Research Article An Energy-Efficient Multipath Routing ...downloads.hindawi.com/journals/ijdsn/2015/642189.pdf · Research Article An Energy-Efficient Multipath Routing Algorithm

Research ArticleAn Energy-Efficient Multipath Routing Algorithm Based on AntColony Optimization for Wireless Sensor Networks

Mengjun Tong,1,2 Yangli Chen,1,2 Fangxiang Chen,1,2

Xiaoping Wu,1,2 and Guozhong Shou1,2

1School of Information Engineering, Zhejiang A&F University, Lin’an City 311300, China2Zhejiang Provincial Key Laboratory of Forestry Intelligent Monitoring and Information Technology Research,Lin’an City 311300, China

Correspondence should be addressed to Mengjun Tong; [email protected]

Received 30 October 2014; Revised 8 March 2015; Accepted 16 April 2015

Academic Editor: Shaojie Tang

Copyright © 2015 Mengjun Tong et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

An energy-efficient ACO-basedmultipath routing algorithm (EAMR) is proposed for energy-constrainedwireless sensor networks.EAMR is a hybrid multipath algorithm, which is reactive in path discovery and proactive in route maintenance. EAMR hasimprovement and innovation in the ant packet structure, pheromone update formula, pheromone updatemode, and themechanismof multipath. Average energy consumption and congestion of path make pheromone update formula more reasonable. Incrementalpheromone update mode may easily lead to local optimum. The pheromone will be thoroughly updated when a node receives abackward ant. EAMRmakes an innovation in multipath mechanism which becomes more reasonable to multipath between sourcenode and destination node. Probabilistic routing mechanism is designed to make stream flow into network more balanced. Thesimulation results show that the proposed algorithm achieves an improvement in energy efficiency, packet delivery ratio, and end-to-end delay.

1. Introduction

Wireless sensor network (WSN) [1–4] is a new generationof sensor networks, integrating sensors and wireless networktechnology. It has very broad application prospects andwill have a tremendous impact on human life. Wirelesssensor network typically has a large number of nodes,which are energy limited and cannot be replenished. Savingand balancing energy consumption of nodes, extending thenetwork lifetime, is one of the primary design goals ofwireless sensor networks [5]. Data transmission between twonodes is typically done with conventional multipath routingalgorithms such as AOMDV through one same path all thetime. It will lead to energy exhaustion of some nodes inthe path, causing node failure and network partitioning,which seriously affects the network lifetime. To achieve betterload balancing throughout the network, packet forwardingshould be based on the current situation of the network.Routing algorithms based on ant colony optimization (ACO),

which use control packets called ants to inform each nodeof the current situation of the network and select nexthop according to the probability formula, are ideal for thedesign of load-balancing multipath routing protocols. ACO-based routing algorithms are one of the hotspots in thepresent study. On the basis of the analysis of various types ofmultipath routing protocols, this paper proposes an energy-efficient ACO-based multipath routing algorithm (EAMR)and presents a series of simulations that are conductedto evaluate the performance of EAMR. ACAD (AutomaticClustering inspired by Ant Dynamics) is a simple heuristicalgorithm that can automatically detect any number of wellseparated clusters, which may be any shape, for example,convex and/or nonconvex. Inspired by the dynamics ofants, the algorithm iteratively partitions the dataset basedon its proximity matrix. It is different from the existingant colony based clustering techniques. NISR (a NatureInspired Scalable Routing protocol for MANETs) protocolis inspired by bees and ant colonies. Ants and bees help

Hindawi Publishing CorporationInternational Journal of Distributed Sensor NetworksVolume 2015, Article ID 642189, 12 pageshttp://dx.doi.org/10.1155/2015/642189

2 International Journal of Distributed Sensor Networks

each other to find food sources, update quality of paths tothese food sources continually, and determine pheromoneof paths. PACONET (an improvised ant colony optimizationalgorithm for MANETs) uses two kinds of mobile agents,FANTs and BANTs. FANTs are transmitted in a controlledbroadcast manner to determine new routes. BANTs establishthe path based on the information gathered by FANTs.Thesemobile agents leave certain amounts of pheromones at thetime they depart from the node.

2. ACO and ACO-Based Routing Algorithms

Ant colony optimization [6] sources from the optimizationmode of ant foraging. Ant colony system (ACS) is a dis-tributed biological system. By collaboration, the ants cancomplete the arduous task that a single individual is incapableof completing, which is themanifestation of biological swarmintelligence. When ants leave the residence to find food,they release the chemicals called pheromones on the path.The pheromone is volatile. Shorter paths can be completedquicker and more frequently by the ants and will thereforebe marked with higher pheromone intensity. These pathswill then attract more ants, which will in turn increase thepheromone level, until there is convergence of themajority ofthe ants onto the shortest path. Ultimately, ants can find theoptimal path through the cooperation. Routing in wirelesssensor networks follows the same principle. The node whichneeds to send data packets releases the ant-like packets to thedestination node, and the ant-like packets are returned fromthe destination node, forming a path to the final destinationnode [7–9].

AntNet [10] is developed according to the principles ofACO. It is also one of themost successful ACO-based routingprotocols by far. In AntNet, the concepts of the forward antsand the backward ants are presented.The forward ants choosethe next hop randomly according to the heuristic informationvalues in the routing tables. And the ID of the node passedwill be appended to the head of the ant. All the forwardants are converted to the backward ants as soon as theyarrive at the final destination. The backward ant travels backto the source node through the reverse route and releasespheromones on each link passed by.

ARA [11] is the earliest on-demand multipath algorithmthat applies ant colony algorithm to ad hoc wireless networks.Routing discovery relies on forward ants and backward ants.In routing discovery, ARA broadcasts forward ants whichonly carry a unique sequence number. If a node receives aforward ant that it has never got, it sets up a reverse path andrebroadcasts the ant to the neighbors. On the contrary, if thenode has received a duplicate ant, it will drop the ant. By thisway, only one path can be formed to the destination. Whena forward ant reaches the destination node, it is convertedto a backward ant and returns following the reverse path. Ifan intermediate node receives the backward ant, it creates apath to the destination node (including next hop, destination,and pheromone) and then continues transmitting along thereverse path. Intermediate nodes set up the correspondingrouting tables instead of dropping same backward ants.

The multipath to the destination node can be formed. Theupdating of the pheromone relies on data packet and time-setting volatilization. No other types of packets are required;thus overhead is reduced.

