End-to-end quality of service-based admission control via virtual sensor nodes

Abstract

Aiming the real-time acquisition of physiological or vital signs, biomedical wireless sensor networks appear as a key technology to support the development of innovative applications or services. However, due to the critical nature of medical data, these networks have to meet strong quality of service requirements and guarantee high levels of confidence and reliability. These goals depend on several factors, such as nodes deployment, network topology, limited throughput, and on the characteristics of the surrounding environment. Harsh environments as hospital facilities can compromise the radio frequency communications and, consequently, the networks' capability to provide the quality of service required by medical applications. Moreover, the impact of harsh environments in the network performance is difficult to manage due to its random and unpredictable nature. So, the admission of new sensor nodes (i.e. patients) to biomedical wireless sensor networks, while preserving the required quality of service, is very hard to plan and manage. This work offers a method based on real-time measurements of the performance provided by the network that allows controlling the admission of new sensor nodes to the network. In order to assess the network, the proposed technique makes use of "virtual sensor nodes" to mimic the presence of the new real sensor nodes. To validate this approach several experiments were made using both simulations and real deployments. The obtained results show that the proposed approach makes possible to detect and avoid situations in which the admission of new sensor nodes may result in a harmful degradation of the quality of service provided by the network.