Size does matter - The case for big motes

A study conducted by the authors on some contemporary designs for wireless sensor nodes, or motes shows a large disparity in the level of resources, in terms of processor power and available memory. There are essentially two schools of thought on the appropriate scale of hardware to include in a mote design: The 'small mote' camp motivates their approach based on an argument centring on cost, power consumption and potential for integration into a single chip design. The 'large mote' camp cites an expected level of needed computational power, derived from top-down design considerations, as a reason for the scale of the resources designed into their motes. The authors have recently begun to plan some real world, large, field sensing applications, typical of the type of task that may be required for any sensor networks deployed to deliver field maps over real surfaces. In this paper the relative merits of 'big' and 'small' motes are compared in three categories, which are considered by most researchers to be important for the construction of successful wireless sensor networks. The categories are: size, power consumption and cost. In none of these is the big mote found to be at a disadvantage compared to the small mote, and there appear to be definite advantages of big motes with respect to cost and power consumption, at least for the types of application for which large sensor networks may be useful. As a result of these consideration, a specification for a 'big mote' is proposed.