In ARAMA [12], when a node needs to establish ormaintain a path to the destination node, it sends a forwardant to a neighbor node rather than flooding. Intermediatenodes’ IDs are appended to the forward ant. What is more,path information of the forward ant (such as hops, remainingenergy, bandwidth, and queue length) is also appended orchanged. ARAMAdefines the concept of the grade.The valueis calculated by the backward ant and saved in nodes. Theformula of the grade relies on the link information such asenergy. When an intermediate node receives a backward ant,the pheromone is updated according to the path gradient ofthe ant. Pheromone of link which is passed by the backwardants is increased, and the other link pheromone volatilizes.The purpose of volatilization is to make nodes forget theold path quicker. The backward ants are deleted when theyreach the source node.The data transmits along the best path.When the best path is destroyed, another path can be used tosend data packets immediately.

Di Caro et al. proposed AntHocNet protocol [13].AntHocNet is an ACO-based multipath hybrid routing pro-tocol.The protocol is reactive in path discovery and proactivein route maintenance. The routing algorithm has four majorphases: reactive route establishment, random data routing,proactive path maintenance and exploration, and link failurehandling.

When a node which does not have route information tothe destination node needs to communicate, it broadcasts aforward ant with all its neighbors for path discovery. Eachnode’s routing table has a pheromone table. If the table haspheromone information to the destination node, the forwardant chooses the next hop based on pheromone informationin the tables.

The forward ant carries a list which saves the nodes thatit has passed. The forward ant changes into a backward antwhen it arrives at the destinationnode.Then the backward antreturns to the source node following the path the forward anthas passed. During the process of returning, the backward antupdates the pheromone table and routing table of the nodesthat it passes, including delay, hops, and other information inthe tables.

Data packet begins to be sent after the path has beenbuilt from the source to the destination. When an interme-diate node has multiple routes to the destination node, itstochastically gets the next hop according to the values inphenomenon table.

The advantage of AntHocNet is that it finds multipathin path discovery so that it can reduce frequency of pathdiscovery. And the phase of route maintenance requires a lotof ants. In addition, each node holds a routing table whichrecords all the destination nodes that the node can reach. It isnot suitable for large scale networks like AntHocNet.

Rosati et al. [14] proposed the DAR (Distributed AntRouting) protocol. DAR is a reactive ACO-based routingprotocol. Unlike proactive routing, reactive routing protocolneed not send the forward ant regularly. When there is

International Journal of Distributed Sensor Networks 3

no data to transmit, it can reduce cost and the energyconsumption of nodes. In the DAR protocol, the forwardants only care about the nodes in cross and choose next hopby only using the pheromone information. The backwardants only release a constant value of flavor in the returnlink. Nodes stochastically choose the next hop according tothe phenomenon when it sends data packets. In the DARprotocol, ants save all the passed nodes’ IDs, so it is suitablefor small networks. The network’s convergence is slow andsometimes also leads to local optimum.

Wang et al. [15] proposed HOPNET protocol which is ahybrid routing protocol. This protocol divides the networkinto a plurality of regions and adopts proactive routing inregions and reactive routing between regions. Because thearea of every region is not large, the cost of internal proactiverouting is not great. HOPNET has more advantages in largenetworks.

In [16], Misra et al. proposed EAAR protocol based onAntHocNet. Minimumhops count and energy path are takeninto account in pheromone update formula. The study ofDucatelle et al. [17] indicates that pheromone update formulais not rational by only considering hops, because fewer hopsmay cause the longer distance between nodes, and signalintensity may be too low. The connection may be loose andeasily leads to network partition. In addition, the fartherdistance between the nodes can also cause more energyconsumption during the data transmission [18].

In [18], it proposes a pheromone diffusion mode of theant colony algorithm based on routing protocol DBACRA.The protocol is divided into two types, the actual and virtualpheromones, which guide the ant packet and data packetto the path search, when the actual pheromones of thebackward ants are from the destination node, releasing thelink of the pheromone. Data transmission also needs theactual pheromone to be completed. Virtual pheromones arespread by the destination node; the whole network formsthe preliminary virtual pheromones by the certain amount oftime.They can guide the forward ants to reach the destinationnode. The forward ant can guide the actual and virtualpheromones to reach the destination node. As a result ofthe virtual pheromones being propagated by the diffusionmethod, in a mobile or relatively large network, the virtualpheromones are also very easy to fall into the loop.

Camilo et al. [19] proposed EEABR (energy-efficientACO-based routing) protocol. EEABR is an energy-efficientACO-based routing algorithm. The forward ant’s head of theEEABR protocol can save nodes’ IDs with the two recenthops which reduces the length of the forward ant packet,saves the energy consumption of nodes, and prolongs thelife of wireless sensor networks. But the EEABR uses thesame packet structure of the forward ant and the backwardant which increases the unnecessary energy cost. EEABRregularly sends forward ants in proactive routing mode. Inthe absence of the data source mode, it reduces the energyconsumption of the whole network. At the time of routingmaintenance, EEABR not only sends the forward ant by theregular method, but also makes the extra overhead of theprotocol. What is more, more packets are forwarded to thepath, causing the network energy consumption imbalance

and reducing the life-span of the network. In the EEABRprotocol, the intermediate nodes discard the ants similar tothe forwarded ants, so it cannot form effectivemultiple paths.

Some of the multipath routing protocols mentionedbefore do not consider the energy of network. If we followthe traditional ant colony algorithm to find the optimalpath, all traffic will be almost focused on a path and itwill lead the nodes on this path to die soon. So this typeof protocols cannot be directly applied to WSNs. Someprotocols, although taking into account the remaining energyon each path, do not consider the rate of the energyconsumption. Other protocols do not take into account thenetwork congestion. Considering factors mentioned beforeand combined with ACO, an energy-efficient ACO-basedmultipath routing algorithm (EAMR) is proposed. EAMRhasimprovement and innovation in pheromoneupdate formulas,pheromone update mode, and the mechanism of multipathestablishment.

