Dois Pesos, Duas Medidas: Gerenciamento de Identidades Orientado a Desafios Adaptativos para Contenção de Sybils

Gustavo Huff MauchFlávio Roberto SantosWeverton Luis da Costa CordeiroMarinho Pilla BarcellosLuciano Paschoal Gaspary

The Sybil attack consists on the indiscriminate creation of counterfeitidentities by a malicious user (attacker). An effective approach to tackle suchattack consists of establishing computational puzzles to be solved prior to grantingnew identities. Despite its potentialities, solutions based on such approachdo not distinguish between identity requests from correct users and attackers,and thus require both to afford the same cost per identity requested. To tacklethis problem, in this paper we propose the use of adaptive computational puzzlesto limit the spread of Sybils. We estimate a trust score of the source of identityrequests in regard to the behavior of others. The higher the frequency a sourcerequests identities, the lower its trust score and, consequently, the higher thecomplexity of the puzzle to be solved by the user(s) associated to that source. Resultsachieved by means of an experimental evaluation evidence our solution'sability to establish more complex puzzles to potential attackers, while minimallypenalizing legitimate users.

Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato:
     Mantida por: