Chapter 24 Using Ant Colony Agents for Designing Energy-Efficient Protocols for Wireless Ad Hoc and Sensor Networks

1
Chapter 24Using Ant Colony Agents for
Designing Energy-Efficient Protocols for Wireless
Ad Hoc and Sensor Networks

• Isaac Woungang
• (Department of Computer Science ,Ryerson
  University, Toronto, Ontario, Canada)
• Sanjay K. Dhurandher
• (Division of Information Technology, Netaji
  Subhas Institute of Technology
• (NSIT),University of Delhi, India)
• Mohammad S. Obaidat
• (Department of Computer Science and Software
  Engineering
• Monmouth University, NJ, USA)

2
Summary

• Literature Survey
• Study of Existing Protocols (MTPR,MMBCR,CMMBCR,
  EAAR etc.)
• Finding their Limitations
• Proposed Protocol (ACO-CMMBCR)
• Design Goals
• Algorithm
• Comparison with Benchmark Protocols
• Proving the correctness of protocol through
  results
• Graphical Plots

3
Introduction

• Mobile Ad-hoc Networks (MANETs)
• Special type of wireless network
• Nodes form a temporary network without any fixed
  infrastructure
• Nodes are mobile
• Each node can act as a source, destination or
  just an intermediate node.

4
Uses of MANETs

• Military services Rescue
  Operations
•    

5
Literature Survey
6
MMBCR Protocol

• Description
• It aims towards efficient usage of residual
  battery of nodes
• Even a single dead node in the network results
  into poor performance of the network, hence its
  very important to have least dead nodes in the
  network
• Uses a variable MBR (Minimum Residual Battery) of
  a path which is equal to the minimum battery left
  of a node along that path
• To select a path from Source S to Destination
  D ,the protocol finds the MBR of all possible
  paths
• The path with highest MBR is selected for Routing

7
EAAR Protocol

• Description
• Provides efficient energy usage in the network
• Based on the Foraging Behavior in Ant Swarms
• Implements the Ant Colony Optimization(ACO)
  scheme on the MMBCR protocol.
• Multipath transmission properties of ant swarms
  increases packet delivery ratio.
• Defines a pheromone variable for each path
• Ti(n,d) MBR / H
• Where MBRMinimum Residual Batter Energy,
• HHops
• The above formula is used in the selection of a
  path

8
EAAR Protocol

• Features and Discussion
• Emphasizes more on maximizing Residual battery of
  nodes
• Doesnt minimizes the total energy consumption
• Its a good approach for a network where lot of
  weak nodes are present
• This is not a good approach when there is not a
  big issue with residual battery of nodes.
• Lets Say if each node in the network has a high
  battery capacity left then the next step should
  be to minimize total transmission energy usage
  rather than focusing more on saving residual
  battery of a node.

9
MTPR Protocol

• Description
• Minimum Transmission Power Routing
• Minimizes the Total Transmission energy in the
  network
• For each path a cost variable is calculated.
• Cost is directly proportional to the Energy
  consumption along the path
• Higher the cost poorer is the path
• Cost ? Energy(i,j)
• Where Energy(i,j) is the energy required to
  transmit data from node I to node j
• Energy(I,j) is directly proportional to the
  square of sidtance between node I and node j

10
MTPR Protocol

• Limitations
• Doesnt bother about
• residual Battery of a node
• Dead Node
• Large number of Dead nodes

11
CMMBCR Protocol(Conditional Max-Min Battery
Capacity Routing)

• Description
• Selects between MTPR and MMBCR scheme
• The selection is done on the basis of a variable
  gamma
• The main reason of doing a selection over here is
  that each protocol has its own advantage in a
  particular circumstance
• So basically CMMBCR tries to mix up the
  advantages of both MTPR and MMBCR in one protocol
• Defines a variable gamma which is used for
  selection
• Value of gamma determines which protocol should
  be used

12
CMMBCR Protocol

• Algorithm
• Step1Find the MBR of each path from source to
  destination
• Step2Compare MBR of each path with gamma. If a
  path has MBR gt gamma then put it in Set(A)
• Step3
• Check If Set(A) is not empty
• MTPR MMBCR
  

13
ACO-based algorithm by Camilo et al.5

• This algorithm maximizes the network lifetime of
  wireless sensor networks.
• The lengths of the routing paths, the node's
  energy level, and the amount of pheromone trail
  available on the connections between the nodes,
  are considered as design parameters to construct
  a routing tree that has optimized energy
  branches.
• The potential energy saving that this scheme may
  have benefited if the node's status was
  investigated or if multiple sink nodes were
  integrated, was not investigated.

