دانلود رایگان مقاله به کارگیری شبکه حسگر بی سیم با توجه به مساله حفره انرژی

Abstract

Heterogeneous sensor networks have been proposed to address some fundamental limits and performance issues present in homogeneous Wireless Sensor Networks (WSNs). Questions such as the number of high-end sensors should be used, and how to deploy them, need proper assessment. In this work, we propose a novel model capable of representing a wide variety of scenarios, from totally random to planned stochastic node deployment in both homogeneous and heterogeneous sensor networks. In particular, this model encompasses networks with the characteristics of small-world networks. Using only about 3% of high-end sensors, and deploying nodes by using the slightly attractive model defined herein, we observe improved characteristics of the network topology, such as: (i) low average path length, (ii) high clustering coefficient, and (iii) improved relay task distribution among sensors. We also provide a guide for deploying nodes in order to improve the network lifetime, showing that the aforementioned model can be used to diminish the energy hole effect. Moreover, we evaluate a topological metric, namely Sink Betweenness, suitable for characterizing the relay task of a node.

5. Final remarks

We proposed a novel modeling solution capable of representing a wide variety of scenarios, from totally random to planned stochastic node deployment, in heterogeneous sensor networks. This model can represent WSNs with characteristics of small-world networks, and can address the energy hole problem. We showed that by using only about 2% of H-sensors (20 out of 1000) and deploying nodes by using the P- and Q-models to distribute the L-sensors around the H-sensors deployed with a repulsive model, we observe important characteristics of the network topology, such as low average path length and high clustering coefficient. Observing the results for coverage, SBet and small world parameters for the two models evaluated herein, we notice that the Q-model is the most suitable for addressing the energy hole in homogeneous networks, while the P-model achieves the highest coverage and overcomes all other models for heterogeneous networks. The P-model acts only in the first corona, while the Q-model acts in all coronas. Regarding the small world properties, the P-model leads to higher cluster coefficient, while Q-model leads to lower average path length, characteristics that are desirable for WSNs. Moreover, we showed that the SBet is a suitable metric for characterizing the relay task of a node. This metric’s ability, in contrast to the relative insensitivity of the classical Betweenness, suggests other possibilities. SBet can be used in a wide variety of applications, both in the design and operation of WSNs. For instance, the designer can assess the best deployment strategy in order to create graphs with a more appropriate SBet distribution. Such an assessment should improve the understanding and management of the network lifetime, since the energy consumption becomes more evenly distributed among the nodes. Studies in that direction require only spatial point process generators (in order to model the deployment models), and tools for graph analysis; therefore there is no need for either complex discrete event simulators or network models. Both are provided by R, a free, multiplatform software environment for statistical computing and graphics, which exhibits excellent numerical properties [3]. We also envision the following research lines: the quantification of the relationship between the metrics used herein, such as the clustering coefficient, the average path length, and the SBet with the fault-tolerance properties, latency and network lifetime, respectively; the introduction of fault-tolerance schemes based on the proposed model and metric; the use of topology control schemes, based on the SBet, to diminish the possibility of interference on nodes that were attractively deployed around the Hsensors and the sink, and the use of SBet to improve the routing performance in WSNs.