3. Energy-Efficient ACO-Based MultipathRouting Algorithm

Based on the limited energy characteristic of WSN, anACO-based multipath routing algorithm is proposed in thispaper. EAMR uses a new multipath mechanism. In thepheromone update formulas, EAMR takes into account theenergy consumption rate of path, the remaining minimumenergy of path, the hops from sink, and the congestionstatus of path. When a source node which does not haverouting information to the destination node needs to senddata packets, it broadcasts forward ants to the destinationsink node. Sink node generates corresponding backwardants, which will travel along the forward path back to thesource node. Backward ants release pheromones on the pathwhen they move back to the source node. In traditionalway of pheromone updating, pheromone is more and moreaccumulated on a path so that more and more data packetswill be sent on this path. As a result, network cannotautomatically balance load. Different from the traditionalincremental pheromone update mode, however, pheromonewill be thoroughly updated when a node receives a backwardant in EAMR. When a node has multiple paths to thedestination node, it will stochastically select one of them inaccordance with pheromone to the sink node.The probabilis-tic routing strategy leads to data load spreading according tothe estimated quality of the paths. If the probabilities are keptup-to-date, this will lead to automatic routing load balancing.

3.1. Path Discovery. EAMR is a mixed multipath routingalgorithm. It is reactive in the route establishment phase.After route establishment, EAMR is a proactive routingalgorithm. When a source node, 𝑠, needs to communicatewith the destination node, 𝑑, and it does not have routinginformation about 𝑑, it broadcasts a reactive forward ant, say𝐹𝑑

𝑠(here an ant is representative of a control packet). Due to

the initial broadcasting, each neighbor of 𝑠 receives a replicaof 𝐹𝑑𝑠, say 𝐹𝑑

𝑠⋅ 𝑘 (the notation “𝑘” refers to indexing; the 𝑘

replica of a single broadcast is represented as 𝐹𝑑𝑠⋅ 𝑘). After


s

d

(a)

s

d

(b)

Figure 1: The paths of forward ant (a) and backward ant (b).

the next hop, the next neighboring node will receive 𝐹𝑑𝑠⋅ 𝑘 ⋅ 𝑙

(𝑘, 𝑙, . . . are integers). As these broadcast ants come from thesame source node and generate at the same moment, we callthem similar ants. Similar ants have the same source nodeaddress and sequence.The task of each ant𝐹𝑑

𝑠⋅𝑘⋅𝑙⋅𝑚⋅𝑛 . . . is to

find a path connecting 𝑠 and 𝑑. At each node, an ant is eitherunicast or broadcast, according to whether or not the currentnode has routing information for 𝑑. If current node does nothave routing information for 𝑑, it broadcasts a forward ant.Otherwise, if routing information is available, the forward antwill choose its next hop 𝑗 with the probability 𝑃𝑖

𝑗𝑑:

𝑃𝑖

𝑗𝑑=

𝜂𝑖

𝑗𝑑

𝛽1

∑𝑘∈𝑁𝑖𝜂𝑖

𝑘𝑑

𝛽1 , (1)

where 𝜂𝑖𝑗𝑑

is the pheromone of link (𝑖, 𝑗) in the routingtables whose destination node is 𝑑. 𝜂𝑖

𝑗𝑑will be described

in detail later in this section. 𝑁𝑖 is the set of neighbors onnode 𝑖 in routing tables, through which node 𝑖 can reach thedestination node. 𝛽1 is a factor which is used to adjust theimportance of pheromone to forward ants choosing next hop,which is greater than or equal to 1.

When a node receives the first forward ant, it forwards theant and saves some information about the ant in the access-list (access-list structure in Section 3.4). If the ant receivedis not one of the similar ants of the first forward ant, thenode could receive the ant and forward it. As a special case,if the node is a neighbor of the source node, it only acceptsthe forward ants that are directly from the source node andrejects any other ants. Therefore, when the node receives asimilar ant, it has to carry out the loop detection at first,that is to say, the current node looks into ant’s head andsearches the identification of itself. If there is a loop, the nodedrops this ant directly; otherwise the similar ant should followthe multipath mechanism to establish multipath between thesource node and the destination node. Steps are elaborated asfollows.

First, make a judgment for whether the first hop ofreceived similar ant and the existing similar ant in access-list is same. First hop is the first node the ant arrived at afterleaving source node. If it has no similar first hop, current nodeshould compare newly received ant’s delay𝑑new and hops ℎnewwith similar ant’s delay 𝑑sim and hops ℎsim in access-list. If𝑑new < 𝜆𝑑 ∗ 𝑑sim and ℎnew < 𝜆ℎ ∗ ℎsim, accept the newforward ant. In this paper, 𝜆ℎ is set to 1.5 and 𝜆𝑑 is set to 2.5.The arguments can be adjusted according to different networkenvironments.

Second, if it has similar first hop, EAMR makes ajudgment for whether the last hop of received ant and theexisting similar ant in access-list are the same. If it has nosimilar last hop, as soon as 𝑑new < 𝜆𝑑 ∗ 𝑑sim and ℎnew <𝜆ℎ∗ℎsim, EAMRaccepts the new forward ant. In this paper,𝜆ℎis set to 1.5 and 𝜆𝑑 is set to 2.5.The arguments can be adjustedaccording to different network environments.

Third, if both first hop and last hop of newly received antare the same as the similar ant saved in access-list, EAMRaccepts the new forward ant only when ℎnew < ℎsim.

Through this method, multiple paths can be obtained. Ifa path fails, another path can be used. Multipath contributesto energy and load balancing of the network. Through thismethod, the forward ant can arrive at the destination node.WSN environment is complex due to network congestionand delay issues. In order to reduce the unnecessary cost,some ants that reach the destination node “too late” will bedropped, and the destination node does not send backwardants accordingly. When the destination node receives aforward ant and has never received its similar ants before,the destination node starts a waiting timer. Those ants whoarrive after waiting timer timeout will be dropped and donothing. Waiting timer timeout interval is calculated by (2).We can see in Section 3.4 that the structure of forward antscontains the Src time field, which records the time when antsleave the source node, so that the end-to-end delay 𝑇𝑠−𝑑 willbe very easy to get. 𝐶𝑑 is a parameter which can be adjustedappropriately according to the network environment:

𝑇𝑑 = 𝐶𝑑 ×𝑇𝑠−𝑑. (2)

During the 𝑇𝑑 time interval, as shown in Figure 1, whenthe destination node receives a forward ant, it will convertthe ant into the corresponding backward ant, which will goback to the source node through the route the same as theforward one did. If an exception occurs with the next hopof the backward ant, such as change of the location of thenode, then the ant will be killed. When a backward ant fromnode 𝑗moves to node 𝑖, the pheromone of node 𝑖 needs to beupdated; node 𝑖 should update or create routing informationto the destination node, according to the following rule:

