A Hybrid Routing Algorithm Based on Ant Colony and ZHLS Routing Protocol for MANET

T.-h. Kim et al. (Eds.): FGCN 2010, Part II, CCIS 120, pp. 112–122, 2010. © Springer-Verlag Berlin Heidelberg 2010

A Hybrid Routing Algorithm Based on Ant Colony and ZHLS Routing Protocol for MANET

Marjan Kuchaki Rafsanjani1, Sanaz Asadinia2, and Farzaneh Pakzad3

1 Department of Computer Science, Shahid Bahonar University of Kerman, Kerman, Iran [email protected]

2 Islamic Azad University Tiran Branch, Tiran, Iran [email protected]

3 Islamic Azad University Khurasgan Branch, Young Researchers Club, Khurasgan, Khurasgan, Iran

[email protected]

Abstract. Mobile Ad hoc networks (MANETs) require dynamic routing schemes for adequate performance. This paper, presents a new routing algo-rithm for MANETs, which combines the idea of ant colony optimization with Zone-based Hierarchical Link State (ZHLS) protocol. Ant colony optimization (ACO) is a class of Swarm Intelligence (SI) algorithms. SI is the local interac-tion of many simple agents to achieve a global goal. SI is based on social insect for solving different types of problems. ACO algorithm uses mobile agents called ants to explore network. Ants help to find paths between two nodes in the network. Our algorithm is based on ants jump from one zone to the next zones which contains of the proactive routing within a zone and reactive routing be-tween the zones. Our proposed algorithm improves the performance of the network such as delay, packet delivery ratio and overhead than traditional routing algorithms.

Keywords: Zone based Hierarchical Link State (ZHLS); Ant Colony Optimiza-tion (ACO); Swarm Intelligence (SI); Mobile Ad hoc Networks (MANETs).

1 Introduction

Mobile ad hoc network (MANET) is an infrastructure-less multi-hop network where each node communicates with other nodes directly or indirectly through inter-mediate nodes. Thus, all nodes in a MANET basically function as mobile routers participating in some routing protocol required for deciding and maintaining the routes. Since MANETs are infrastructure-less, self-organizing, rapidly deployable wireless networks, they are highly suitable for applications communications in re-gions with no wireless infrastructure, emergencies and natural disasters, and military operations [1,2].

Routing is one of the key issues in MANETs due to their highly dynamic and dis-tributed nature. Numerous ad hoc routing algorithms exist to allow networking under various conditions. They can be separated into three groups, proactive, reactive and hybrid algorithms. In proactive routing algorithms maintain continuously updated

A Hybrid Routing Algorithm Based on Ant Colony and ZHLS Routing Protocol 113

state of the network and the existing routes; however, in some cases it may generate an unnecessary overhead to maintain the routing tables and then may be better to create routes only on demand, the case of reactive routing algorithms. In reactive routing algorithms require time-consuming route creations that may delay the actual transmission of the data when sources have no path towards their destination and then, in this case may be better to use a proactive routing algorithm. In hybrid proto-cols try to profit the advantages of both reactive and proactive protocols and combine their basic properties into one. These protocols have the potential to provide higher scalability than pure reactive or proactive protocols thanks to the collaboration be-tween nodes with close proximity to work together and therefore reduce the route discovery overhead [3].

Recently, a new family of algorithms emerged inspired by swarm-intelligence, which provides a novel approach to distributed optimization problems.

The expression “Swarm Intelligence” defines any attempts to design algorithms in-spired by the collective behavior of social insect colonies and other animal societies. Ant colonies, bird flocking, animal herding and fish schooling are examples in nature that use swarm intelligence. Several algorithms which are based on ant colony were introduced in recent years to solve the routing problem in mobile ad hoc networks.

This paper provides the description of a hybrid routing scheme based on both an Ant Colony Optimization (ACO) and a Zone based Hierarchical Link State (ZHLS) protocol that pretends to profit the advantages of both reactive and proactive algorithms.

Ant Colony Optimization (ACO) is a family of optimization algorithms based on real ants' behavior in finding a route to food nest. It has been observed available routes, ants find shortest route to food nest. To achieve this, ants communicate through deposition of a chemical substance called pheromone along the route. Short-est path has highest concentration leading to more and more ants using this route [4]. There are some successful ant-based algorithms for the network that we will introduce them in next section.

