Energy-Aware Routing in Sensor Networks A Large System Approach

短文En e r g y B a s e d Ult r a-Wid e b a n dMu lt ip a t h Ro u t in g Alg o r it h m inAd Ho c S e n s o r Ne t w o r k sLiu Jie1,Ca o Y a ng1,21School of Electronic and Information,Wuhan Univers ity,Wuhan430072,P.R.China2State Key Laboratory of Software Engineering,Wuhan University,Wuhan430072,P.R.ChinaAb st r act:The Energy based Ultra-Wideband Multipath Routing(EUMR)algorithm for Ad hoc sensor network is proposed.It utilizes the fun ction of UWB positioning to reduce the network communication delay and route overhead. Fu rth ermo re,the algorith m considers energ y consumption,the residual energy and node hops of communication paths to make energy consumption more balanced and extend the network lifetime. Then routing which is stable,energy-saving and low-delay is realized.Simulation results show that the algorithm has better performance on saving energy,route overhead,stability and extending network lifetime.Key wor ds:UWB;multipath;ad hoc network; energy;EUMRI.INTRODUCTIONUWB-Ad hoc senso r networ ks u se Ad h oc network formation and ultra wideband technology as the physical layer transmission.UWB has the characteristics of high transmission rate,large communication capacity,high hiding,low power consumption,precision orientation,etc.UWB is conducive to achieve positioning in the Ad hoc sensor networks dynamic composition of wireless mobile nodes and get high data transfer rate in the short distance.Also,it can achieve long-distance transmission through a multi-hop routing.Thus, UWB is very suitable for Ad hoc sensor network physical layer bination of UWB and Ad hoc technology,the problems such as high-speed access and the frequency resource allocation can be good solved,and the wireless network is developed into a more efficient and extensive practical application.Due to energy constraints,topology changes and b an dwidth co nstrain ts,wireless Ad ho c senso r n etwor k is high r eq uired on ro uting algorithm.multipath routing with the advantagesof small delay,load balanced,large throughput, has significant advantages of effective use of bandwidth and impr ovem ent of tran sm ission reliability and fault tolerance.The current multi-2011.01935path routing protocols mainly are SPIN,DD[1], SAR,TORA,SMR[2],MSR,etc.At present, energy-based multi-path routing protocols are less and representative routing protocols mainly have EAR[3],MMER[4]and so on.In Ref.[3], energy multi-path routing protocol EAR proposed by Shah RC et al sets up multiple paths between source nodes and destination nodes.As the node chose probability is energy related,the energy consumption of communication is distributed to multiple paths,so data transfer balanced consume the energy of the whole network and the network lifetime is extended.EAR protocol can achieve balanced energy consumption in the network to maximum prolong the network lifetime.In Ref.[4],accor ding to the mode of disjoint m ulti-path data tr an smissio n,MMER pr otoco l is designed by the global energy minimization of the routing mechanism to achieve balanced energy consumption and energy ef ciency.In order to reduce energy consumption,reduce routing load,balance energy consumption of the network and extend the network lifetime,based on EAR protocol proposed by Shah RC et al, a scheme called energy based ultra-wideband multipath routing(EUMR)is proposed.The results show that EUMR can balanced consume energy of the network,reduce the transmission delay,extend the network lifetime and reduce the impact on the routing algorithm due to network changes.II.EUMR ALGORITHMA.Basic Idea of EUMR Algor ithmTh e nodes nu mber and data r ed undancy are much more in Ad hoc network.In order to reduce energy consumption and to ensure reliability of data transmission,EUMR consider s energy consumption,the residual energy and node hops of communication paths to make energy consumption more balanced and extend the network lifetime. UWB-Ad h oc senso r n etwo rk usin g UWB tech nolo gy of MAC,UWB techno logy can provide accurate positioning information for the upper layer.This feature will be applied to EUMR algorithm.(1)As data is to be sent,according to UWB positioning message,if destination node is in the range of a hop of the node,source node direct sends data packet without broadcasting path setup massage by destination node.(2)In route setup process,if the intermediate node want to send path setup message,it judges whether the target node in its a hop communication range.If in the range, the target node is directly added to the self route recording table,without broadcasting the message. Using the UWB positioning function,through there is only one hop,several network load can be reduced.B.Descr iption of EUMR AlgorithmIn UWB-Ad hoc sensor network,we suppose:sN is source node;DN is the destination node;()iCost Nis the communication cost of nodeiN reaching the destination node D N.Route setup:The important step of EUMR algorithm is route setup.Every node needs to know all the next hop nodes which can reach the destination node and calculate probability of choosing each next hop node.The process of route setup includes the following steps:(1)Destination node constructs a path setup m ess ag e and br oadcas ts the message to all neighbor nodes.The path setup message has the following format:interest=(type,interval,rect, timeamp,expiresat,cost).The end of data packet of path setup message contains a cost domain, which express the cost of nodes communication.The initial value is0,()0DCost N=.