Mixnet optimization methods

We propose a method to optimally select mix network parameters for given deployment context and adversarial model. Our considers both worstcase average-case anonymity selects configurations that meet worst-case constraints while maximizing average anonymity. apply our methods mixnet size optimization determine the number width of layers, provide results various scenarios. For cases where suddenly changes (drop in user traffic) we evaluate countermeasures based on mix-generated dummy traffic show inexpensive link dummies can significantly boost protection some these cases.