2 Related Work

Routing in MANETs has traditionally used the knowledge of the connectivity of the network with emphasis on the state of the links. To overcome the problems associated with the link-state and distance vector algorithms, numerous routing protocols have been proposed. The routing protocols proposed for MANETs are generally categorized into three groups: table driven (also called proactive) and on-demand (also called reac-tive) and hybrid protocols which are both proactive and reactive in nature [3].

2.1 Routing in Mobile ad Hoc Networks

In Proactive routing protocols, each node continuously maintains up-to-date routes to every other node in the network. Routing information is periodically transmitted throughout the network in order to maintain routing table. Thus, if a route has already existed before traffic arrives, transmission occurs without delay. Otherwise, traffic packets should wait in queue until the node receives routing information corresponding to its destination. However, for highly dynamic network topology, the proactive

114 M.K. Rafsanjani, S. Asadinia, and F. Pakzad

schemes require a significant amount of resources to keep routing information up-to-date and reliable.

Proactive protocols suffer the disadvantage of additional control traffic that is needed to continually update stale route entries. Since the network topology is dynam-ic, when a link goes down, all paths that use that link are broken and have to be re-paired. This protocol is appropriate for a network with low mobility.

Certain proactive routing protocols are Destination-Sequenced Distance Vector (DSDV) [5], Wireless Routing Protocol (WRP) [6] and so on. The main differences among them are the number of used tables, the information that is kept and the forward packet police to maintain the tables updated.

Reactive Routing Protocols in contrast to proactive approach, a node initiates a route discovery throughout the network, only when it wants to send packets to its des-tination. For this purpose, a node initiates a route discovery process through the net-work. This process is completed once a route is determined or once a route has been established, it is maintained by a route maintenance process until either the destination becomes inaccessible along every path from the source or until the route is no longer desired. In reactive schemes, nodes maintain the routes to active destinations. A route search is needed for every unknown destination. Therefore, theoretically the communi-cation overhead is reduced at expense of delay due to route research. Furthermore, the rapidly changing topology may break an active route and cause subsequent route searches.

Reactive strategies are suitable for networks with high mobility and relatively small number of flows. Some reactive protocols are Ad hoc On-Demand Distance Vector (AODV) [7], Dynamic Source Routing (DSR) [8], Temporally Ordered Routing Algo-rithm (TORA) [9] and Associativity-Based Routing (ABR) [10].

Hybrid Protocols, each node maintains both the topology information within its zone and the information regarding neighboring zones that means proactive behavior within a zone and reactive behavior among zones. Thus, a route to each destination within a zone is established without delay, while a route discovery and a route main-tenance procedure is required for destinations that are in other zones.

The Zone Routing Protocol (ZRP) [11], Zone-based Hierarchical Link State (ZHLS) routing protocol [12] and Distributed Dynamic Routing algorithm (DDR) [13] are three hybrid routing protocols. The hybrid protocols can provide a better trade-off between communication overhead and delay, but this trade-off is subjected to the size of a zone and the dynamics of a zone.

The hybrid approach is an appropriate candidate for routing in a large network. Joa-Ng et al. [12] proposed a hybrid routing protocol is called Zone-based Hierarchical Link State (ZHLS) routing protocol in the effort to combine the features of proactive and reactive protocols. In ZHLS routing protocol, the network is divided into non-overlapping zones.

Unlike other hierarchical protocols, there is no zone-head. ZHLS defines two levels of topologies - node level and zone level. A node level topology tells how nodes of a zone are connected to each other physically. A virtual link between two zones exists if at least one node of a zone is physically connected to some node of the other zone. Zone level topology describes how zones are connected together. There are two types of Link State Packets (LSP) as well - node LSP and zone LSP. A node LSP of a node contains its neighbor node information and is propagated with the zone where as a


zone LSP contains the zone information and is propagated globally. So, each node has full node connectivity knowledge about the nodes in its zone and only zone connectivi-ty information about other zones in the network. So given the zone id and the node id of a destination, the packet is routed based on the zone id till it reaches the correct zone. Then in that zone, it is routed based on node id. A <zone id, node id> of the destination is sufficient for routing so it is adaptable to changing topologies. In ZHLS, Zone LSPs are flooded throughout the network so that all nodes know both zone level and node level topologies of the network. This simplifies the routing but introduces communication overhead [12].

2.2 Ant-Based Routing Algorithms for MANETs

There exist some successful ant-based algorithms to network control, being the most prominent AntNet [14], and Ant-based Control (ABC) [15], which have a number of properties desirable in MANETs. AntNet and ABC use two ants, forward and back-ward ants to find the shortest route from the source to the destination.

