Safeguarding web services using self-adaptive schema hardening algorithm

Abstract

Web Services in production often evolve over time due to changes in business and security requirements. Often various Web Service standards such as WS-Security, WS-Trust, WS-Routing etc. are introduced or revoked. Such changes alter the structure of an input message accepted by web services. Message validation mechanism becomes in-effective if schemas in use are not updated in line with aforementioned changes. Also, Web Services become prone to different attack vectors if the schemas are loosely defined. Here, we present algorithms that help fine tune schemas by the process of iterative deduction. Also, our work helps to identify patterns of attack vectors that demarcate themselves from genuine messages. Our adaptive schema refining algorithm classifies logged requests into set of schema classes based on a measure of similarity. This classification of messages in to schema classes enables us to tighten the schemas to prevent bad requests or expand the schemas to accommodate newer requests. © 2011 Springer-Verlag.