14
ACO-based algorithm by Wen et al.6

• This algorithm is designed for minimizing the
  time delay in wireless sensor networks when
  transferring the data, while accounting for the
  energy level of a node as constraint.
• In their scheme, ant agents routing-tables of
  each node are built based on partial pheromones
  and heuristic values.
• These values are then updated by a back round ant
  that holds the network load and delay
  information.
• A reinforcement learning technique is employed
  to address the tradeoff between delay and energy
  level at each node.
• It results to an energy efficient scheme compared
  to the AntNet scheme 24, in terms of energy
  consumed by each packet during transmission.
• However, this scheme did not address the
  situation when the traffic load at a node might
  turn out to be heavy.

15
A two-steps algorithm by Salehpour et al.25

• This algorithm combines a Clustering technique
  with an ACO-based heuristic to design an
  energy-efficient routing scheme for wireless
  sensor networks.
• In the first step, the Low-Energy Adaptive
  Clustering Hierarchy algorithm (LEACH) 5 is
  used to achieve clustering and message
  transmission in the network, resulting to evenly
  distributed energy consumption among all the
  nodes in the network.
• In the second step, an ACO-based heuristic (the
  AntNet scheme 24) is invoked by the cluster
  heads (which are inherited from the first step)
  to send the aggregated data packets to the base
  station, and this process repeats iteratively
  until convergence is reached.

16
A two-steps algorithm by Salehpour et
al.25(Contd.)

• In the latter, backward and forward ant agents
  are used in collaboration to explore the routing
  possibilities of the data packets throughout the
  network.
• These are based on the information gathered by
  each node regarding the amount of pheromone on
  the paths to its neighbors and the decision made
  by ants based on the energy level of the neighbor
  nodes.
• One major concern with this scheme is that the
  heuristic value associated with each node is
  dependent on the energy level of that node.
• No method was disclosed to estimate this value,
  and the impact of this parameter on the obtained
  optimized solutions was not investigated.

17
ACO-based algorithm by Wang et al.7

• This algorithm uses quality-of-service (QoS)
  provisioning and balanced energy consumption as
  target to achieve energy efficiency.
• In this scheme, service differentiation between
  Real time (RT) and Best effort (BE) traffic is
  made through designing a specific pheromone model
  where artificial ants are extended.
• This yields ants that are endowed with the
  capability of emitting two types of pheromone
• (1) RT pheromone scheme - used to achieve
  the above-mentioned balance energy consumption
  for BE traffic considering the path hop count and
  minimum residual energy along the path as
  constraint parameters.
• (2) BE heuristic scheme - which focuses on
  ensuring the necessary QoS provisioning on the
  selected routing path between a sensor and a
  sink.

18
ACO-based algorithm by Wang et al.7(Contd.)

• The routing tables at each node are updated
  according to these BE and RT pheromone matrices.
• Although this scheme was proved to balance the
  energy consumption in the whole network in real
  world situations.
• The authors neglected to compare their scheme
  against similar state-state-of-the-art well-known
  schemes, in terms of efficiency, or energy
  related performance metrics.

19
Protocol by Dhurandher et al. 8

• The authors proposed an ant swarm-based algorithm
  that integrates both the power consumption at
  each node when routing a data packet and
  multi-path transmission features of artificial
  ants.
• In their proposed scheme, the energy usage is
  minimized by means of
• The path discovery process, inspired from the
  features of AntHocNet 9.
• And designed based on parameters such as route
  hop count and minimum battery energy remaining
  from the weakest node of the route.