(2)The path setup message is sent between intermediate nodes.Every intermediate n ode f orwards the path setup message o nly to the neighbors that closer to the sour ce node than oneself and farther away from the destination node. Thus at the node i N,the message is sent only to a neighbor node j N,which satis es:(,)(,)i s j sd N N d N N≥102011.06336短文2011.011EUMR:Not only the hops constraint is added, but also route load is optimized by using UWB positioning message.In EUMR-1,the hops value is contained in path setup message,thus it is propitious to choose path with less hops.Also,the hops value is taken into account in calculating chose probability of the next hop,thus in the case of the same energy consumption of the nodes,according to choose the path with less hops,data packet communication delay can be reduced to improve communication ef ciency.In EUMR-2,node is judged whether in the range of communication by using UWB positioning message,through there is only one hop.Thus several load of network can be reduced.EUMR utilizes orientation identity of UWB to reduce the network communication delay and route cost.Furthermore,the routing algorithm considers energy consumption,the residual energy and node hops of communication paths to make energy consumption more balanced and extend the network lifetime.Then energy-saving and low-delay routing was realized.Based on EAR protocol proposed by Shah RC et al,EUMR algorithm is proposed.The routing algorithm utilizes the function of UWB positioning to reduce the network communication delay and route cost.Furthermore,it considers energy consumption,the residual energy and node hops of communication paths to make energy consum ption mor e balanced and exten d the network lifetime.Then routing which is stable, energy-saving and low-delay is realized.III.S IMUL AT IO N R E S ULT S AN D ANALYSISA.Simulation Environment and Par ameter Setting Using NS2simulation platform,the relev ant simulation parameters are as follows:Table I Simulation parameter settingsimulation parameter valuethe size of network area(m2)1000×1000channel bandwidth(Mbps)500(UWB) wireless transmission distance(m)100MAC protocol TDMA the nodes number of network30the initial energy of node(J)5packet generation rate(s)0.001mobile model Random Waypoint node movement intervals(s)10 the average velocity of node(m/s)(0,30)simulation time(min)80α,β,A,B1,50,0.5,0.5B.Per for m ance Standar dEUMR algorithm is on-demand routing algorithm. SMR protocol is a typical on-demand routing protocol with good performance.Many multipath routing protocols,such as MSR,TORA,all adopt a similar route setup mechanism as SMR.EUMR is compared with classic on-demand routing protocol SMR,we analyze performance of routing algorithm on packet delivery rate,average delay,routing overhead,the number of dead nodes, the network lifetime.(1)Packet delivery rate:It is the ratio of the number of data packets received by destination node and the number of packets sent by source node.The higher packet delivery rate is,the less lost packets in the transmission process are and the better performance of the network is.(2)Average end-to-end delay:It is an average difference between the packets received time of destination node and the packets transmitted time of source node.The smaller delay is,the faster response is,and the better performance of the network is[7].(3)Norm alized routin g overhead:It is the ratio of the number of control packets sent by all nodes and the number of total sent packets. This parameter reflects the stability of path in dynamic environment,the smaller the overhead is,and the more stable the path is.122011.063363短文(4)The number of dead nodes:It is the number of nodes,of which the energy is reduced to zero.(5)Network lifetim e:I t is the time when the first node appears in the network.This parameter reflects the overall vitality of the net work.In general,t he longer the node lifetime i s,the network energy balanced consumption will be better controlled [8].C.Analysis of SimulationResultsFig.1Packet delivery rateIt can be observed in the gure that when the network is stabile (for node speed smaller),the packet delivery ratio EUMR and SMR are both high.With the increase of node mobile speed,SMR takes more time for route setup and the lost packets are also more.Thus,the packet delivery rate of SMR protocol decreases significantly.However,the packet delivery rate of EUMR with slow decreased trend is much higher than SMR.In addition,as the character istics of UWB positioning can only provide information in one hop,the network performance is limited improved for EUMR-2,and EUMR-2is not more apparent than EUMR-1.It is suggested that it needs to further excavate thepotential of UWB to improve