A Fuzzing-Based Test-Creation Approach for Evaluating Digital TV Receivers via Transport Streams

Digital TV (DTV) receivers are usually submitted to testing systems for conformity and robustness assessment, and their approval implies correct operation under a given DTV specification and protocol. However, many broadcasters inadvertently misconfigure their devices and transmit the wrong information concerning data structures and protocol format. Since most receivers were not designed to operate under such conditions, malfunction and incorrect behavior may be noticed, often recognized as field problems, thus compromising a given system’s operation. Moreover, the way those problems are usually introduced in DTV signals presents some randomness, but with known restrictions given by the underlying transport protocols used in DTV systems, which resembles fuzzing techniques. Indeed, everything may happen since any deviation can incur problems, depending on each specific implementation. This error scenario is addressed here, and a novel receiver robustness evaluation methodology based on non-compliance tests using grammarbased guided fuzzing is proposed. In particular, devices are submitted to unforeseen conditions and incorrect configuration. They are created with guided fuzzing based on real problems, protocol structure, and system architecture to provide resources for handling them, thus ensuring correct operation. Experiments using such a fuzzing scheme have shown its efficacy and provided opportunities to improve robustness regarding commercial DTV platforms.