20
Protocol by Dhurandher et al. 8(Contd.)

• On the other hand, multi-path transmission is
  used to divert the packet flow in case of link
  failure (assumed to occur one at a time), leading
  to less number of dead nodes compared to the
  AntHocNet 9 scheme.
• The merit of this protocol is that
  energy-awareness is used as a factor to increase
  the time that the protocol takes to judge the
  best possible route to be used for the data
  packets transmission.
• As pointed out by the authors, their proposed
  protocol was not tested in a real test-bed
  environment using in real-life scenario
  applications.

21
ACO-based solution by Okdem and Karaboga 10

• By considering the energy conservation at each
  node, their routing scheme is designed in such a
  way as
• (1) To deal with failure in communication node -
  this is addressed by sustaining multiple paths
  alive in the routing task
• (2) To deal with the energy level at each node
  and the length of the paths - these are handled
  by implementing a mechanism that chooses the
  nodes with more energy when performing the
  routing process
• (3) To incorporate the ACO-based approach - where
  artificial ants contribute in designing effective
  multi-path data transmission from source to sink
  based on the information gathered at each node
  about the amount of pheromone on the available
  paths.
• In order to validate their approach, the authors
  introduced a real-time test environment made of a
  router chip, implemented in the form of a small
  sized hardware component.
• However, the case of multiple sink nodes was not
  investigated.

22
ACO-based multipath routing algorithm by Xia and
Wu 11

• This algorithm uses the energy consumption of
  each path and the available power of nodes as
  criteria for selecting the optimal routing path
  (among multiple available paths) for the delivery
  of packets from source nodes to the sink node.
• It improves the simple ACO (SACO) scheme 26,
  in the sense that an optimized state transition
  and global pheromone update rules are introduced
  to increase the possibility of ants to find a new
  path
• To avoid local optimization.
• To maintain the multi-paths possibility when
  transferring the data packets from the source
  nodes to the sink respectively.
• However, the mobility of sensor nodes was not
  taken into consideration.

23
ACO-based energy-efficient routing protocol by
Misra et al.12

• This Protocol combines the effect of power
  consumption when transmitting a packet, the
  residual battery capacity of a node, and the
  multi-path transmission properties of artificial
  ant swarms.
• In their scheme, the path discovery phase is
  inspired from AntHocNet 9, but with distinct
  functionally.
• The routes are maintained through new pheromone
  reinforcement and evaporation techniques, leading
  to the use of multi-path transmission through the
  "good routes" only rather than all the possible
  paths.
• Even though this scheme showed good promises, the
  effectiveness of the proposed scheme was not
  tested in real test-bed using practical
  scenarios.

24
A self-governed ACO-inspired routing scheme by
Mahadevan and Chiang 13

• The authors proposed a self-governed ACO-inspired
  routing scheme to solve the packet routing
  problem with minimal energy consumption for each
  hop communication, leading to maximum lifetime of
  the network.
• Their scheme is inspired from the Max-Min ant
  system (MMAS) 14 to produce optimized routing
  paths to transfer the data from source nodes to
  the sink, while considering energy efficiency and
  self-healing as design criteria.
• However, their proposed scheme was not compared
  against few other state-of-the-art benchmarks,
  nor implemented in a real tested in order to
  judge its efficiency when dealing with practical
  scenarios.

25
ACO-based routing protocol by Hui et al. 15

• This Protocol considers the node energy, the
  frequency of a node acting as a router to achieve
  the routing, and the path delay, as design
  criteria.
• Their scheme is based on the idea that using the
  lowest energy path does not necessary mean
  obtaining the long-term network lifetime due to
  the fact that the optimal path may quickly get
  energy depleted.
• The authors have followed the basic ACO principle
  for selecting the optimal path to transmit the
  data form source nodes to sink.
• The originality of their scheme stands in the
  fact that in its route discovery and maintenance
  phase, the routing tables at each node were
  updated according to the pheromone and energy
  levels at that node.
• However, node mobility was not considered.

