Page 233 - DCAP311_DCAP607_WIRELESS_NETWORKS
P. 233

Unit 14: Authentication




               Client-generated  one-time  passwords  (figure  14.3)  are  similar  to  conventional  OTP   Notes
               hardware tokens (such as RSA SecurID, VeriSign VIP OTP, and so on). However, with
               the level of near-ubiquitous proliferation of mobile devices today, cellular phones have
               become a viable alternative as the soft-token (or "something held") authentication factor.
               In this case, individualized cryptographic software components can be installed on mobile
               devices to generate time-based or event-based OTPs. Users then can use the OTP to sign-
               in to the application after authenticating simple Web-based credentials (examples include
               RSA SecurID software, Java ME applications, and so on). This approach has the benefits
               of OTPs, not having to deal with out-of-band delivery latencies and inconsistent service
               coverage.

                                  Figure 14.4: Out-of-band Authentication



















          Source: http://msdn.microsoft.com/en-us/library/cc838351.aspx#_Architectural_Perspectives
               Out-of-band  (OOB)  (figure  14.4)  authentication  leverages  the  second  factor  for
               authentication, instead of delivery of individualized information. Current implementations
               include speech recognition that is facilitated via out-bound or in-bound voice calls (to/
               from land or cellular lines) or KBA via SMS request/reply. This type of solution offers
               higher authentication strength, as both the browser and a second factor are used to verify
               credentials, which works to impede common phishing attacks.
               In general, these high-level approaches rank in increasing relative authentication strength.
               However, higher authentication strength does not necessarily represent the best-of-breed
               solution, as security often requires trade-offs in user convenience. Studies indicate that
               different user populations respond  differently to strong-authentication methods.  The
               choice of an approach (or combinations of approaches) should consider user segments, as
               well as the types of activities and interactions that they perform with the application.


                       Example: Online consumers who view account balances (considered a relatively
               low-risk type of activity) should require comparatively less effort to access than consumers
               who conduct account transfers or physicians who view patient's medical records.

               Finally, the implementation of strong user authentication often is a balancing act between
               security and usability. Many implementations have not achieved their intended effectiveness
               by either delivering significantly deteriorated user experiences or compromising security
               postures by simply satisfying requirements on paper. Furthermore, there is no one-size-
               fits-all solution approach. For example, not all Web sites require multifactor authentication,
               while enhanced knowledge-based authentication might be "secure enough"; and some
               carefully designed single-factor methods actually might be "stronger" than some multifactor
               methods. A well-balanced design tends to be more cost-effective than force-fitting various
               best-of-breed approaches into one solution.






                                           LOVELY PROFESSIONAL UNIVERSITY                                   227
   228   229   230   231   232   233   234   235   236   237   238