Spatial Distribution Patterns, Power Law, and the Agent-based Directed Diffusion Sensor Networks

Abstract

The directed diffusion sensor networking was proposed by Intanagonwiwat et al. (2003) as an alternative to the IP-based end-to-end and the hop-by-hop ad-hoc routing architectures. This paradigm was apparently inspired by Alan Turing's (1952) seminar paper on the morphogenesis and pattern formation theory. Among numerous follow-up studies to the directed diffusion architecture, Malik and Shakshuki (2007) recently extended the directed diffusion networking by introducing mobile agents (MA) with the goal of increasing the operation efficiency, and the extended paradigm is termed agent-based directed diffusion (AbDD). We suggest two potential improvements to the AbDD, introducing spatial distribution modeling and optimizing the number of mobile agents (MA) employed. We choose to seek inspiration from the spatial distribution patterns of insect populations, which is an emergent property of the individual behaviors at the population level. An insect population can be envisioned as a multi-modal wireless network in which individuals are capable of processing multimodal signals (e.g., auditory, visual, and chemosensory) from environment and their peers. From evolutionary theory, we can expect that insect spatial distribution patterns should be optimized to facilitate the communications among population individuals since communications regulate or control the insect behaviors, which in turn, play crucial roles in the survival and prospering of insects.