26
ACO-Energy Saving Routing (A-ESR) algorithm by
Kim et al. 16

• The energy saving problem was formulated as an
  energy-consumption minimized network (EMN)
  optimization problem.
• It is based on the concept of traffic centrality
  of a node, defined as a measure involving the
  traffic volume (in bytes) on a link and the
  density of traffic carried on that link then
  solved using the ACO method where only a single
  artificial ant is considered.
• The optimized energy efficiency level produced by
  the proposed by the algorithm is dependent on a
  controlling factor that was used to weight the
  traffic centrality.
• However, the authors neglected to indicate how
  the value of this factor can be allocated in a
  dynamic manner.

27
ACO-based energy-aware Routing(ABEAR) by Ren et
al. 17

• Their proposed scheme introduces a congestion
  matric and uses it along with a combination of
  reactive route setup procedure and proactive
  neighbor maintenance procedure in its routing
  phase to find suitable paths for transferring the
  data from source to destination.
• In this process, the link quality, remaining
  energy at each node, and pheromone values are
  integrated as design variables in the ACO
  approach when performing the routing computation,
  with the goal to reduce the network lifetime.

28
Energy-Aware ACO Routing Algorithm (EAACA) by
Cheng et al. 18

• In their scheme, the residual energy of the
  one-hop neighbor of each node, and the distance
  from source to sink are used as design criteria
  in the selection of the paths to route the data
  packets.
• In the route discovery phase, the information
  gathered by each node regarding the amount of
  pheromone on the paths and the decision made by
  ants based on the residual energy of its one-hop
  neighbor are used to establish all valid paths
  between the sensor nodes and the destination node
  before the source node starts releasing the data
  packets.
• In the route maintenance phase, probe packets are
  sent to the destination node periodically to
  monitor the quality of the chosen transmission
  paths.
• Although this scheme was shown to balance the
  energy consumption at each node, the case of
  mobile sensor nodes was neglected.

29
Adaptable and balanced ACO-based routing
algorithm by Dominiquez-Medina and Cruz-Cortes
19

• This algorithm considers memory and power supply
  as criteria to minimize the energy consumption
  and latency in data transmission.
• The ACO design of the proposed scheme is a
  combination of
• The ACO-based Location Aware Routing for WSNs
  (ACLR) 20 - which attempts to establish an
  equilibrium between the sensor nodes lifetime and
  the delay of the transmissions.
• and the Energy Efficient Ant Based Routing
  Algorithm (EEABR) 5 - which considers the
  energy efficiency in order to maximize the
  network lifetime.
• However, the proposed scheme was not implemented
  in a real tested in order to judge its efficiency
  when dealing with practical scenarios.

30
ACO-CMMBCR

• Consists of 2 main parts
• 1) Implementing the ACO scheme
• Defining pheromone for each path
• Greater the pheromone means better the path
• 2) Dynamic Protocol Selection
• Does an intelligent selection
• Selection between (MTPRACO) and (MMBCRACO)
  depending upon the value of gamma
• Selection is done to minimize energy usage

31
Implementing ACO on CMMBCR

• Defining the pheromones
• A-CMMBCR considers the combination of two routing
  schemes , hence it uses two pheromones
• Pheromone(mm) for MMBCR and pheromone(mt) for
  MTPR.
• Pheromone(mt)1/(Total Transmission energy of
  path Number of Hops)
• Pheromone(mm) MBR/(Number of Hops)
• where, MBRMinimum battery of a node in the
  path.
• Total transmission power is the sum of
  transmission power to send data to next hop for
  each node in the path.

32
Implementing ACO on CMMBCR

• Using the pheromones
• The two pheromones are used for deciding the path
  to be chosen for routing.
• Routing table is organized in such a way that
  paths from source S to destination D are
  stored.
• For each path two pheromones are stored and MBR
  of that path is also stored.
• The purpose of storing these values for a path is
  that these are used when the selection is done.

33
A-CMMBCR

• Algorithm
• If a source node 'S' wants to send data to a
  destination node 'D' then following steps must
  take place
• Step1
• The node S checks its routing table to find
  whether a path to D exists or not. If a path
  exists, it sends the data to the next Hop else
  Step 2 is performed.
• Step2
• The node S broadcasts route request packet
  (RREQ). Then Step 3 is performed.
• Step3
• If any neighbor nodes routing table has a path
  to D exists it replies back to node S through
  Route Reply packet (RREP) else it broadcasts the
  RREQ. Step3 is followed for each intermediate
  node thus receiving the RREQ. If no path for D is
  available, the intermediate node relays the RREQ
  packet.

