TDMA Scheduling and Channel Assignment for Cognitive Tactical Networks

Shortage of spectrum constrains military applications that need to support different classes of traffic. In this work, we study efficient utilization of limited spectrum and propose joint time division multiple access (TDMA) MAC scheduling and channel assignment. First, we formulate the problem as a linear program and propose a centralized algorithm that utilizes the concept of "independent link sets" for scheduling slots and allocating channels. Compared to previous work, our approach supports partial flow loss and allows prioritization of traffic flows. Next, we propose extensions to the distributed USAP MAC protocol (widely used in military applications) to jointly assign TDMA slots and channels to links. Our extensions allow USAP to minimize either slots or channels used in the schedule; a function that is needed for dynamic behavior of future SDR/cognitive radios. We implement our scheduling algorithms in the context of a network design tool (NEDAT) and evaluate their performance using realistic mission scenarios.