the performance ofnetwork.Fig.2A verage end-to-end delayFigure 2shows the comparisons of average end-to-end delay changing with node mobile speed.It can be seen that when the node mobile speed is small and the network is stable,the average end-to-end delay of SMR protocol is less than EUMR.The delay of SMR and EUMR increase withnode mobile speed increasing.In general,because EUMR chooses the path with smaller hops inthe case of equal energy consumption and utilizes UWB positioning function to reduce certain data packet transmission delay to improve communication efciency.Fig.3Normalized routing overhead2011.0163Figure 3shows the comparisons of normalized routin g over head changing with node mobile speed.In the results,routing overhead of EUMR is the lowest and routing overhead of SMR is the highest.When node does not move,although SMR discards duplicate request packets,as EUMRutilizes UWB positioning message and path setup message packet is not needed to sent for data communication in the range of one hop,route load is decreased.Also,path setup message can decrease a hop of broadcast,and routing overhead is effectively limited.Due to EUMR-2have no constraint on hop,routing overhead of EUMR-2is less than SMR and more than EUMR-1.As EUMR combines the optimization of EUMR-1and EUMR-2,routing overhead of EUMR is lowest.As the node mobilespeed increases,the routing overhead of SMR and EUMR both increase,EUMR algorithm has obviousadvantages.Fig.4The number of dead nodesFigure 4shows the number of dead n odes changing with node mobile speed.With node mobile speed increasing,the numbers of dead nodes increase.Because EUMR-1and EUMR-2consider energy consumption rate and more efficient consume the energy of node,the dead nodes number of EUMR-1or EUMR-2is less than SMR.The dead nodes numbers of EUMR-1,EUMR-2and EUMR are close to each other,as the node move randomly lead to predictionsinaccurate.Fig.5The network lifetimeFigure 5shows the network lifetime changing with node mobile speed.With node mobile speed increasing,the network lifetime decrease.Mainly because the network topology changes increase and control overhead increases with nodes mobile speed incr easing,the network lif etime arrive in advance.The network lifetime of EUMR-1,EUMR-2or EUMR is longer than SMR.IV .CONCLUSIONSIn order to reduce energy consumption,ensure reliability of data transmission and improve network performance,EUMR algor ithm is proposed.The routing algorithm utilizes thef unction of UW B po sitio ning to r ed uce th e network communication delay and route overhead.It considers energy consumption,the r esidual energy and node hops of communication paths and calculate probability of choosing each next hop node on communication paths.The results show that by choosing stable ,energ y-sav ing an d low-delay routing basedon UWB positioning function and energy,EUMR can balan ced co ns um e th e en erg y of network,r edu ce the transm ission delay ,extend the network lifetime and improve the performance of network.142011.0短文References[1]INT ANAGONWIW AT C,GOVINDAN R,ESTRIN D,etal.Directed Diffusion for Wireless Sensor Networking[J].IEEE/ACM working,2003,11(1):2–16.[2]LEE S J,GERLA M.Split Multipath Routing withMaximally Disjoint Paths in Ad hoc Networks[R].IEEE ICC2001,Helsinki,2001:3201–3205.[3]SHAH R C,RABAEY J M.Energy Aware Routing forLow Energy Ad Hoc Sensor Networks[C]//Proceedings of IEEE Wireless Communications and Networking Conference.Piscataway,USA,2002:17–21.[4]YIN Shouyi,LIN Xiaocang.Multipath Minimum EnergyRouting in Ad Hoc[R].IEEE International Conference, 2005:3182–3186.[5]INTANAGONWIWAT C,GOVINDAN R,ESTRIND.D irect ed Di ffus ion:A Scalab le and Rob us tCommunication Paradigm for Sensor Networks[C]// Proceedings of ACM MobiCom2000,Boston,MA,2000: 56–67.[6]WANG Quandi,LI Bing,LIU Qingsong.Research onenergy Multipath Routing for Wireless Sensor Network[J].Information and Control,2006,35(2):130–132.[7]H EI NZ E LM AN W R,CHA N D RAK A SA N A,BALAKRISHNAN H.Energy-Efficient Communication Protocol for Wireless Mi crosensor N etwo rks[C]//Proceedings of the33rd Hawaii International Conferenceon System Sciences,2000:1–10.[8]MANN R P,NAMUDURI K,PENDSE R.Energy AwareRouting Protocol(EARP)for Ad-Hoc Wireless SensorNetworks[C]//Proceedings of IEEE CommunicationsSociety Globe com2004Workshops,2004:419–427. BiographiesLiu J ie,received h er Mas ter degree ofCommunication and Information System in July2006at Wuhan University of technology.She iscurrently PhD candidate at School of Electronic and Information,Wuhan University.Her research interests includes UWB wireless communication and wireless sensor networks.E-mail:LJ_IVY@Cao Y an g,received his Bachelor degree ofElectronic Engineering in1967at HuazhongUniversity of Science and Technology.He iscurrently a professor at School of Electronic and Information,Wuhan University.He is also an assistant directorof Information Network Technology Committee in China Institute of Communication,China.His research interests the next Generation internet and security,wireless sensor networks and applications,the design theory and methods of Soc.E-mail:caoyang@2011.01365。

