Multiagent systems have been considered as a potential solution for developing adaptive systems. In this research, a cellular self-organizing (CSO) approach is proposed for developing such multiagent adaptive systems. The design of CSO systems however is difficult because the global effect emerges from local actions and interactions that are often hard to specify and control. In order to achieve high-level flexible and robustness of CSO systems and retain the capability of specifying desired global effects, we propose a field-based regulative control mechanism, called field-based behavior regulation (FBR). FBR is a real-time, dynamical, distributed mechanism that regulates the emergence process for CSO systems to self-organize and self-reconfigure in complex operation environments. FBR characterizes the task environment in terms of “fields” and extends the system flexibility and robustness without imposing global control over local cells or agents. This paper describes the model of CSO systems and FBR, and demonstrates their effectiveness through simulation-based case studies.
