Most stored program controlled telephone switching systems are designed with centralized control architectures. Evolving processor technology and requirements of smooth growth over a wide size range have led system designers to take advantage of distributed processing architectures. This paper provides a conceptual framework for software partitioning on a distributed stored-program controlled switching system and discusses the resulting software structure as present in No. 5 ESS. To facilitate the software development process, a methodology based on current software engineering principles is employed. Sufficiently decoupled software components have expedited parallel development of software and provided reliability by reducing interaction and limiting propagation of errors among software components. The structuring strategy described for telephone switching is equally applicable to other distributed real-time control systems.

Known solutions of a large number of important (and difficult) computational problems called NP-complete problems depend on enumeration techniques which examine all feasible alternatives. This paper considers the design of enumeration schemes in a distributed environment in an attempt to exploit the parallel activities inherent in enumeration algorithms.

A distinguishing feature of modular design from ad hoc design is the establishment of an intermodule protocol to which all modules adhere. The problem of representing and analyzing intermodule protocol for the control portion of register transfer level systems is outlined. An introduction to two existing graph models of computation indicates that existing register transfer level module sets are representable by various "token flow" models. A single model that is capable of representing the token flow models and some of its analytical properties are illustrated by example. Finally, three examples of deadlocks in existing modules sets are presented. These deadlocks were uncovered by the analytic properties of the new model. One example is due to incorrect interconnection of modules at the user level. The other two illustrate incorrect signaling conventions between modules necessitating a redesign of some modules.