►个人简介郜帅，博士，1980年8月生。

副教授，河南济源人，2004 年至今在北京交通大学电子信息工程学院工作，2008 年2 月至2009 年2 月期间获国家公派资助赴美国得克萨斯州大学阿灵顿分校交流访问一年。

先后主持或参与国家重大专项、973 项目、863 项目、国家自然科学基金项目等二十余项，获得授权发明专利10 余项，提交国际标准草案4 项（其中IETF 工作组草案1 项）。

►联系方式办公电话：51684274电子邮箱：shgao@办公地点：机械工程楼D701C►研究方向1. 无线传感器网络：系统体系结构、路由协议、移动性、定位、拓扑控制等。

2. 未来信息网络体系：路由交换、网络性能评估、分析等。

3. 移动互联网：系统架构、移动路由、移动组播等。

►科研项目►部分在研课题1. 科技部863子课题网络信息安全管理体系架构及理论-2 2012-2013主持2. 国家重大专项子课题移动互联网网络与信息安全技术研究-22011-2012 主持3. 国家自然基金青年基金机会传感器网络感知路由关键技术2012-2014 主持►学术著作1.Shuai Gao, Hongke Zhang and Sajal K. Das. Efficient data collection in wireless sensor networks with path-constrained mobile sinks. IEEE Transactions on Mobile Computing. 2011, 10(4): 592-608.2.Shuai Gao, Hongke Zhang. Energy Efficient Path-constrained Sink Navigation in Delay Guaranteed Wireless Sensor Networks. Journal of Networks. 2010, 5(6), pp. 658-665.3.郜帅，霍宏伟，张宏科等，基于数据采集量均衡的移动无线传感器网络节能机制，通信学报，2009，30(9)，pp. 109-116.4.郜帅，张宏科，徐怀松，sink轨迹固定传感器网络的高效数据采集机制，软件学报，2010，21(1)，pp.147-162.5.郜帅等，时延受限传感器网络移动sink 路径选择方法研究，电子学报, 2011, 39(4): 742-747.6.Shuai Gao, Hongke Zhang and Sajal Das. Efficient Data Collection in Wireless Sensor Networks with Path-constrained Mobile Sinks. In: Proc.of the 11th IEEE Int’l Symp. on a World of Wireless, Mobile and Multimedia Networks (WoWMoM), 2009.7.Shuai Gao, Yanchao Niu, Hongwei Huo and Hongke Zhang. An Energy Efficient Communication Protocol Based on Data Equilibrium in Mobile Wireless Sensor Network. In: Proc. of the 3rd Int’l Conf. on Mobile Ad-hoc and Sensor Networks(MSN), 2007.8.Shuai Gao, Hongke Zhang. Optimal Overlapping Time Partition in Sensor Networks with a Path-constrained Mobile Sink. In: Proc. of the 1st IEEE Int’l Conf. on Communication Technology and Application (ICCTA), 2009.9.Shuai Gao, Hongke Zhang, Tianfei Song and Ying Wang. Network Lifetime and Throughput Maximization in Wireless Sensor Networks with a Path-constrained Mobile Sink. In: Proc of Int’l Conf. on Communications and Mobile Computing (CMC), 2010.10.Shuai Gao, Linjuan Zhang and Hongke Zhang. Energy-aware spray and wait routing in mobile opportunistic sensor networks. Proc.3rd IEEE Int’l Conf. on Broadband Network & Multimedia Technology (IC-BNMT), 2010.11.Lili Wang, Shuai Gao and Hongke Zhang. Distributed PMIPv6 based on the prefix aggregation. Proceedings of AIAI, 2011.12.Huaming Guo, Shuai Gao, et al., Inter-domain routing with ASnumber: A traffic engineering perspective. Proc. 1st Int’l Symp. on Computer Network and Multimedia Technology (CNMT), 2009.13.Jianfeng Guan, Shuai Gao, et al., The analysis and simulation of multicast join delay. Proc. IEEE Int’l Conf. on Network Infrastructure and Digital Content (IC-NIDC), 2009.14.霍宏伟，郜帅等，基于室内传播模型的无线传感器网络节点部署策略研究，中国工程科学，2008，10(9).15.李昭桦，郜帅等，IEEE802.15.4网络的IPv6报头压缩研究，北京交通大学学报，2007，31(5).16.霍宏伟，郜帅等，一种实时轨道监测无线传感器网络与服务模型研究，铁道学报，2008，30(6).►获奖荣誉1.BJTU IPv6无线移动路由器，2005年度北京市科学技术一等奖（排名5）2.BJTU IPv6微型传感路由器，2008年度电子学会电子信息科学技术二等奖（排名5）3.高性能IPv6路由器协议栈软件，教育部科技成果鉴定，2004年（排名8）4.BJTU IPv6无线移动路由器，教育部科技成果鉴定，2004年（排名5）5.BJTU IPv6微型传感路由器，教育部科技成果鉴定，2005年（排名2）6.BJTU IPv6网络性能分析报告系统，教育部科技成果鉴定，2005年（排名5）7.一体化标识网络系统，教育部科技成果鉴定，2009年（排名4）。

