Recent advances in low power microprocessors and power storage technology have allowed for the design of a broad range of sensor network platforms. Current sensor network platforms fall into one of two main categories, low power, application specific platforms and full featured general purpose platforms with higher power consumption. The former generally consist of a power-efficient microprocessor and a few application specific resources, while the latter generally have a general purpose processor and a full complement of resources available via a GPIO interface. General purpose platforms are readily available and can be adapted to almost any application but are less efficient with respect to power consumption than application specific solutions. Application specific solutions are extremely efficient and well suited to the problem they are designed for, but are time consuming and expensive to build. MASS is a modular sensor node architecture that enables rapid development of an application specific solution. While not as power-efficient as a completely custom solution, MASS provides application flexibility while being much more efficient than common-off-the-shelf general purpose sensor platforms. Because MASS is modular in both it's software and hardware design, it is straightforward to extend both the software platform and the hardware supporting it to suit an application. What this means is that a custom sensor network solution is now available at a much lower cost, and in a much shorter time-frame than was previously available. This makes a whole new class of sensor network applications feasible.

The system level hardware architecture of individual nodes in a distributed wireless sensor network has not received adequate attention. A novel hardware architecture based on the concept of task specific modular computing provides both the high flexibility and power efficiency required for effective distributed sensing solutions. This paper presents this newly developed architecture and provides an analysis of two possible methods for module interaction. Results from simulations based on this analysis are given. Finally, a brief description of the hardware and software design and prototype implementation of the modular architecture for sensor systems (MASS) is given to show the ease of use of the conceptual architecture.