <p>As is well known, the password requirements were set out
in the Orange Book and were actually reduced to password
length (6 and 8 characters depending on the system protection
class). Current documents define the availability, as a
minimum, of a password protection policy at an organization or
in a computer system. Some requirements can be made more
specific, while others are left in the hands of system
administrators (Table 1).</p>
      <p>IV.</p>
    </sec>
    <sec id="sec-5">
      <title>PASSWORD SECURITY METRICS</title>
      <p>Listed below are some of the best known password security
index classes:</p>
      <p>Numerical metrics (e.g. Orange Book);</p>
    </sec>
    <sec id="sec-6">
      <title>Probabilistic metrics [2, 7, 8]; Shannon informational entropy [6]; Heuristic entropy modifications [3, 10, 11]; Probabilistic entropy modifications [9].</title>
      <p>NIST-recommended password entropy can be calculated by
statistics for specific systems, which is not always practicable.</p>
      <p>This work will consider the Shannon entropy and heuristic
entropy (recommended by NIST SP 800-22). The methods
differ in that the Shannon entropy assumes a password to be
generated by a random-number generator, while heuristic
entropy implies a human created password.</p>
      <p>The Shannon entropy is calculated as follows:
 = log2| | =  ∙ log2| | = 
ln| |
ln 2
,
where |A| is alphabet capacity, n is password length.</p>
      <p>The metric suggests that the more complicated the alphabet
and the longer the password, the more secure the latter is.</p>
      <p>Given below is the Shannon entropy calculation example
(Table 2).
of a password containing non-alphanumeric or uppercase
characters.</p>
      <p>This formula can be described as follows: the first password
character receives a value of 4 bits, each subsequent character
from the second to the eighth one receives 2 bits, from the ninth
to the twentieth – 1.5 bits and each subsequent – 1 bit. If there
are non-alphanumeric or uppercase characters, 6 bits are added
to the obtained result.</p>
      <p>
        For these metrics, a password is considered secure if it
conforms to the entropy [
        <xref ref-type="bibr" rid="ref11">11</xref>
        ]:
- according to Shannon – 56 bits or more;
- on NIST recommendations – 24 bits or more.
      </p>
      <p>The above criteria should be restricted, i.e. if a password is
recorded in password brute forcing databases (dictionaries),
entropy is reduced to zero.
last year. Presented below is certain statistics obtained for each
password</p>
      <p>database. The outcomes of the study into the
compromised
password
database
of</p>
    </sec>
    <sec id="sec-7">
      <title>Yandex (1,261,809 passwords) are listed in tables 3-5, Mail.ru (45,000) – tables 68, Google (4,926,673) – tables 9-11.</title>
      <p>Used alphabet
PWD including numerals only
PWD composed of characters
PWD composed of lowercase letters only
PWD similar to mobile phone number
PWD coinciding with login
PWD similar to dates
PWD suitable for secure password
informal description
PWD suitable as per Shannon security
PWD suitable as per NIST security
Number of
passwords
774,669
1,968,873
1,968,873
22,751
45,010
156,142
0</p>
      <p>
        Comparative analysis of the obtained and earlier known
statistics [
        <xref ref-type="bibr" rid="ref12 ref13">12, 13</xref>
        ] showed a trend for slight strengthening of
password protection. This is because some Internet services
defined stricter rules for interfaces, for example, strengthened
requirements for password length (at least 6 characters) and the
use of a relatively complicated alphabet. The statistics suggests,
however, that the above fact does not stop unorganized and
careless users from choosing easily hackable passwords, and
the number of Top 500 passwords has hardly changed over the
years.
      </p>
      <p>In general, the study confirmed that the authentication
system remains highly vulnerable (only 10% of passwords can
be considered reliable), which prompts the creation of
integrated information protection systems and the improvement
of information security management systems.</p>
      <p>Finally, it should be noted that using entropic metrics
instead of verbal descriptions is more practical in defining
technical requirements for information security systems, as they
are easier to automate and control. Besides, the use of formal
indices helps diminish the degree of subjectivity inherent in
system security analysis.</p>
    </sec>
  </body>
  <back>
    <ref-list>
      <ref id="ref1">
        <mixed-citation>
          [1]
          <string-name>
            <surname>Bonneau</surname>
            <given-names>J.</given-names>
          </string-name>
          <article-title>Guessing human-chosen secrets</article-title>
          .