第19卷 第2期 太赫兹科学与电子信息学报Vo1．19，No．2 2021年4月 Journal of Terahertz Science and Electronic Information Technology Apr．，2021文章编号：2095-4980(2021)02-0224-05WSNs中优化移动信宿路径的数据收集算法吕 虹(贵州广播电视大学(贵州职业技术学院)信息工程学院，贵州贵阳 550001)摘要：收集数据是部署无线传感网络(WSNs)的根本目的。

采用移动信宿策略可有效缓解 WSNs的能耗问题，信宿的移动路径是该策略的关键。

为此，提出基于伪驻留点的数据收集(VRDC)算法。

VRDC算法先依据驻留点规划信宿路径，再依据路径选择伪驻留点(VRPs)。

VRPs可通过一跳直接向移动信宿传输数据，而其他的节点则将数据传输至最近的VRPs，进而减少传输跳数，降低能耗。

仿真结果表明，提出的VRDC算法能有效降低能耗，并平衡节点间的能耗。

关键词：无线传感网络；数据收集；路径规划；伪驻留点；能耗中图分类号：TN926；TP393 文献标志码：A doi：10.11805/TKYDA2019493Data collection algorithm for optimizing mobile accommodation path in WSNsLYU Hong(Electronic and Computer Engineering，Guizhou Radio & TV University，Guiyang Guizhou 550001，China)Abstract：Data collection is the fundamental purpose of deploying Wireless Sensor Networks(WSNs).The adoption of mobile accommodation strategy can effectively alleviate the problem of energyconsumption of WSNs. The mobile path is the key of the strategy. Therefore, Virtual Rendezvous points-based Data Collecting(VRDC) algorithm is proposed in this paper. The VRDC algorithm first planslodging paths based on the host Points, and then selects Virtual Rendezvous Points(VRPs) based on thehost Points. VRPs can transmit data directly to the mobile address via one hop, while other nodes cantransmit data to the nearest VRPs, thereby reducing the number of transmission hops and energyconsumption. Simulation results show that VRDC algorithm can effectively reduce energy consumption andbalance energy consumption among nodes.Keywords：Wireless Sensor Networks；data collection；path planning；Virtual Rendezvous Points；energy consumption无线传感网络(WSNs)[1]已在多个应用范围内使用，如军事监测、环境勘察、智慧农业、智能家居、康复医疗等。

无线传感器网络路由协议无线传感器网络（Wireless Sensor Network，WSN）是由大量低成本、低功耗的传感器节点组成的网络系统，用于感知和收集环境信息。

无线传感器网络的路由协议起着关键作用，它决定了数据在网络中的传输路径和方式，影响着整个网络的性能、能耗以及生存时间。

1. LEACH（Low-Energy Adaptive Clustering Hierarchy）是一种经典的层次化路由协议。

它将网络中的节点划分为若干个簇（Cluster），每个簇有一个簇首节点（Cluster Head）。

簇首节点负责收集和聚合簇内节点的数据，并将聚合后的数据传输给基站节点，从而减少了网络中节点之间的通信量，节省了能耗。

2. AODV（Ad Hoc On-Demand Distance Vector）是一种平面路由协议，适用于无线传感器网络中节点数量较少且网络拓扑较稳定的情况。

AODV协议通过维护路由表来选择最短路径，当节点需要发送数据时，它会向周围节点发起路由请求，并根据收到的响应建立起路由路径。

3. GPSR（Greedy Perimeter Stateless Routing）是一种基于地理位置的路由协议。

它通过利用节点的地理位置信息来进行路由选择，具有低能耗和高效的特点。

GPSR协议将整个网络划分为若干个区域，每个节点知道自己的位置以及周围节点的位置，当需要发送数据时，节点会选择最近的邻居节点来进行转发，直到达到目的节点。

除了以上几种常见的路由协议，还有很多其他的无线传感器网络路由协议，如HEED（Hybrid Energy-Efficient Distributed clustering）、PEGASIS（Power-Efficient Gathering in Sensor Information Systems）等，它们各自具备不同的优势和适用场景。

