Information complexity, information matching and system integration

This paper explores information transaction between different entities comprising a system to arrive at a unified way of describing the efficiency of integrated working of a system. Such a unified view would make it possible to optimize an entire system even if it is comprised of very dissimilar elements such as hardware, software, humans and intelligent machines. The approach is based on the connections between communication theory, computational complexity, thermodynamics and system integration. We formulate the integrated working of two entities as an information engine and examine its optimality in the thermodynamic sense by analogy with the Carnot cycle. Concepts of information generation efficiency and utilization efficiency are derived. Some applications are discussed