An Overview of Memory Provisioning in Real-Time Wireless Sensor Networks

Wireless Sensor Networks and the smart applications designed to operate upon them have enjoyed a rapid increase in popularity over the last decade. The main challenge currently is the provision of real-time service delivery for wireless sensor networks to cater for new applications with guaranteed Quality of Service (QoS) requirements. However each application has a different service requirement. In order to deliver real-time services the dimensioning of such networks is important to service providers in order to meet these service requirements. If packets cannot be stored due to insufficient memory they are lost. Lost packets result in the resending of the packets and hence an increase in delay in delivery of the application traffic. It is this memory provisioning of these wireless sensor networks that is the focus of the work presented in this paper. More specifically the relationship between the application design, implementation and memory resources required to run the service are explored using a stack analysis tool. This stack analysis tool enables the stack footprint to be measured. Results of memory usage for two different WSN applications are presented. Recommendations based on this study for efficient memory provisioning and ultimately real-time service delivery are given.