The MPSAS has two distinctly separate modes of operation. The first mode is the simplest and will likely be the most often used mode, and it concerns the ability to simulate the processing of any one MPS algorithm on a single AP. The direct inputs to the AP/algorithm are configured by the user and then run through a simulation. If a non-supervisor AP is being simulated, these direct inputs correspond to the MPS devices within the algorithm to be simulated. However if a supervisor is being simulated, these direct inputs correspond to MPST stopper words (which are used to determine the stopper configuration) and the requested rates of subordinate APs.
The second mode of operation, simulation of a supervisor at the device-level, is a bit more complex and represents the complete software simulation of the MPS system. Rather than simulating only a single algorithm, this MPSAS mode simultaneously simulates all APs and their supervisor AP in an integrated way representative of the entire MPS system. The devices to all relevant non-supervisor APs are configured and each AP is simulated in parallel. The output of these simulations is then used as input to a simulation of the supervisor AP/algorithm. Results are then provided for each algorithm simulation performed as a branch in this ``tree'' of distributed algorithm processing.
The following two sections provide details on how to go about using MPSAS to simulate in either mode. By design, the simulation of a supervisor at the device-level is more or less made up of many simulations of single APs. Hence the section on this mode simply builds upon basics discussed in the section on simulating a single AP/algorithm. Therefore we begin with a discussion of simulating a single algorithm (and its AP).