No abstract

No abstract

Designing a Wireless Sensor Network (WSN) to achieve a high Quality of Service (QoS) (network performance and durability) is a challenging problem. We address it by focusing on the performance of the 802.15.4 communication protocol because the IEEE 802.15.4 Standard is actually considered as one of the reference technologies in WSNs. In this paper, we propose to control the sustainable use of resources (i.e., energy consumption, reliability and timely packet transmission) of a wireless sensor node equipped with photovoltaic cells by an adaptive tuning not only of the MAC (Medium Access Control) parameters but also of the sampling frequency of the node. To do this, we use one of the existing control approaches, namely the viability theory, which aims to preserve the functions and the controls of a dynamic system in a set of desirable states. So, an analytical model, describing the evolution over time of nodal resources, is derived and used by a viability algorithm for the adaptive tuning of the IEEE 802.15.4 MAC protocol. The simulation analysis shows that our solution allows ensuring indefinitely, in the absence of hardware failure, the operations (lifetime duration, reliability and timely packet transmission) of an 802.15.4 WSN and one can temporarily increase the sampling frequency of the node beyond the regular sampling one. This latter brings advantages for agricultural and environmental applications such as precision agriculture, flood or fire prevention. Main results show that our current approach enable to send more information when critical events occur without the node runs out of energy. Finally, we argue that our approach is generic and can be applied to other types of WSN.

No abstract

This paper presents a simple and distributed clustering algorithm suitable for large-scale wireless sensor networks (WSNs) consisting of several thousands of sensor nodes and few sink nodes. A two-tiered hierarchical architecture is used to increase scalability and ensure performances and durability of such a system: Level 1 called sensor network is partitioned into several equilibrate clusters with one leader or sink by cluster; Level 2 also called sink network is composed by N sink nodes placed in planned manner into monitored region and is connected through IEEE 802.11 radio interfaces. A multi-channel system is used to create a cellular structure by assigning one frequency channel per cluster. We use simulation technique to evaluate and compare the impact of two distributed schemes (a single channel one and a multi-channel one) on network capacity like traffic load, energy consumption, medium access delay, end-to-end delay and data delivery ratio.