总之，无线传感器网络的路由协议在保证数据传输可靠性和网络能耗方面起着重要的作用。

α Tree in Sensor NetworkPatrick Y.H. Cheung, and Nicholas F. Maxemchuk, Fellow, IEEERouting Problem in Sensor NetworkWe view the objective of the routing layer to be to choose paths through the sensor network to the sinksthat maximize the lifetime of the network by minimizing energy consumption. Two competing propertiesof a sensor network must be considered when selecting paths:(a)Data from different sensors can be aggregated together, reducing the overall energy costs. However,only data that traverses common paths can be aggregated together.(b)The message arrival process is an impulse. When an event occurs, many sensors detect the event andreport their measurements. During an impulse, those sensors that are chosen to act as forwarders willconsume energy at a much higher rate. If the paths are not carefully provisioned, “popular” routes willrun out of energy before the transmission of the impulse is complete.There are two competing effects. On the one hand concentrating the data on a small number of pathsincreases the compression and reduces the energy, while on the other hand concentrating the data on asmaller number of paths increases the energy expended by those nodes and decreases the network lifetime.Our objective is to determine the optimum concentration, taking both effects into account.We will attack the problem in three phases. In the first phase, we minimize the total energy, taking intoaccount the amount of aggregation that can be performed along the paths. In the second phase, we alsoconsider the energy expended by the intermediate nodes. In the third phase, we take into accountcongestion and energy deficits and use deflection routing to move packets in directions that may besuboptimal on the average, but are preferable based on the actual use that the network has experienced. Inthe next section, we will describe the α tree algorithm, which is a response to the challenged proposed inthe first phase.α Tree AlgorithmWhen data is not aggregated or compressed, the routing structure that uses the least energy to transmit thedata from the sensors to a destination is a minimum depth tree (MDT) that is rooted at the sink, with linkweights indicating the energy needed to transmit on a link. In contrast, when the information from multiplesensors is completely redundant, only one unit of message is forwarded to the destination, regardless of thenumber of incoming messages. The tree that uses the least energy to collect the data in this case is aminimum spanning tree (MST). In most cases, the data from multiple sensors is not completely redundant,but has some redundancy. In the α tree algorithm, a single parameter α can be adjusted according todifferent levels of data redundancy in order to find routes that minimize the overall energy consumption.In reference [1] we relate algorithms that generate minimum spanning trees and minimum depth trees byjoining one node at a time to a tree that is rooted at the destination. In the minimum depth tree the nodesthat are connected to the rooted tree are labeled with the distance to the destination. In the minimumspanning tree the node labels are zero. We plan on labeling the nodes with α× (the distance to the destination) and relating α to the amount of data compression that is performed at the intermediate nodes.(a) (b) (c)Figure 1 (a) Minimum Spanning Tree (b) Minimum Depth Tree (c) α Treeα trees, with 0 ≤α≤ 1 are between the minimum spanning and minimum depth trees. How α affects the “shape” of a tree is illustrated in Figure 1. As α→ 0, we only consider the energy needed to connect the next node to the tree. The resulting tree is a minimum spanning tree, as depicted in Figure 1(a). When thedata is compressed very little additional energy is required to get the rest of the way to the destination. As α→ 1 equal weight is given to transmitting the data to the first node as along the rest of the path to the destination, which is true when there is no compression. The resulting tree is a minimum depth tree, as depicted in Figure 1(b). Intermediate values of α yield trees that fall somewhere in between the minimum spanning tree and minimum depth tree, as depicted in Figure 1(c). Our objective is to determine the α tree that best models the amount of data compression that can be performed.In this initial description we have made α a constant. We can use an algorithm that is similar to Dykstra’s shortest path algorithm to generate an α tree. We will also consider making α a function of the number of hops to the destination. When there is more than one sink for the data, we have a forest of trees. In order to find the forest of α trees we will start the algorithm with all of the destinations labeled as zero. Then, run the α tree algorithm for each sink successively. A node should switch its connection to a new tree whenever its distance to the new sink is shorter than its distance to the previous one.Preliminary Results and ImpactsTo visualize the behavior of the α tree algorithm, an α tree simulator has been built. In our implementation, the link weights are set to be (distance between two nodes)n, where n is the path loss exponent. As a result, the link weight is proportional to the transmission power for forwarding the data from one node to another. Under this assumption, the α tree algorithm can always find the optimum network topology with the minimum overall energy cost if a fixed compression ratio c is assumed at each forwarding node. The value of α in that case should be set to c, and the proof will be given in the workshop. Figure 2 shows topologies generated with α equals 0 (MST), 1 (MDT), and 0.8 respectively. In this example, 200 sensors are spread randomly over a 30 × 30 region with a sink at the center, and the path loss exponent is chosen to be four. By visual inspection, the simulation results primarily agree with our expectation. We have also calculated the energy cost of the topologies based on different values of α and n, assuming a fixed compression ratio of 0.8. Table 1 summarizes the results. The energy cost of a topology is defined as follows:(a) (b)× (Distance)nNo. of bits transmitted on a link after data aggregationNo. of bits in a messageEnergy Cost =∑all links(c)Figure 2 α tree topology with (a) α = 0, (b) α = 1, and (c) α = 0.8.Path Loss Exp. (n) α = 0 α = 1 α = 0.82 2302 2763 21183 4360 5512 41034 8892 10033 8574Table 1 Overall energy costs of α trees under the assumption of a fixed compression ratio of 0.8The α tree algorithm makes a pioneer attempt on relating data aggregation performance to the generation ofrouting topologies which minimize the total energy cost for data funneling. The performance of data aggregation is dependent on the type of application or the nature of the data. The α tree algorithm caneasily adapt to the variations in compression performances through the adjustment of a single parameter. In general, the optimum value of α decreases as the amount of data reduction increases.Work in ProgressIn last section, we mentioned that α tree works best for networks having a fixed data compression ratio ateach forwarding node. Obviously, such a model is not completely accurate. In the real world, data shouldhave temporal and spatial correlations, e.g. temperature over a region. Therefore, a more realistic data compression model should be developed. We are applying information theory to defining a generic compression model, taking into consideration possible temporal and spatial correlations. Based on therefined compression model, we will then evaluate the performance of α tree, e.g. percentage reduction ontotal energy cost with respect to node density and sensor-to-sink ratio, as compared to MST and MDT. Wewill also investigate the relationship between the choice of α and the data aggregation performances. In addition, we may use optimal routing [2] to generate optimal trees and compare these trees with best αtrees. Finally, we may study the overhead in generating α trees and find out the response of the α tree algorithm at different levels of node mobility.References[1] N.F. Maxemchuk. Video Distribution on Multicast Networks. IEEE JSAC, 15(3): 357-372, April 1997.[2] D. Bertsekas and R. Gallagher. Data Networks. Prentice-Hall, 1992.。

