Middleware for wireless sensor network virtualization

Abstract

Sensor and network virtualization technology are used in smart home, smart grid, smart city and many other applications of Internet of Things (IoT) that deploy Wireless Sensor Network (WSN) to facilitate multiple sensor data transmission over multiple networks. Existing WSNs are designed for a specific application running on low data rate network. The challenge is how to ensure multiple sensor data for multiple applications be transmitted over multiple heterogeneous networks having different transmission rates while ensuring Quality-of-Service (QoS). The research has developed a middleware that provides sensor and network virtualization with guaranteed QoS. The middleware was designed comprising of two layers: Application Dependent Layer Middleware (ADLM) and Network Dependent Layer Middleware (NDLM). The ADLM combined multiple sensor data to form services based of Service Oriented Application (SOA). It is comprised of service handling manager that combines various sensor data and form services, QoS manager that assigns priority and service scheduling manager that forwards the service frames. The NDLM facilitated seamless transmissions of various service data over multiple heterogeneous networks. It consists of hypervisor which is composed of flowvisor and the powervisor. The flowvisor is madeup of transmit and routing managers responsible for routing and transmitting service packets. The powervisor consists of a resource manager that determines and selects the node with the highest battery power. The middleware was implemented and evaluated on a real experimental testbed. The experimental results showed that the middleware increased throughput by 8.7% and reduced the numbers of packets transmissions from the node by 68.7% compared to proxy middleware using SOA. In addition, end-to-end transmission delay was reduced by 85.2% when compared to SenShare using SOA. The flowvisor at the gateway decreased the waiting time of packets in the queue by 59.8%, when the flowvisor raised the output rate up to 2.5 times the maximum arrival rate of WSN packets. The powervisor increased the node’s life time by 17.6% when compared to VITRO by limiting the transmission power to the existing battery voltage level. In brief, the middleware has provided guaranteed QoS by increasing throughput, reducing end-to-end delay and minimizing energy consumption. The middleware is highly recommended for IoT applications such as smart city and smart grid.