𝜂𝑖

𝑗,𝑑=

Min𝑘1

(ETX𝑘2 ∗ 𝑇𝑘3𝑠−𝑖∗ 𝐸𝑘4avg), (3)

where ETX [20] (the expected transmission count metric)of a link is the minimum expected pathway which the totalnumber of data transmissions (including retransmissions)


required to successfully send a data packet to the destina-tion. ETX which incorporates the effects of link loss ratios,asymmetry in the loss ratios between the two directionsof each link, and interference among the successive linksof a path is based on packet delivery ratios, which findshigh throughput of multihop paths. The ETX of a path isestimated as the accumulation of ETX values over all linksconstituting the path, which minimizes the transmissioncost. Updating of the routing ETX is compared with otherrouting, which chooses the value of ETX to be the least asthe best route, sends the information to the routing module,and is recorded in the routing table. The ETX can decreasethe energy consumed per packet. 𝑇𝑠−𝑖 is the delay from theforward ant leaving source node to the backward ant arrivingat the current node. 𝐸avg is the average energy consumptionon a path from the source node 𝑠 to the destination nodeand can be obtained from the backward ant packet. Min isthe energy value of the minimum energy node on the paththat backward ant passes.𝐸avg shows the energy consumptionconditions on the path. The larger value of 𝐸avg indicatesthat more energy has been consumed on the link and moredata packets have been transmitted on the path. Hence thenodes on the path should be distributed less load traffic infuture data transmission.Min indicates the energy bottleneckof a path. Paths that have energy bottleneck should be alsodistributed less load traffic in future data transmission. Thishelps extend the life of the nodes and the network. Sometimesa node’s death will lead to the segmentation of the entirenetwork, especially in the case of mobile nodes, where anode may move to the junction of two parts of the networkand die early, leading to the segmentation of the network.ETX value of a route which is the sum of the ETX foreach link is the link count which the successful transmission(including data retransmission) of a packet required fromcurrent node 𝑖 to the destination node 𝑑. According to thefeedback mechanism through the MAC layer, if the data getsACK response information, it will successfully deliver to thedestination node. Otherwise the source node has to resendthe data packets. When the retransmission reaches a certainthreshold, it will discard the packet and the MAC layer willinform the upper network layer that the error occurred inthe process of transmitting the data packets. It is a metricbased on link capacity; this parameter can not only considerthe length of the path, but also consider the packet loss rate.Packets should reach the destination node along the pathwith the ETX of a link as small as possible. The fewer theETX of the link counts are, the more pheromones will bereleased on the link, and packets are more likely to choosethis link. What is more, it will shorten the time delay of thedata transmission. If the next hop chooses the ETX valuewhich is smaller than itself, so as not to form a loop, 𝑇𝑠−𝑖 isobtained by current timeminus Src time in the backward antpacket (Src time will be described in Section 3.4). In general,the path with small transmission delay is better than the bigdelay path which may be at the congestion state because ofrelatively large load. According to (3), pheromone is releasedless in longer delay link. 𝑘1, 𝑘2, 𝑘3, and 𝑘4 are coefficientsthat can be adjusted appropriately according to the networkenvironment.

3.2. Data Session. Each node establishes the routing infor-mation by ants. Once backward ant returns, the data storedin the buffer can be sent out. The node that needs tosend data packets should choose the next hop from a setof neighbors with the probabilistic routing strategy. Theprobability formula is defined as

𝑃𝑖

𝑗𝑑=

𝜂𝑖

𝑗𝑑

𝛽2

∑𝑘∈𝑁𝑖𝜂𝑖

𝑘𝑑

𝛽2, (4)

where 𝛽2 is a factor which can adjust pheromones onprobabilistic routing selection. The greater the value of 𝛽2is, the more likely it is to choose the link which has morepheromones as next hop. It can be adjusted appropriatelyaccording to the network environment. In this paper, 𝛽2 is aninteger value not smaller than 1.Thepaths that havemore loadtransmission tend to have longer delay and less remainingenergy, and through the automatic updates of pheromones,data packets will automatically be sent to the path which hasmore energy and no congestion. Finally, the entire network isload balanced.

Pheromone evaporation occurs periodically. For every 𝜏time period, nodes evaporate the pheromone value automat-ically. Equation (5) is the pheromone evaporation formula:

𝜂𝑖

𝑗,𝑑= 𝜂𝑖

𝑗,𝑑× (1−𝜌) , (5)

where 𝜌 lies between 0 and 0.05.

3.3. Route Maintenance and Link Failures. Each node needsrouting maintenance during the data transmission. EAMRuses the way to send forward ant regularly for routingmaintenance. When a node begins to send data packets, itproduces the forward ant regularly. The interval of sendingforward ant is based on the speed of the node movementand network condition. In this paper, interval is set for 5seconds. But whether to send the forward ant needs to takeinto account the pheromones in each link of the routing tablesand the length of access-list. If the maximum pheromone ofall links is less than a threshold and the length of access-listis less than limit value, forward ants could be sent. It canguarantee that forward ants are asynchronously sent in thenetwork. Similarly, when source node receives a backwardant, it can also decide whether to send forward ant accordingto the above method. If the maximum pheromone exceedsthe threshold or the length of access-list is greater than upperlimit, it should reset send timer. In this paper, pheromonethreshold is set to 30% of original value.The length of access-list upper limit can be set based on network scale and nodesmemory size. Upper limit is set to the 20% of total number ofnodes in the experiment of this paper.

In order to avoid nodes in the same path sendingduplicate forwarding ants, when a node receives a forwardant, it needs to reset the forward ant send timer. As a result,the node which is transmitting data between source nodeand destination node forwards the ant it received and delayssending the forward ant which is produced by itself. By thisway, the node can reduce the number of data packets it sends,thus reduces the cost of network.


Type Src_address Seqno Visitednode Esum TTLSrc_time

Figure 2: Packet structure of forward ant.

Type Visitednode Src_time Emin Eavg ETX

Figure 3: Packet structure of backward ant.

