On security-aware transmission scheduling

The problem of interest is to characterize to what extent nodes independently following certain transmission schedules can be hijacked to relay flows of information packets. Information flows can be embedded in given transmission schedules by properly adding delays and inserting dummy packets. Such hidden flows are usually indicators of network intrusion, and it is of interest to know their rates. The maximum rate of information flow that can be transmitted without causing the transmission activities to deviate from given transmission schedules is used to measure the covert capacity under these schedules. Based on the assumption that information flows have bounded delays, a theoretical framework is constructed to quantitively analyze the covert capacity under transmission schedules modeled by renewal processes. Explicit solution is obtained for Poisson processes. The results suggest a close correlation between the covert capacity and the traffic burstiness.