34
A-CMMBCR

• Algorithm
• Step 4
• As the RREQ packet is broadcast in the
  network, it can eventually reach the destination
  node D. At the destination node, Route Reply
  packet (RREP) is generated and reply is sent back
  to S. RREP is passed to node S through the
  intermediate nodes along the path from which RREQ
  was received. Now as each node receives the RREP
  packet, it updates its
• routing table

35
Performance Evaluation of the
A-CMMBCR Protocol

• In this work, we have used the GLOMOSIM simulator
  24 to compare
• The ETB-MDSR scheme against the CMMBCR 12,
• The Minimum Transmission Power Routing (MTPR)
  27), and the Energy-Aware Routing protocol
  (EAAR) 8 schemes.
• On the basis of the performance metrics
• The network energy usage.
• The load distribution (in terms of number of
  packets per node).

36
Simulation Parameters
Number of nodes 40
Simulation time 500 (s)
Initial energy of nodes All Nodes were initiated with an equal energy value
Terrain dimension 2000 (m) x 2000 (m)
Traffic Type CBR, with the following scenarios CBR 17 100 1536 1S 0S 250S CBR 12 19 100 1536 1S 250S 400S CBR 14 27 100 1536 1S 400S 500S
MAC protocol IEEE 802.11
Mobility model Random waypoint (when applicable) or none.
37
Simulation Scenarios
Scenarios Data size Node speed Mobility
1 100 times Control Packet Size - None
2 125 times Control Packet Size - None
3 150 times Control Packet Size - None
4 100 times Control Packet Size 10 m/s Random waypoint
5 125 times Control Packet Size 10 m/s Random waypoint
6 150 times Control Packet Size 10 m/s Random waypoint
38
Results
Benchmark Protocols used for comparison are
MTPR,EAAR and CMMBCR Graph1. Energy usage vs.
number of node
39
Results Analysis

• Graph1 Analysis
• It is observed that in terms of energy consumed
  per packet, the A-CMMBCR scheme outperforms all
  other schemes.
• This is attributed to the features of the ACO
  scheme used in A-CMMBCR as it generates
  multi-paths from one node to another, thus
  provides an alternative for path overloading.
• In addition, the path with highest pheromone on
  it consumes less energy than selected using the
  EAAR scheme.

40

• Graph2. Network Energy usage vs. number of node.

41
Results Analysis

• Graph2 Analysis
• It can be observed that compared to the other
  schemes, the A-CMMBCR scheme uses least energy as
  the number of nodes increases.
• This can be justified by the fact that the ACO
  framework used in A-CMMBCR optimizes the energy
  usage.
• Indeed, as the number of nodes increases, the
  density increases, which requires an efficient
  usage of energy.
• In the A-CMMBCR scheme, ACO helps in serving this
  purpose by equally distributing the packets to
  the nodes, thereby boosting the residual battery
  of nodes, and hence saving the energy at each
  node.

42

• Graph3. Average traffic distribution vs. number
  of node
• It can be observed that in the case of the
  A-CMMBCR scheme, the traffic distribution is
  even.
• This can be justified by the fact that thanks to
  the design features of its ACO framework, the
  A-CMMBCR systematically distributes the packets
  to the paths that are less condensed.

43
Conclusion

• In this Presentation, we overviewed recent
  proposals on the use of ACO-based algorithms for
  designing energy-efficient routing protocols for
  ad hoc wireless and sensor networks.
• It was reported that the studied family of ACO
  heuristics yielded a much better solution to the
  energy consumption problem compared to
  conventional approaches.
• We also introduced an enhancement to a recently
  proposed ACO-based routing protocol (called
  A-CMMBCR), which belongs to the aforementioned
  family of protocols.