南京邮电大学硕士研究生学位论文术语表术语表Adaptive Threshold sensitive Energy APTEEN 自适应敏感阀值节能型传感网络协议CDMA码分多址Code Division Multiple AccessCSMA 载波侦听多路访问Carrier Sense Multiple AccessDD 定向扩散Directed DiffusionGEAR 地理和能量感知路由Geographic and Energy Routing LEACH 低功耗自适应分簇协议介质访问控制Media Access ControlMCU 微控制单元Micro-Controller UnitPEGASIS Po-Efficient Gathering in SensorInformation System服务质量Quality of Service信息协商传感协议Sensor Protocol for Information viaNegotiationTCP 传输控制协议Transfer Control ProtocolTDMA 时分多址Time Division Multiple AccessTEEN 敏感阀值节能型传感网络协议Threshold sensitive Energy Efficient sensorNetwork protocol用户数据包协议User Datagram ProtocolWSN 无线传感器网络Wireless Sensor Network南京邮电大学学位论文原创性声明本人声明所呈交的学位论文是我个人在导师指导下进行的研究工作及取得的研究成果。

尽我所知，除了文中特别加以标注和致谢的地方外，论文中不包含其他人已经发表或撰写过的研究成果，也不包含为获得南京邮电大学或其它教育机构的学位或证书而使用过的材料。

与我一同工作的同志对本研究所做的任何贡献均已在论文中作了明确的说明并表示了谢意。

无线传感器网络的能源管理方法与优化策略无线传感器网络（Wireless Sensor Networks，WSN）是一种由许多通过无线通信相连接的低功耗传感器节点组成的网络。

