An evaluation of current and future botnet defenses

Botnets are large networks of compromised machines under the control of a bot master. These botnets constantly evolve their defences to allow the continuation of their malicious activities. The constant development of new botnet mitigation strategies and their subsequent defensive countermeasures has lead to a technological arms race, one which the bot masters have significant incentives to win. This dissertation analyzes the current and future states of the botnet arms race by introducing a taxonomy of botnet defences and a simulation framework for evaluating botnet techniques. The taxonomy covers current botnet techniques and highlights possible future techniques for further analysis under the simulation framework. This framework allows the evaluation of the effect techniques such as reputation systems and proof of work schemes have on the resources required to disable a peer-to-peer botnet. Given the increase in the resources required, our results suggest that the prospects of eliminating the botnet threat are limited.