Forward ants have two ways for sending, unicasting,and broadcasting, during route maintenance. Unicasting aforward ant whose next hop is decided by (1) only updates thepheromone of the existing path. There is 10% probability tobroadcast forward ant.There are two reasons why EAMR hasto send broadcast ants: on the one hand, if a unicast packetwas sent, it only updates the pheromones which alreadyexists but does not explore new paths (especially under thecondition that the nodes had moved or the energy wasconsumed), this may cause the original routing fail to work.On the other hand, broadcasting ants can explore new pathsand avoid falling into local optimum. When a broadcastingant reaches the neighborhood and the neighbor may have norouting information at the moment, then the neighbor nodewill continue to broadcast the ant. It will soon cause the floodin the entire network and it is costly. So we are limited to thebroadcasting ants. TTL of broadcasting ant is set to 2 (thisvalue can be changed according to the environment) in thispaper. That is to say, if the ant does not find a new path aftertwo hops, it will be killed. Because of taking into account thefact that the new path is near the current path, EAMR usesthis hop restriction mechanism.

The environment of WSN is complex, so the node in thedata transmission processmay lose connectionwith neighborand cause link failure. Like many other protocols, EAMRprotocol also adopts the Hello mechanism, sending Hellomessage to neighbors regularly. It is easy to find whetherthe neighbor is reachable and judge whether the link to thisneighbor has failed. Of course, it is necessary to inform theneighbor initiatively when the node itself has a fault andmakes neighbor nodes find the link failure in time.

When the node finds a link to a neighbor node thathas fault, the node should delete the neighbor’s informationfrom its neighbor table and routing table. Before deleting,the node should judge whether the removed neighbor is theonly next hop node to the destination. If it is the only nexthop, the node should send link failure notification to otherneighbors. Even though the neighbor node is not the onlynext hop to the destination node, the route is the best routeby viewing the pheromones. The node should also send linkfailure notification to other neighbors. During the forwardant transmission, the current node needs to save “last hop,”by which the ant sends to current node. If node needs to sendlink failure notification, it only sends back to the “last hop”without broadcast. It can reduce the cost. The node receivingthe link failure notification processes as the same method.

When a node knows the link failure by sending packetfailure and there is no other path used to send data packets,

the node should start to repair route by sending repairingants. The mechanism is similar to AODV and the distinctionis that, in order to reduce the energy consumption, therepairing ants are not broadcast but sending several repairingants according to the number of neighbors. Repairing antshave maximum hops limit (hop limit is 3 in this paper). Thenode starts timer at the time of sending the first repairingant. If it does not receive the backward repairing ants at theend of the timer, it is obvious that there are no paths to thedestination node. As a result, the node has to discard cachedata and send link failure notification to its “last hop.”

3.4. Packet Structure of EAMR. In the packet structure ofEAMR, the forward ant needs to save all the nodes’ IDs whichit has visited in Visitednode list. The list acts as a blacklistwhen selecting next hop by using probability formula inneighbor table. That is to say, only the nodes which existin neighbor table and do not exist in the Visitednode havethe opportunity to become the next hop. The Visitednodefield increases the packet length of the forward ant butavoids a loop. The backward ant also followed the nodes inVisitednode list to return to the source node.

Figure 2 is the packet structure of the forward ant. Typerepresents the type of packet. Src address is the sourcenode’s address which produces the forward ant. Seqno isthe sequence when the source node generates the forwardant. Every forward ant has a unique sequence number.The node which receives the forward ant can use flag⟨Src address, seqno⟩ to determine whether it receives asimilar ant. Esum is the sum of energy consumption of theVisitednode by the ant. Esum is used for pheromone update.The Src time field records the timewhen ants leave the sourcenode. TTL reflects ants’ life-span in the network. It has twomain functions; first, it can limit the range of ant search;second, it prevents ants unrestricted in the network circlesfromwasting source of network.TheVisitednode field is usedto save all the nodes’ IDs which the forward ant has visited.

Forward ants die out after reaching the destination node,and those which meet the requirements can be transformedinto a corresponding backward ant. Figure 3 suggests thepacket structure of backward ant. Eavg is average energy con-sumption in a path from the source node 𝑠 to the destinationnode. Esum is divided by the node count of Visitednode toget the Eavg which is calculated by destination node. Eavgis used for pheromone update in (3). Emin is energy valueof the minimum energy node on the path that backward antpasses. Src time is the time when corresponding forward antleaves the source node. Visitednode is the path which the


Src_address Seqno Src_timeF_hop TTLL_hop Hops

Figure 4: Structure of access-list element table.

Nei_addr Nei_energy Hops Pheromone Last_update_time

Figure 5: Structure of the neighbor table.

forward ant passes and gets from the forward ant. Backwardantswill go back to the source node through the path the sameas the forward one did and update pheromone according toformula (3) in link. There are differences between backwardants and forward ants that the backward ants have no TTLfield, because the backward ants have already carried thepath information Visitednode. If backward ants’ next hopis unreachable, the backward ants will be dropped. ETXof a path is estimated as the accumulation of ETX valuesover all links constituting the path. Using ETX for pathselectionminimizes the transmission cost and achieves a highthroughput.

If an intermediate node receives a forward ant whichmeets the requirement of forwarding, it should save theant in access-list. Figure 4 is the element table structureof access-list. Each element table corresponds to an ant.The combination of marker ⟨Src address, Seqno⟩ is used todetermine whether packet is the similar ants. F hop is thefirst hop from source node to current node. L hop is thelast hop from source node to current node. The two marksare mainly used for multipath rule of similar ants. Access-list does not save the whole Visitednode of the forward ant;it can save limited resources of sensor nodes. Hops are thehop count of the source node to the current node. Src time’smeaning is similar with that of the structure of forward ants,which recorded the time when the ant leaves the source node.Hops and Src time are all used for multipath rule of similarants. TTL is the survival time of the element table. When anelement table is timeout, it will be deleted.

Figure 5 is the structure of neighbor table which savesthe neighbor’s information and pheromone to the destinationnode. Nei addr is the address of neighbor. Nei energy isneighbor’s current energy. Hops are the distance of destina-tion node to the neighbor node and hop is the unit. Hops’initial value is a constant BIGGEST HOPS when neighbortable is initialized in this paper.The value is set to 9999 whichmeans the neighbor cannot be used as the next hop for packetforwarding. When the node receives a backward ant, thenode updates the value of hops and releases the pheromoneaccording to formula (3) on the link. Releasing pheromonemeans updating the pheromone in Figure 5. Pheromone isto vaporize at regular intervals. The initial value of thepheromone is zero. If it finds there are no pheromones in allthe links when a node receives a forward ant, it will broadcastthe forward ant. Last update time is the time that updates theneighbor node table recently.