          <source>Technical Report UCAMCL-TR-819</source>
          .
          <year>2012</year>
          . 161 p.
        </mixed-citation>
      </ref>
      <ref id="ref2">
        <mixed-citation>
          [2]
          <string-name>
            <surname>Bonneau</surname>
            <given-names>J.</given-names>
          </string-name>
          <article-title>The Science of Guessing: Analyzing an Anonymized Corpus of 70 Million Passwords</article-title>
          .
          <source>In 2012 IEEE Symposium on Security and Privacy</source>
          , IEEE,
          <year>2012</year>
          , pp.
          <fpage>538</fpage>
          -
          <lpage>552</lpage>
          . DOI:
          <volume>10</volume>
          .1109/SP.
          <year>2012</year>
          .
          <volume>49</volume>
          .
        </mixed-citation>
      </ref>
      <ref id="ref3">
        <mixed-citation>
          [3]
          <string-name>
            <surname>Boothroyd</surname>
            <given-names>V.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Chiasson</surname>
            <given-names>S.</given-names>
          </string-name>
          <article-title>Writing down your password: Does it help?</article-title>
          <source>In Proc. of the 2013 Eleventh Annual Conference on Privacy, Security and Trust</source>
          , IEEE,
          <year>2013</year>
          , pp.
          <fpage>267</fpage>
          -
          <lpage>274</lpage>
          . DOI:
          <volume>10</volume>
          .1109/PST.
          <year>2013</year>
          .
          <volume>6596062</volume>
          .
        </mixed-citation>
      </ref>
      <ref id="ref4">
        <mixed-citation>
          [4]
          <string-name>
            <surname>Burnett</surname>
            <given-names>M.</given-names>
          </string-name>
          <string-name>
            <surname>Perfect Password</surname>
          </string-name>
          : Selection, Protection, Authentication. Syngress Publishing,
          <year>2006</year>
          .194 p.
        </mixed-citation>
      </ref>
      <ref id="ref5">
        <mixed-citation>
          [5]
          <string-name>
            <surname>Burr</surname>
            <given-names>W.E.</given-names>
          </string-name>
          <article-title>and etc</article-title>
          .
          <source>Electronic Authentication Guideline. NIST Special Publication 800-63-1</source>
          .
          <year>2011</year>
          . 110 p.
        </mixed-citation>
      </ref>
      <ref id="ref6">
        <mixed-citation>
          [6]
          <string-name>
            <surname>Christiansen</surname>
            <given-names>M.M.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Duffy</surname>
            <given-names>K.R.</given-names>
          </string-name>
          <string-name>
            <surname>Guesswork</surname>
          </string-name>
          , Large Deviations, and Shannon Entropy,
          <source>IEEE Transactions on Information Theory</source>
          ,
          <year>2013</year>
          , volume
          <volume>59</volume>
          , issue 2, pp.
          <fpage>796</fpage>
          -
          <lpage>802</lpage>
          DOI: 10.1109/TIT.
          <year>2012</year>
          .
          <volume>2219036</volume>
          .
        </mixed-citation>
      </ref>
      <ref id="ref7">
        <mixed-citation>
          [7]
          <string-name>
            <surname>Galbally</surname>
            <given-names>J.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Coisel</surname>
            <given-names>I.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Sanchez</surname>
            <given-names>I.</given-names>
          </string-name>
          <article-title>A probabilistic framework for improved password strength metrics</article-title>
          .
          <source>In Proc. of the 2014 International Carnahan Conference on Security Technology (ICCST)</source>
          , IEEE,
          <year>2014</year>
          , pp.
          <fpage>1</fpage>
          -
          <lpage>6</lpage>
          . DOI:
          <volume>10</volume>
          .1109/CCST.
