Study Of Quality Of Service In Wireless Sensor Networks

Wireless Sensor Network(WSN)is a distributed system consists of nodes which are discretely distributed in the environment,equipped with specialized sensors,capable of network wireless communication and collaborate with each other to monitor targets/events in the area of interest WSN has become a fundamental technology in the industry and social life,widely applied in field monitoring,environment monitoring,industry monitoring,city management and healthcare,etcThe Quality of Service(QoS)is essential in evaluating the design and actual performance of WSN,and has been extensively studied in the literature.Different from the traditional network which focuses mainly on data communication,WSN also fulfills the task of sensing even changing the physical environment.Thus,the QoS of WSN can be divided into two major parts:reliability of sensing and reliability of data transfer.The reliability of sensing evaluates WSN's capability of accurately,completely and efficiently acquiring the information from the surrounding environment or spotting the target event.Reliability of data transfer evaluates whether the obtained information can be delivered to the control/data center in a timely and accurate fashion.Moreover,constrained by its low cost,large scale and limited power source,WSN's QoS is also highly dependent on network management technology,such as time synchronization.This thesis investigates WSN's QoS from the aspect of reliability of sensing,reliability of data transfer and network management,respectively.1.The reliability of sensing reflects WSN's capability of interacting with the physical envi-ronment.The sensing capability of a single node is usually limited.Thus it is common for multiple nodes to collaborate for better sensing coverage.In WSN with mobile nodes,because not all nodes are equipped with GPS,their precise locations are usually unknown.Locations obtained through reference of other nodes contain non-negligible errors,which make it difficult to move the nodes to expected location.This thesis considers the problem of barrier coverage,where the area of interest is of a narrow belt shape,and studies how to efficiently move the nodes to achieve the expected coverage,taking into account the location errors.We first investigate the relation between the un-covered holes and the location error,and find that the probability of the hole is decided by the accumulative distribution of the location error.Then we propose an efficient method to calculate the probability of the holes.At last,we conduct extensive experiments and prove the effectiveness of the proposed method2.The quality of data transfer is essential in QoS analysis,especially for systems provid-ing real-time services.It has been a research challenge to get end-to-end delay from hop-by-hop,single-link delay.Traditional methods work in the time domain and its computational cost increases significantly with the size of the network,thus can hardly be scaled up.This thesis proposes a novel framework that computes the end-to-end delay in the frequency domain.After modeling the end-to-end delay distribution with the signal flow graph,we prove that the end-to-end delay distribution is actually the inverse Laplace transform of the transfer function of the signal flow graph.Two ef-ficient methods,Cramer' s rule-based method and Mason gain rule-based method,are adopted to obtain the transfer function.By analyzing the time responses of the transfer function,we obtain the end-to-end delay distribution.To improve the QoS of network transmission in WSN,we present an efficient method for spotting network bottleneck by utilizing the dominant poles of the transfer function,a method to analyze the network performance of different retransmission protocols using the presented framework,as well as a method to analyze and pick out the energy efficient network path for transmission.Theoretical analysis and extensive evaluations show the effectiveness of the proposed approach3.Network management is also referred as the middleware,which supports the WSN.Time synchronization is a fundamental technology used in network management.Instead of the tradi-tional view of providing as high accuracy as possible to all time synchronization scenarios,this thesis insists that the accuracy of time synchronization should be case-specific.We take the tar-get tracking system as an example and exploit the synchronization requirements by different QoS views such as network lifetime,sensing delay and sensing accuracy.We propose an energy effi-cient Sensor Array Synchronization Protocol(SASP),which provides the required synchronization accuracy to guarantee the QoS.Specifically,when no target appears,to guarantee system lifetime,coarse synchronization is achieved with little overhead by piggybacking time information onto pe-riodical network maintenance packets.Once targets appear,SASP achieves high inter-array and relatively higher intra-array synchronization accuracy rather than the traditional network-wide on average high accuracy.In this way,it guarantees reliable communication and accurate data fu-sion while reducing energy consumption.Theoretical analysis and extensive evaluations show the effectiveness of the proposed protocol.