Figure 6 is the structure of Hello packet. Src addr is theaddress of neighbor that sends Hello packet. Node energy is

Type Src_addr Node_energy

Figure 6: Structure of Hello packet.

the current energy of neighbor. Nodes maintain the neighbortable by sending Hello packet periodically to each other.Hello’s sending interval can be different according to thedifferent scenes. Hello interval can be smaller if the mobilenode moves fast. In this paper, Hello interval is set to 2seconds. If it does not receive the neighborHello packet morethan 6 seconds, the neighbor information should be removedfrom the neighbor table.

3.5. Algorithm of EAMR. See Algorithm 1.

4. Simulation Results and Discussion

In order to test the effect of algorithm applied to WSN,this experiment made a comparative analysis of EEABR,EAMR, AOMDV [21], and AntHocNet. NS2 is open sourcesoftware and an effective simulation tool; it is the mostlyused simulation for studies on network. In the simulationenvironment, all levels can be controlled by the experimenterin the network structure. Users only need to concentrateon the basic element in the experiment, for the rest ofthe part can be “transparent.” Result can be achieved byconfiguring the environment parameters of the ideal networkenvironment and can be real-time tracking and record thekey nodes of important information. In order to get real-timeinformation about network performance evaluation, experi-ment follows the reproduction of some special circumstances.It is difficult to do it in the real network. NAM tool in NS2can demonstrate the operation of the network animation;users can visually see the network protocol, can understandall kinds of environments or other factors on the shadowof the network, and also can be compared to demonstratethe advantages and disadvantages of various methods. Thetrace of NS2 object in the process of communication can berecorded in the specific event at the trace files; users can usethe data processing tools in NS2 gawk to statistical data tracefiles, analysis throughput, time delay, packet loss rate, andother parameters. The simulation platform is ubuntu 10.04 +NS2.34. We completed the EAMR source codes which matchto integrate the NS2, including EAMR packet structuresand EAMR agent class, and added the EAMR source codes


Input:The following blank tables of all nodes are input:

(1) The neighbor table: A table containing all nodes in the neighborhood of a node.(2) Routing table: A table containing the next hop to transfer packets.(3) Access-list: A table containing all distinct paths initialized by source node.Initial pheromone for all nodes = 0.

Output:Update tables with all the values required to transmit data.Pheromone value which takes into account the energy consumption rate of path, the remaining minimum energy path,

the hops from sink and the congestion status of path for selected nodes.Steps:

(1) At its initial stage, nodes should create a connection with their neighbors through broadcasting of Hello packet.Initialize the access-list of every node as NULL.

(2) a source node which needs to transfer data checks in the routing table:if the source node exists which there is routing information to destination node:the forward ant is sent to next step according to the routing information.

elserebroadcast the forward ant and add the sent forward ant into the access-list.

(3) if an intermediate node is the neighboring one of source node:accept the forward ant which sends form the node.if the intermediate node receives a forward ant which the value of TTL is 1:discard it.

if what the node received is not a similar ant:add the ant into the access-list.if the ant exists which there is routing information to destination node:

choose the next step according to the formula (1).else

broadcast a forward ant.elseif the ant which creates a loop:

discard it.elsefollow the multipath mechanism in Section 3.1.

(4) if the destination node received forward ant for the first time:discard it, create the backward ant which will go back to the source node through the route as same as theforward one did and the destination node starts a wait timer.if the ants arrive within the time:

change into backward ants and be back.else

discard them.(5) if the intermediate nodes receive a backward ant:

delete the comparative ant information which it saved in access-list.if the intermediate node is not the neighboring one of destination node:calculate the value of pheromone based on formula (3) and release it on link.

elsesend packet to the destination node and needn’t to release pheromone.

(6) When the source node receives the first backward ant, the pheromone should be released on link right away. And alsothe routing table to destination node should be created. What’s more, the source deletes the forward ant information it saved inaccess-list. Meanwhile, Data transmission is initiated with each packet. the data saved in cache is sent out. The data packet willbe chosen next hop based on formula (4). If the nodes are the neighboring ones of destination node, it’s unnecessary to chooseby formula, it could send packets to the destination node directly.

(7) On link failure, Step (1) is repeated from the node that has’t route to send data packet.

Algorithm 1: EAMR algorithm.

into the NS2 simulation software. In accordance with theAntHocNet protocol functions, we also realized this protocolfor NS2 to make a comparison with EAMR performance.

4.1. Simulation Scene and Factor Setting. The mobile scenefile is created by the CMU tool set, sink node on the central

location of network. The factor setting of mobile scene isshown in Table 1. CBR is selected to be the scenario of Dataflow, which is created by the CMU tool cbrgen.tcl. The factorsetting of data stream is shown in Table 2.

In order to facilitate study of the protocol improvementin energy consumption, in this paper, the first-order ratio


Table 1: Factor setting of mobile scene.

Factor ValueNodes number of network 100Range of movement 1500m ∗ 1500mNode communication range 250mNode moving speed 10, 20, 30, 40, 50m/sMaximum pause time 0 sSimulation time 100 s

Table 2: Factor setting of data stream.

Factor ValueBandwidth 2MbpsType CBRNumber of CBR sources 20Packet sending rate 10 packets/s (40Kbps)Data packet size 512 byte

Table 3: Factor setting of sensor nodes.

Factor ValueType of channel Channel/wireless channelMAC type Mac/802 11Network interface type Phy/WirelessPhyInterface queue model Queue/drop tail/Pri queueTransmission of radio Propagation/two-ray groundEnergy model Energy model

Routing protocol AOMDV, EEABR, EAMR, andAntHocNet

Antenna type Antenna/omniantennaNode initial energy 2 JSink node energy 100 JLength of the network interfacequeue 50

Table 4: Factor setting of EAMR.

Factor Value𝛽1 1𝐶𝑑 2𝑘1 1𝑘2 1𝑘3 1𝑘4 1𝛽2 1𝜏 10𝜌 0.05

model [22] is selected for the energy transfer model. Thefactor setting of sensor nodes is shown in Table 3. The factorsetting of EAMR algorithm is shown in Table 4.

