Enhancing the trustworthiness of service on-demand systems via smart vote filtering

Abstract

Service on-demand (SoD) systems allow their users to regulate the sharing of common resources via a voting process. A common application example is the collaborative scheduling of multimedia transmissions in e-radio or video streaming services. Therefore, high user commitment and participation is critical to the success of a SoD system. Securing a SoD system against common attacks, such as vote flooding, can impose client anonymity retraction, online registering and access control mechanisms. Nonetheless, such processes can degrade the users’ quality of experience, discouraging user participation. The present study proposes a defense mechanism against vote flooding attacks that can operate under complete vote anonymity and without any user access restrictions. The novel scheme is implemented as a vote filtering scheme, executed prior to each service scheduling decision. The proposed scheme has linear complexity and is shown via simulations to considerably mitigate or completely negate the effects of several attacks types.