LESS Is More: Host-Agent Based Simulator for Large-Scale Evaluation of Security Systems

摘要: Recently proposed network security systems have demonstrated the benefits of scale for achieving many goals, including detection worm outbreaks, botnets, and denial service attacks. However, is also a barrier to further advancement such systems: obtaining working with appropriately large data sets difficult, existing simulation techniques are ill suited this domain. To overcome these challenges, we propose host behavior simulator, LESS, designed evaluating systems. LESS build automatically configures behaviors agents using background traffic samples malicious models. In turn, communicate each other throughout simulation, generating records. We demonstrate applicability by tuning it publicly available traces, then generated records reproduce results from several recently used extend evaluations systems, highlighting dimensions system performance that would be difficult study without simulation.