4.2. Performance Analysis. The following performance met-rics were used for the sake of comparison of our proposed

20

30

40

50

60

70

10 20 30 40 50Node speed (m/s)

Pack

et d

eliv

ery

ratio

AOMDVEEABR

AntHocNetEAMR

Figure 7: Packet delivery ratio.

protocol, EAMR, with the benchmark protocols discussedearlier: delivery ratio, delay, routing overhead, and energyefficiency. Several simulation experiments were performed.Some of the key results obtained are reported below.

The relationship between the packet delivery ratio offour kinds of protocol and maximum movement velocityof nodes is shown in Figure 7. Besides AOMDV, the otherthree protocols for data transmission use the mechanism ofmultipath probabilistic routing, and the delivery ratio of thethree protocols is higher than AOMDV. EEABR performsbetter than AOMDV, which is designed for the packetdelivery ratio. As we introduce mobility in the environment,the performance of these protocols decreases substantially.AntHocNet uses the mechanism of multipath probabilisticrouting, but the packets will be more likely to send to thelink which has more pheromones. With data packets sent,pheromones will be increased on the link and then a “good”path will come to many load streams.Thus, it will cause locallink busy and increase packet loss probability. So AntHocNetis not as good as the EAMR in the slowly moving scene.EAMR uses a better multipath mechanism; pheromone willadjust with current energy and load condition of the path intime, and the data will be more balanced injected into eachpath of the network. It leads to automatic load balance innetwork at last. It will be more reasonable for EAMR whenthe data flow is busy and node movement speed is fast. Inaddition, themechanismof link failure recovery of the EAMRis more reasonable. As a result, EAMR outperforms the otherthree protocols when the routing problems arise. In themoving scene, EAMR protocol shows a better performancein packet delivery than the other three protocols.

Figure 8 shows the relationship between the averagedelays on end-to-end with nodes’ maximum moving speed


AOMDVEEABR

AntHocNetEAMR

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

10 20 30 40 50Node speed (m/s)

End-

to-e

nd d

elay

(s)

Figure 8: End-to-end average delay.

in four protocols. Obviously, in the aspect of average delay,EEABR, AntHocNet, and EAMR are much better thanAOMDV. AntHocNet is more likely to send data packetsto delay link of the minimum. AntHocNet has better finalresults compared to the other two protocols when packetsending rate is slow. When the node moves faster, EAMRpresents the better results. On the one hand, the multipathmechanism of EAMR has completed more paths to thedestination node than the other two protocols. Pheromone ineach link will adjust with current energy and load conditionin time, resulting in load balancing. On the other hand, theEAMR will use a mechanism of link failure recovery. Whenthe routing problems arise, it can quickly resume routing andreduce the packet sending blindly and reduce the time delaycorrespondingly. When a node moves faster, unreachablepossibilities of the next hopwill increase, and the data packetsdropout and routing reconstructs will be increased. In thiscase, the delay of all the four protocols has tendency toincrease.

ACO-based routing protocols show a good performancein terms of delivery ratio and delay, but such a goodperformance is increased overhead for the price. In order toadapt to the constantmoving and location changing of nodes,EAMR sends the forward ant regularly in route maintenance,through the backward ant update pheromone on paths.Whatcan be seen from the preceding simulation results is thatEAMR improves data delivery ratio and reduces the trans-mission delay, but the regular ant transmission also increasesrouting overhead. Both AntHocNet and EEABR also regu-larly send ant package in routemaintenance. Inmany aspects,the routing overhead of EAMR is less than AntHocNet.These aspects include the method of forward ant sendin route maintenance, multipath mechanism, pheromone

AOMDVEEABR

AntHocNetEAMR

5

10

15

20

25

10 20 30 40 50Node speed (m/s)

Rout

er co

nsum

ptio

n

Figure 9: Routing overhead.

update mechanism based on energy, and improved linkmaintenance mechanism. As Figure 9 shows, the routingoverhead of AOMDV outperforms other protocols.

As Figure 10 shows, the definition of average energyconsumption is that the average energy is consumed whenthe final destination node accepts a packet. The smaller theaverage energy is consumed, the higher the energy efficiencyof the network will be. EEABR takes into account the variousfactors such as the power consumed in transmitting a packet.EEABR, being a multipath energy-aware routing protocol,shows the better performance compared to the past ACO-based existing protocols. The multipath mechanism and themechanism of link failure recovery of EAMR show the betterperformance followed by AntHocNet. What is more, EAMRroute maintenance mechanism also makes EAMR send lessforward ants, which reduces the average energy consumption.The average energy consumption of network is minimal inEAMR. As nodes move faster, the data packets drop androuting request also has sharp increase. It is clear that in thecases with high mobility the results show a higher variation,which decreases with less mobility.

5. Conclusions

In this paper, we studied the existing ACO-based routingprotocols and proposed EAMR. In the EAMR, the ant packetstructure, pheromone update formulas, pheromone updatemode, and the mechanism of multipath established are allimproved.

In the pheromone update formulas, EAMR takes intoaccount the energy consumption rate of path, the remain-ing minimum energy path, the hops to the sink, and thecongestion of path. Different from traditional incremental


0.45

0.5

0.55

0.6

0.65

0.7

0.75

0.8

10 20 30 40 50Node speed (m/s)

Ener

gy effi

cien

cy

AOMDVEEABR

AntHocNetEAMR

Figure 10: Average energy consumption of network.

pheromone update mode, the pheromone will be thoroughlyupdated when the node receives a backward ant. With anew multipath mechanism, EAMR can be more reasonableto establish multiple paths between the source node and thefinal destination node. Probabilistic routing mechanism isdesigned to make data flow into network more balanced.

In particular, we can make conclusions from the perfor-mance of EAMR, EEABR, AOMDV, and AntHocNet. Fromthe simulation results, it is clear that EAMR achieves animprovement in energy efficiency, packets delivery ratio, andend-to-end delay.

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper.

Acknowledgments

This work is supported in part by NSFC, National BasicResearch Program of China under Grant no. 61174023,Zhejiang Provincial Key Laboratory of Forestry IntelligentMonitoring and Information Technology Research, and Zhe-jiang A&F University research program foundation talent tostart Project no. 2014FR015. It is also supported by fundsfrom the preresearch project of the ResearchCenter for SmartAgriculture and Forestry, Zhejiang A&F University, whichstarts Project no. 2013ZHNL02. Zhejiang Provincial ScienceTechnology Plan Projects Key Science Technology SpecificProject under Grant no. 2012C13011-1.