44
Conclusion (Contd.)

• We have showed through simulations that A-CMMBCR
  outperformed the CMMBCR, EAAR and MTPR schemes,
  in terms of energy consumed per packet, energy
  usage, average traffic distribution, used as
  performance metrics.
• We believe that the ACO paradigm will continue
  to be used as a powerful algorithmic framework
  that can contribute in solving various types of
  optimization problems, including energy-related
  problems that may arise in next generation
  networks, including green networks.

45
References

• 1 C.-K. Toh, Maximum Battery Life Routing to
  Support Ubiquitous Mobile Computing in Wireless
  Ad Hoc Networks, IEEE Communications Magazine,
  Vol. 39, No. 6, June 2001, pp. 138-147.
• 2 Sanjay K. Dhurandher , Sudip Mishra and
  Mohammad S. Obaidat , An Energy-Aware Routing
  Protocol for Ad-Hoc Networks Based on Foraging
  Behavior in Ant Swarms, IEEE ICC 2009, pp. 1-5.
• 3 K. Scott and N. Bambos, Routing and Channel
  Assignment for Low Power Transmission in PCS,
  Proc. Intl. Conf. Universal Personal
  Communications (ICUPC?96), Cambridge, MA, 1996,
  pp. 498-502.
• 4 C. E. Perkins, E. M. Belding-Royer, and S.
  Das, Ad Hoc On-demand Distance Vector (AODV)
  Routing, IETF Internet Draft, 2001.
• 5 T. Camilo, C. Carreto, J. Silva, F. Boavida,
  "An Energy-Efficient Ant Base Routing Algorithm
  for Wireless Sensor Networks", In ANTS 2006, 5th
  International Workshop on Ant Colony Optimization
  and Swarm Intelligence, 4150, pp. 49-59, 2006.

46
References

• 6 Y. Wen, Y. Chen, and D. Qian, "An Ant-based
  approach to Power-Efficient Algorithm for
  Wireless Sensor Networks", In Proc. of the World
  Congress on Engineering (WCE 2007), vol. II, July
  2-4, London, U.K., 2007.
• 7 J. Wang, J. Xu, and M. Xiang, "EAQR An
  Energy-ecient ACO Based QoS Routing Algorithm in
  Wireless Sensor Networks", Chinese Journal of
  Electronics, vol. 18, No.1, Jan, 2009.
• 8 S. K. Dhurandher, S. Misra, M. S. Obaidat, P.
  Gupta, K. Verma, and P. Narula, "An Energy-Aware
  Routing Protocol for Ad-Hoc Networks Based on the
  Foraging Behavior in Ant Swarms", In Proc. of the
  IEEE International Conference on Communications
  (ICC'09), Dresden, Germany, June 14-18, pp. 1-5,
  2009.
• 9 G. Di Caro, F. Ducatelle and L. M.
  Gambardella, "AntHocNet An Adaptive
  Nature-Inspired Algorithm for Routing in Mobile
  Ad Hoc Networks, Telecommunications (ETT), vol.
  16, No. 2, 2005.
• 10 S. Okdem and D. Karaboga, "Routing in
  Wireless Sensor Networks Using an Ant Colony
  Optimization (ACO) Router Chip", Sensors 2009,
  vol. 9, pp. 909-921, doi 10.3390/s90200909,
  2009.

47
References

• 11 S. Xia, S. Wu, "Ant Colony-based
  Energy-Aware Multipath Routing Algorithm for
  Wireless Sensor Networks", In Proc. of 2nd
  International Symposium on Knowledge Acquisition
  and Modeling (KAM'09), Nov. 30-Dec 1, Wuhan,
  China, pp. 198-201, 2009.
• 12 S. Misra, S. K. Dhurandherb, M. S. Obaidatc,
  P. Guptab, K. Verma, P. Narula, "An ant
  swarm-inspired energy-aware routing protocol for
  wireless ad-hoc networks", The Journal of Systems
  and Software 83 (2010) 21882199.
• 13 V. Mahadevan and F. Chiang, "iACO A
  Bio-inspired Power Efficient Routing Scheme for
  Sensor Networks", Intl. Journal of Computer
  Theory and Engineering, vol.2, No.6, pp.
  1793-8201, Dec, 2010.
• 14 T. Stutzle and H. Hoos, "MAX-MIN Ant system
  and local search for combinatorial optimization
  problems", In S. Voß, S. Martello, I.H. Osman,
  and C. Roucairol, editors, Meta-Heuristics
  Advances and Trends in Local Search Paradigms for
  Optimization, Kluwer Academics, Boston, MA, USA,
  pp. 313329, 1999.