          <year>2014</year>
          .
          <volume>6986985</volume>
          .
        </mixed-citation>
      </ref>
      <ref id="ref8">
        <mixed-citation>
          [8]
          <string-name>
            <surname>Galbally</surname>
            <given-names>J.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Coisel</surname>
            <given-names>I.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Sanchez</surname>
            <given-names>I.</given-names>
          </string-name>
          <article-title>A New Multimodal Approach for Password Strength Estimation-Part I: Theory and Algorithms</article-title>
          .
          <source>IEEE Transactions on Information Forensics and Security</source>
          ,
          <year>2017</year>
          , volume
          <volume>12</volume>
          , issue 12, pp.
          <fpage>2829</fpage>
          -
          <lpage>2844</lpage>
          . DOI: DOI: 10.1109/TIFS.
          <year>2016</year>
          .
          <volume>2636092</volume>
          .
        </mixed-citation>
      </ref>
      <ref id="ref9">
        <mixed-citation>
          [9]
          <string-name>
            <surname>Groza</surname>
            <given-names>B.</given-names>
          </string-name>
          <article-title>Analysis of a Password Strengthening Technique and Its Practical Use</article-title>
          .
          <source>In Proc. of 2009 Third International Conference on Emerging Security Information, Systems and Technologies</source>
          , IEEE,
          <year>2009</year>
          , pp.
          <fpage>292</fpage>
          -
          <lpage>297</lpage>
          . DOI:
          <volume>10</volume>
          .1109/SECURWARE.
          <year>2009</year>
          .
          <volume>52</volume>
          .
        </mixed-citation>
      </ref>
      <ref id="ref10">
        <mixed-citation>
          [10]
          <string-name>
            <surname>Kelley</surname>
            <given-names>P.G.</given-names>
          </string-name>
          <article-title>and etc. Guess Again (and Again and Again): Measuring Password Strength by Simulating Password-Cracking Algorithms</article-title>
          .
          <source>In proc. of the 2012 IEEE Symposium on Security and Privacy</source>
          , IEEE,
          <year>2012</year>
          , pp.
          <fpage>523</fpage>
          -
          <lpage>537</lpage>
          . DOI:
          <volume>10</volume>
          .1109/SP.
          <year>2012</year>
          .
          <volume>38</volume>
          .
        </mixed-citation>
      </ref>
      <ref id="ref11">
        <mixed-citation>
          [11]
          <string-name>
            <surname>Markov</surname>
            <given-names>G.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Sharunov</surname>
            <given-names>V</given-names>
          </string-name>
          .
          <source>About Information Security of Email Services. Voprosy kiberbezopasnosti [Cybersecurity issues]</source>
          ,
          <year>2015</year>
          , No
          <volume>5</volume>
          (
          <issue>13</issue>
          ), pp.
          <fpage>55</fpage>
          -
          <lpage>59</lpage>
          . DOI:
          <volume>10</volume>
          .21681/
          <fpage>2311</fpage>
          -3456-2015-5-
          <fpage>55</fpage>
          -59.
        </mixed-citation>
      </ref>
      <ref id="ref12">
        <mixed-citation>
          <source>[12] The Top 500 Worst Passwords of All Time</source>
          ,
          <year>2008</year>
          . URL: http://www.whatsmypass.com/the-top-500
          <string-name>
            <surname>-</surname>
          </string-name>
          worst
          <article-title>-passwords-of-all-time. (Last access</article-title>
          :
          <volume>01</volume>
          /12/
          <year>2017</year>
          ).
        </mixed-citation>
      </ref>
      <ref id="ref13">
        <mixed-citation>
          <source>[13] The Top 500 Worst Passwords of All Time</source>
          .
          <year>2010</year>
          . URL: https://www.symantec.com/connect/blogs/top-500
          <string-name>
            <surname>-</surname>
          </string-name>
          worst-passwords-alltime. (Last access:
          <volume>01</volume>
          /12/
          <year>2017</year>
          ).
        </mixed-citation>
      </ref>
    </ref-list>
  </back>
</article>