The architecture for the RDRN link management and control is shown in Figure 14. The topology modules are used only on ES nodes capable of becoming a master ES. The remaining modules are used on all ES nodes and RNs. The beamform module determines an optimal steering angle for the given number of beams which connects all RNs to be associated with this ES. It computes an estimated signal to noise interference ratio (SIR) and generates a table of complex weights which, once loaded, will control the beam formation. Note that this table is not loaded until the table fill trigger is activated. The connection table is used by the Adaptive HDLC and ATM protocol stacks for configuration via the adaptation manager.
Figure 14: Network Control System Architecture.
The emulation uses as much of the actual network control code as possible. The packet radio driver, GPS driver, and Network Control Protocol state machine are implemented in Maisie; tables, data structures, and decision functions from working NCP code are used. Figure 15 shows the structure of the Maisie entities. The entity names are shown in the boxes and the message types are shown along the lines. Direct communication between entities is represented as a solid line. The dashed lines indicate from where entities are spawned.
Figure 15: Emulation Design.
The RN entity which performs ATM VC setup (HSLRN entity in Figure 15) generates calls as a Poisson process which the ES node (HSLES entity in Figure 15) will attempt to accept. If the EN moves out of range or the ES has no beam or slot available the setup will be aborted. As the RN moves, the ES will hand off the connection to the proper ES based on closest distance between RN and ES.