48
References

• 15 X. Hui, Z. Zhi-gang, N. Feng, "A Novel
  Routing Protocol in Wireless Sensor Networks
  based on Ant Colony Optimization", Intl. Journal
  of Intelligent Information Technology
  Application, vol. 3, No. 1, pp. 1-5, 2010.
• 16 Y-M. Kim E-J. Lee H-S. Park, "Ant Colony
  Optimization Based Energy Saving Routing for
  Energy-Efficient Networks", IEEE Communications
  Letters, vol. 15, Issue 7, pp. 779-781, July,
  2011.
• 17 J. Ren, Y. Tu, M. Zhang, and Y. Jiang, "An
  Ant-Based Energy-Aware Routing Protocol for Ad
  hoc Networks", In Proc. of Intl. Conference on
  Computer Science and Service System (CSSS),
  Nanjing, China, pp. 3844-3849, June 27-29, 2011.
• 18 D. Cheng, Y. Xun, T. Zhou, W. Li, "An Energy
  Aware Ant Colony Algorithm for the Routing of
  Wireless Sensor Networks", Journal on
  Communications in Computer and Information
  Science, Springer, USA, vol. 134, 395-401, 2011.
• 19 C. Dominquez-Medina, N. Cruz-Cortes,
  "Energy-efficient and location-aware ant colony
  based routing algorithms for wireless sensor
  networks", In Proc. of the 13th ACM Annual
  Conference on Genetic and Evolutionary
  Computation (GECCO'11), New York, NY, USA, ISBN
  978-1-4503-0557-0, 2011.

49
References

• 20 X. Wang, Q. Li, N. Xiong, and Y. Pan, "Ant
  Colony Optimization-Based Location-Aware Routing
  for Wireless Sensor Networks", Springer-Verlag,
  Lecture Notes in Computer Science, 5258109120,
  2008.
• 21 S. K. Dhurandher, M. S. Obaidat, and M.
  Gupta, "Application of Ant Colony Optimization to
  Develop Energy Efficient Protocol in Mobile
  Ad-Hoc Networks", Journal of Network and Computer
  Applications 34(5) 1498-1508, 2011.
• 22 K. Scott and N. Bambos, Routing and Channel
  Assignment for Low Power Transmission in PCS, In
  Proc. of the International Conference Universal
  Personal Communications (ICUPC96), Cambridge,
  MA, USA, pp. 498-502, 1996
• 23 S. Singh, M. Woo, and C. S. Raghavendra,
  Power-Aware Routing in Mobile Ad Hoc Networks,
  In Proc. of the 4th Annual ACM/IEEE International
  Conference on Mobile Computing and Networking,
  Dallas, TX, USA, pp. 181-190, 1998.
• 24 G. A. Di Caro and M. Dorigo, "AntNet A
  mobile agents approach to adaptive routing",
  Technical report IRIDIA/97-12, IRIDIA, Université
  Libre de Bruxelles, 1997.

50
References

• 25 A.-A. Salehpour, B. Mirmobin, A.
  Afzali-Kusha, S. Mohammadi, "An energy efficient
  routing protocol for cluster-based wireless
  sensor networks using ant colony optimization",
  In Proc. of the International Conference on
  Innovations in Information Technology (IIT), Al
  Ain, United Arab Emirates, Dec. 16-18, pp.
  455-459, 2008.
• 26 A. P. Engelbrecht, "Computational
  Intelligence An Introduction", 2nd Edition,
  University of South Africa, Pretoria, John Wiley
  Sons Ltd., ISBN 978-0-470-03561-0, 2007.
• 27 L. M. Gambardella and M. Dorigo, "Solving
  Symmetric and Asymmetric TSPs by Ant Colonies",
  In Proc. of IEEE International Conference on
  Evolutionary Computation, Nagoya , Japan, pp.
  622627, May 20-22, 1996.