Joint Architecture Standard Overview Profile
The selection of network topology is a critical component when developing multi-node or multi-point system architectures. A good choice of topology will require less power, have less complexity, higher reliability, and will orchestrate network traffic smoothly and quickly between nodes. Conversely, a poor choice of topology will introduce complex and power-hungry logic, reduced performance due to bottlenecks and large distances between nodes, and increase the probability of system failure due to dropped messages and lack of fault tolerance.
Topology selection should begin by identifying the required and desired qualities of the network. These qualities may include a certain level of speed and performance, power usage, wiring or routing complexity, cost, redundancy and reliability, or a combination of these or other factors. Furthermore, individual system requirements may introduce data flow requirements that map to some topologies more readily than others. Only after the requirements of the network have been established can the various topologies be analyzed to determine which will work best for a particular system.
During topology selection, it is important to remember that often there will not be a single ideal topology for most systems. Rather, selecting a topology becomes a matter of balancing and trading off various properties until a reasonable solution is achieved. Another issue to consider is the fact that several topologies may be acceptable given a set of requirements. In this case, selecting among the potential candidates becomes a judgment call on the part of the designer.