References

[1] J. Li and H. Gao, “Survey on senor network research,” Journalof Computer Research and Development, vol. 45, no. 1, pp. 1–15,2008.

[2] I. F. Akyildiz,W. Su, Y. Sankarasubramaniam, and E. Cayirci, “Asurvey on sensor networks,” IEEE Communications Magazine,vol. 40, no. 8, pp. 102–105, 2002.

[3] Z.Ma, Y. Sun, and T.Mei, “Survey on wireless sensors network,”Journal of China Institute of Communications, vol. 25, no. 4, pp.114–124, 2004.

[4] K. C. Rahman, “A survey on sensor network,” Journal ofComputer and Information, vol. 1, no. 1, pp. 76–87, 2010.

[5] G. M. Shafiullah, A. Gyasi-Agyei, and P. J. Wolfs, “A surveyof energy-efficient and Qos-aware routing protocols for wire-less sensor networks,” in Novel Algorithms and Techniques inTelecommunications, Automation and Industrial Electronics, pp.352–357, Springer, Dordrecht, The Netherlands, 2008.

[6] M. Dorigo and T. Stutzle, Ant Colony Optimization, TsinghuaUniversity Press, Beijing, China, 2007.

[7] M. Saleem, G. A. Di Caro, and M. Farooq, “Swarm intelligencebased routing protocol for wireless sensor networks: surveyand future directions,” Information Sciences, vol. 181, no. 20, pp.4597–4624, 2011.

[8] H. Wang and N. Luo, “An improved ant-based algorithm fordata aggregation in wireless sensor networks,” in Proceedingsof the International Conference on Communications and MobileComputing, vol. 3, pp. 239–243, Shenzhen, China, April 2010.

[9] I. Ullah, “Bee-sensor: a step towards meta-routing strategies inhybrid Ad hoc networks,” in Swarm Intelligence: Proceedings ofthe 7th International Conference, ANTS 2010, Brussels, Belgium,September 8–10, 2010, vol. 6234 of Lecture Notes in ComputerScience, pp. 392–399, Springer, Berlin, Germany, 2010.

[10] G. Di Caro and M. Dorigo, “AntNet: distributed stigmergeticcontrol for communications networks,” Journal of ArtificialIntelligence Research, vol. 9, no. 1, pp. 317–365, 1998.

[11] M. Gunes, U. Sorges, and I. Bouazizi, “ARA: the ant-colonybased routing algorithm for MANETs,” in Proceedings of theInternational Conference on Parallel Processing Workshop, pp.79–85, Aachen, Germany, 2002.

[12] O. Hussein and T. Saadawi, “Ant routing algorithm for mobilead-hoc networks (ARAMA),” in Proceedings of the 22nd IEEEInternational Performance, Computing, and CommunicationsConference (IPCCC ’03), pp. 281–290, April 2003.

[13] G. Di Caro, F. Ducatelle, and L. M. Gambardella, “AntHocNet:an adaptive nature-inspired algorithm for routing in mobile adhoc networks,” European Transactions on Telecommunications,vol. 16, no. 5, pp. 443–455, 2005.

[14] L. Rosati, M. Berioli, and G. Reali, “On ant routing algorithmsin ad hoc networks with critical connectivity,”AdHoc Networks,vol. 6, no. 6, pp. 827–859, 2008.

[15] J. Wang, E. Osagie, P. Thulasiraman, and R. K. Thulasiram,“HOPNET: a hybrid ant colony optimization routing algorithmfor mobile ad hoc network,” Ad Hoc Networks, vol. 7, no. 4, pp.690–705, 2009.

[16] S. Misra, S. K. Dhurandher, M. S. Obaidat, P. Gupta, K. Verma,and P. Narula, “An ant swarm-inspired energy-aware routingprotocol for wireless ad-hoc networks,” The Journal of Systemsand Software, vol. 83, no. 11, pp. 2188–2199, 2010.

[17] F. Ducatelle, G. A. Di Caro, and L. M. Gambardella, “Ananalysis of the different components of the AntHocNet routing


algorithm,” in Proceedings of the 5th International Conference onAnt Colony Optimization and Swarm Intelligence (ANTS ’06),pp. 37–48, 2006.

[18] B. Rong, P. Hao, and D. Qifeng, “Ant colony algorithm based onpheromone diffusion mode by the research of wireless sensornetworks,” The Journal of Sensors and Actuators, vol. 24, no. 11,pp. 1644–1648, 2011.

[19] T. Camilo, C. Carreto, J. S. Silva, and F. Boavida, “An energy-efficient ant-based routing algorithm for wireless sensor net-works,” in Ant Colony Optimization and Swarm Intelligence:5th International Workshop, ANTS 2006, Brussels, Belgium,September 4–7, 2006. Proceedings, vol. 4150 of Lecture Notes inComputer Science, pp. 49–59, Springer, Berlin, Germany, 2006.

[20] D. S. J. De Couto, D. Aguayo, J. Bicket, and R. Morris, “Ahigh-throughput path metric for multi-hop wireless routing,”Wireless Networks, vol. 11, no. 4, pp. 419–434, 2005.

[21] M. K. Marina and S. R. Das, “Ad hoc on-demand multipathdistance vector routing,” Wireless Communications and MobileComputing, vol. 6, no. 7, pp. 969–988, 2006.

[22] W. B. Heinzelman, A. P. Chandrakasan, and H. Balakrish-nan, “An application-specific protocol architecture for wirelessmicrosensor networks,” IEEE Transactions onWireless Commu-nications, vol. 1, no. 4, pp. 660–670, 2002.

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014


Active and Passive Electronic Components

Control Scienceand Engineering

Journal of



RotatingMachinery


Hindawi Publishing Corporation http://www.hindawi.com

Journal ofEngineeringVolume 2014

Submit your manuscripts athttp://www.hindawi.com

VLSI Design



Shock and Vibration


Civil EngineeringAdvances in

Acoustics and VibrationAdvances in



Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation http://www.hindawi.com

Volume 2014

The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014

SensorsJournal of


Modelling & Simulation in EngineeringHindawi Publishing Corporation http://www.hindawi.com Volume 2014


Chemical EngineeringInternational Journal of Antennas and

Propagation




Navigation and Observation



DistributedSensor Networks


Documents

Research Article An Energy-Efficient Multipath Routing ...downloads.hindawi.com/journals/ijdsn/2015/642189.pdf · Research Article An Energy-Efficient Multipath Routing Algorithm