Building Biologically-Inspired Self-Adapting Systems

Abstract

Biological systems are far more complex than systems we design and build today.  The human body alone has orders of magnitude more complexity than our most-intricate designed systems.  Further, biological systems are decentralized in such a way that allows them to benefit from built-in error-correction, fault tolerance, and scalability.  It follows that if we can extract certain properties of biological systems and inject them into our software design process, we may be able to build complex self-adaptive software systems.  

Biological systems’ complexity makes them not only desirable to guide software design, but also difficult to fully understand.  Thus one approach to building software similar to biological systems is by first building models of biology that we can understand.  Then these models can guide the high-level design, or architecture of the software systems, resulting in systems that retain the model’s fault tolerance, scalability, and other properties.  

I present a general outline of how one might use biology to create a model to guide the architecture of a software system, and develop one such model and the resulting architectural style, the tile style, for computational systems that can use a large distributed network of computers, such as the internet, to solve computationally-intensive problems in a discreet, fault-tolerant, and scalable manner.