这些节点以无线方式相互通信并协同工作，以收集、处理并传输环境信息。

然而，WSN节点的能源供应是一个关键问题，因为它限制了网络的生命周期和性能。

因此，能源管理方法和优化策略对于延长网络寿命、提高性能和可靠性至关重要。

能源管理方法是指通过最小化能源消耗和优化能源利用来延长整个无线传感器网络的生命周期。

以下介绍几种常见的能源管理方法：1. 睡眠调度：通过管理传感器节点的睡眠和唤醒状态，可以降低整个系统的能量消耗。

根据工作负载情况，可以动态调整节点的睡眠/唤醒时间，以平衡能源消耗和网络性能。

2. 路由优化：设计高效的路由算法可以减少数据传输的距离和节点之间的通信成本。

通过优化路由路径，可以减少节点间的通信频率，降低能源消耗。

3. 数据聚合：数据聚合是将多个传感器节点收集到的数据进行合并和压缩，减少不必要的数据传输和处理。

聚合后的数据可以通过更少的传输量传送到基站，节约能源。

4. 能量回收：利用无线传感器网络中节点的能量回收技术，如光能、振动能等，来补充节点的能量消耗，从而延长网络寿命。

除了上述能源管理方法外，还有一些优化策略可以进一步提高无线传感器网络的能源管理效果。

1. 拓扑优化：通过重新部署或调整节点之间的连接关系，可以优化网络拓扑结构，减少能量消耗和通信开销。

采用最佳拓扑结构可以减少能量耗尽的风险，并提高网络的可靠性。

2. 节能硬件设计：采用低功耗、高效的硬件设计可以降低节点的能源消耗。

例如，使用低功耗的处理器和通信模块，或者采用节能的传感器设备，可以大幅提高网络的能源利用效率。

3. 数据压缩算法：使用高效的数据压缩算法可以降低数据传输量，进而减少节点间的通信开销。

通过减少数据传输量，可以大幅降低网络能量消耗。

4. 动态能量管理：通过实时监测节点的能源消耗情况，动态调整能量分配和节点的工作状态，以最大限度地延长网络寿命。

面向无线传感器网络的能量感知路由算法研究无线传感器网络（Wireless Sensor Networks，WSNs）由大量的分布式传感器节点组成，这些节点能够自主感知、采集信息并将其传输到其他节点或基站进行处理。

然而，节点的能源限制是WSNs面临的主要挑战之一。

为了延长网络的生命周期，降低能源消耗是至关重要的，因此研究面向无线传感器网络的能量感知路由算法显得非常重要和紧迫。

能量感知（Energy Awareness）路由算法是一种将能源消耗作为重要指标的路由选择算法。

它在选择传输路径时考虑节点的剩余能量、节点间的通信质量以及距离等因素，以降低网络的能耗。

下面将讨论面向无线传感器网络的能量感知路由算法的一些关键研究内容。

1. 能量感知路由算法的需求和目标能量感知路由算法的需求和目标主要包括以下几个方面：1.1 能源均衡性（Energy Balance）：在整个网络中实现节点能量的均衡消耗，避免部分节点能量过早耗尽而导致网络中断。

1.2 路径稳定性（Path Stability）：选择稳定的传输路径，减少路径的变动，降低由于路径切换引起的能耗。

1.3 距离优化（Distance Optimization）：根据节点之间的距离选择最短路径，减少能量消耗和传输延迟。

1.4 覆盖率（Coverage）：根据节点的覆盖范围选择传输路径，以保证网络的全面覆盖。

2. 能量感知路由算法的研究内容2.1 距离感知路由算法距离感知路由算法根据节点之间的距离选择最短路径，以减少能量消耗和传输延迟。

常用的距离感知路由算法包括基于最短路径树（Shortest Path Tree，SPT）的算法和基于距离向量（Distance Vector）的算法。

这些算法通过计算节点之间的距离来选择最佳传输路径，从而降低能耗。

2.2 能量均衡路由算法能量均衡路由算法旨在实现网络中节点能量的均衡消耗，避免部分节点能量过早耗尽而导致网络中断。

文章编号：1〇〇9 -2552(2017)05 -0046 -04DOI：10. 13274/ki.hdzj. 2017. 05. O il基于压缩感知的三维水下传感器网络路由算法王萌萌，李博，刘功亮，杜睿(哈尔滨工业大学（威海）信息与电气工程学院，山东威海264209)摘要：由于水下传感器网络中节点能量有限且工作环境恶劣，设计高能效的路由算法实现数 据采集尤为重要。

文中利用水下传感器网络中原始信号的相关性，将压缩感知技术与不均匀分 层多跳路由算法联合设计，提出一种基于分布式压缩感知的数据采集方案一D C S-U L M。

结果表 明，该算法在保证原始数据重构精度的同时可有效延长网络生命周期。

关键词：水下传感器网络；压缩感知；三维路由；能耗中图分类号：T N929.5; TP212.9 文献标识码：AEnergy-efficient routing algorithm in 3-dimension underwater sensor networks based on compressed sensingWANG Meng-meng，LI Bo，LIU Gong-liang，DU Rui(School of Information and Electrical Engineering，Weihai Campus，Harbin Institute of Technology，Weihai 264209,Shandong Province，China)Abstract ：Because the b a tte ry-d riv e n nodes in underw ater sensor netw orks w ork in adverse e n v iro n m e n t，the design o f e n e rg y-e fficie n t ro u tin g a lgo rithm re a lizin g data gathering is im p o rta n t.U sing the corre latio n o f o rig in a l signal in underw ater sensor n e tw o rk s，in th is p a p e r，an uneven-la y e re d，m u lti-h o p ro u tin g based on d is trib u te d compressed sensing(DC S-U L M)is proposed to achieve data c o lle c tio n.The sim u latio n results show that D C S-U LM can effe ctive ly prolong the life tim e o f netw orks w h ile it ensures thereco nstructio n accuracy o f o rig in a l d a ta.Key words:underw ater sensor n e tw o rk s;compressed se n sin g;3-dim ension ro u tin g;energy consum ptiony信息疼术2017年第5期0引言海洋是维持人类生存发展的重要基地。