دانلود مقاله ISI انگلیسی شماره 28799
ترجمه فارسی عنوان مقاله

روش شبکه های بیزی برای یک استراتژی حرکت الهام گرفته از بیولوژیک برای شبکه های تلفن همراه حسگر بی سیم

عنوان انگلیسی
A Bayesian network approach to a biologically inspired motion strategy for mobile wireless sensor networks
کد مقاله سال انتشار تعداد صفحات مقاله انگلیسی
28799 2009 12 صفحه PDF
منبع

Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)

Journal : Ad Hoc Networks, Volume 7, Issue 6, August 2009, Pages 1217–1228

ترجمه کلمات کلیدی
شبکه های بیزی - اتصال به کامپیوتر - پوشش - طول عمر - تحرک - شبکه های حسگر بی سیم - خطا شبکه های تحمل -
کلمات کلیدی انگلیسی
Bayesian networks, Connectivity, Coverage, Lifetime, Mobility, Wireless sensor networks, Fault tolerant networks,
پیش نمایش مقاله
پیش نمایش مقاله   روش شبکه های بیزی برای یک استراتژی حرکت الهام گرفته از بیولوژیک برای شبکه های تلفن همراه حسگر بی سیم

چکیده انگلیسی

Mobility strategies for wireless sensor networks (WSNs) are presented. We introduce a grazing mobility strategy for mobile WSNs, inspired by the foraging behaviour of herbivores grazing pastures. We present Bayesian network GRAZing (BNGRAZ) that implements the proposed WSN grazing strategy. BNGRAZ uses local neighbourhood information to predict coverage and connectivity performance changes related to sensor node motion characteristics. This enables a sensor node to predict the performance implications related to its direction of movement. We implement the BNGRAZ approach to grazing in a custom built mobile WSN simulator. The WSN performance criteria considered during the validation process include coverage, redundancy, connectivity, and network lifetime.

مقدمه انگلیسی

Recent advances in sensor technology (in terms of size, power consumption, wireless communication and manufacturing costs) have enabled the prospect of deploying large quantities of sensor nodes to form wireless sensor networks (WSNs). These networks are created by distributing large quantities of usually small, inexpensive sensor nodes over a geographical region of interest in order to collect data relating to one or more variables. These nodes are primarily equipped with the means to sense, process and communicate data to other nodes and ultimately to a remote user(s). WSN nodes can also have mobility capabilities which enable them to roam the region of interest to harvest information. Sensor nodes may cooperate with their neighbours (within communication range) to form an ad-hoc network. WSN topologies are generally dynamic and decentralized. WSNs have a wide range of applications including military, environmental monitoring, health, home, space exploration, chemical processing, and disaster relief. The majority of WSN research has assumed that the nodes are static and that once deployed they are unable to relocate. This limits the ability of WSNs to adapt to changing operating environments. A large number of applications involve a dynamic environment and/or do not necessarily need the deployment of large quantities of static nodes. A mobile WSN varies from traditional static WSNs in the obvious sense. A fraction of the sensor nodes can have motion capabilities which enable the WSN to change position over time according to some strategy. This motion may be achieved by including motors and servos onto the node platform. Mobility capabilities may also be possible by attaching the nodes to other mobile entities. This gives the nodes the ability to physically change position in relation to neighbouring nodes and also the environment in which the nodes are situated. The nodes may move individually to optimize the performance of the network. Mobility strategies aimed at this have been presented by Kansal et al. [5], [6] and [7], and Rao and Biswas [9]. Nodes may also move cooperatively around the geographical region of interest to adapt to the environment and/or application criteria. When deploying a WSN in a region of interest a number of alternative approaches can be adopted. The region could be covered with a large quantity of static nodes, which achieve the desired coverage criteria at initial deployment. The network would generally incorporate a high level of redundancy in order to extend its lifetime, with redundant nodes providing a level of fault tolerance. The drawback of this approach is that optimum deployment is required to achieve the desired level of performance. When considering deployment to unmanned remote regions, optimum deployment becomes difficult to achieve. An alternative approach is to deploy a smaller quantity of mobile nodes. These nodes would not achieve the desired level of coverage at any instantaneous point in time; however, over a finite period of time total coverage (specified coverage criterion) could be achieved. In order to achieve this nodes migrate around the geographical region collecting data and thus providing total coverage. This concept will be referred to as a grazing strategy, by analogy to a herd of herbivores feeding off pastures. This paper presents a mobility scheme based on a decentralized Bayesian network GRAZing (BNGRAZ) algorithm that adopts the proposed WSN grazing strategy. BNGRAZ uses discrete Bayesian networks to predict the likelihood of deterioration in performance given that the sensor node moves in a particular direction. The evidence used for prediction is obtained from local neighbourhood information, which minimizes the communication overheads and provides scalability. The performance criteria considered under the BNGRAZ algorithm include connectivity and coverage. The paper also presents a distributed coverage approximation (CA) algorithm. This algorithm enables the sensor nodes to approximate the collective coverage of the WSN using only local knowledge of neighbouring node configuration. The CA algorithm is required for the successful operation of the BNGRAZ algorithm. The simulation results have been obtained by implementing the proposed BNGRAZ algorithm using a custom built simulator which allows the evaluation of the performance of mobile WSNs. Section 2 of this paper outlines related work on mobile WSNs, and their inherent performance implications and benefits. Section 3 discusses the grazing strategy in detail outlining the performance implication and benefits. Section 4 discusses the proposed decentralized BNGRAZ algorithm that aims to achieve the grazing motion behaviour in WSNs. In Section 5 we present the simulation results and compare the results to a generic fixed path approach. Finally, 6 concludes the paper and discusses future work.

نتیجه گیری انگلیسی

This paper has presented a novel biologically inspired mobility strategy for WSNs. We argue that instead of deploying a large number of static sensor nodes which provide total coverage at any instant in time, a relatively smaller number of mobile nodes can be deployed within the region of interest. These nodes adopt a grazing strategy to achieve the desired coverage over time. We proposed a decentralized discrete Bayesian network to implement the grazing strategy (BNGRAZ). The algorithm predicts the WSN’s performance with respect to a sensor node’s mobility. This is to enable the latter (sensor node) to derive the optimum direction of motion. Performance predictions are derived at the level of the sensor node based on neighbour positions and a measure of the overall WSN coverage. A sensor node’s position is assumed to be Gaussian in nature with a variable spread; this approach takes into account the fact that the knowledge about a particular node’s position will become less accurate the longer one goes without communicating with it. Simulation of the impact of varying node speed, total number of nodes, reconfiguration duty time and desired coverage period has been carried out. The results have been presented and show promising prospects for the use of the grazing strategy. Work is under way to compare the performance of the latter with that of alternative mobility strategies. The work presented here has also generated models which are realistic in essence and enable the WSN designer to test various mobility scenarios and evaluate their performance quantitatively ahead of implementation/deployment in the face of uncertainty–uncertainty associated with measurements and uncertainty resulting from constrained resources with a view to maximize system lifetime.