AntNet [14] is a proactive ACO routing algorithm for packet switch networks. In this algorithm, a forward ant is launched from the source node at regular intervals. A forward ant at each intermediate node selects the next hop using the information stored in the routing table of that node. The next node is selected with a probability propor-tional to the goodness of that node which is measured by the amount of pheromone deposited on the link to that node. When a forward ant reaches the destination, it gene-rates a backward ant which takes the same path as the corresponding forward ant but in opposite direction. The backward ant updates pheromone values as it moves on its way to the source node.

ARA (Ant colony based Routing Algorithm) proposed by Gunes et al. [16] is a reac-tive ACO routing algorithm for mobile ad hoc networks. ARA has two phases: route discovery, and route maintenance. In route discovery phase, the sender broadcasts a forward ant. The ant is relayed by each intermediate node until reaches the destination. After receiving a forward ant in the destination, the ant is destroyed and a backward ant is sent back to the sender. The backward ant increases the pheromone value corres-ponding to the destination in each intermediary node until it reaches the sender. When the sender receives a backward ant, the route maintenance phase starts by sending data packets. Since the pheromone track is already established by the forward and backward ants, subsequent data packets will perform the route maintenance by adjusting the pheromone values.

ARAMA (Ant Routing Algorithm for Mobile Ad hoc networks) proposed by Hossein and Saadawi [17] is a proactive routing algorithm. The main task of the for-ward ant in other ACO algorithms for MANETs is collecting path information. How-ever, in ARAMA, the forward ant takes into account not only the hop count factor, as most protocols do, but also the links local heuristic along the route such as the node’s battery power and queue delay. ARAMA defines a value called grade. This value is calculated by each backward ant, which is a function of the path information stored in the forward ant. At each node, the backward ant updates the pheromone amount of the node’s routing table, using the grade value. The protocol uses the same grade to update pheromone value of all links. The authors claim that the route discovery and mainten-ance overheads are reduced by controlling the forward ant’s generation rate. However, they do not clarify how to control the generation rate in a dynamic environment.


AntHocNet is a hybrid ant based routing protocol proposed by Di Caro [18] in the effort to combine the advantages from both AntNet and ARA. AntHocNet reactively finds a route to the destination on demand, and proactively maintains and improves the existing routes or explore better paths.

In AntHocNet, ant maintains a list of nodes it has visited to detect cycles. The source node sends out forward ants and when it receives all the backward ants, one generation is completed. Each node i keeps the identity of the forward ants, the path computation, number of hops, number of the ant from the source to node i, and the time the ant visited node i. Note that more than one ant may have reached node i and therefore the identity of the ant is important. When an ant arrives at a node, the node checks the ant’s path computation and the time it reached node i. If the path computa-tion and time are within a certain limit of those produced by another ant of the same generation then the ant is forwarded. Otherwise, the ant is discarded.

In case of a link failure at a node and no alternative paths are available, the node sends a reactive forward ant to repair the route locally and to determine an alternative path. If a backward ant is received for the reactive forward ant, the data packets are sent along the newly found path and all its neighbors are notified about the change in route. Otherwise, the node sends a notification to all its neighbors of the lost destina-tion paths which in turn initiate forward ants from the neighbors. In the next section, we present the main ideas of our algorithm.

3 The Our Proposed Routing Scheme

Our algorithm uses the ZHLS protocol which consists of the proactive routing within a zone and reactive routing between the zones. The network is divided into zones which are the node’s local neighborhood. The network divides into non-overlapping zones; a node is only within a zone. The zone size depend on node mobility, network density, transmission power and propagation characteristics. Each node knows its physical location by geo-location techniques such as Global Positioning System (GPS). The nodes can be categorized as interior and gateway nodes.

Fig. 1. Example of our scheme structure

Zone7

Zone1

S

AE

B

Zone6

P

F

M

C

I H

L

D

G

KZone4

Zone5

Zone3 Zone2

N


In Fig 1 for node S, nodes C, D, and E are gateway nodes, and nodes A, B are inte-rior nodes. All other nodes are exterior nodes (outside the zone). To determining gate-way and interior nodes, a node needs to know its local neighbors. This is achieved by a detection process based on replies to hello messages transmitted by each node. Each node only knows the connectivity within its zone and the zone connectivity of the whole network.

3.1 Routing Table

The algorithm has two routing tables, Intrazone Routing Table (IntraRT) and Interzone Routing Table (InterRT). IntraRT is a routing table maintained proactively. A node can determine a path to any node within its zone immediately. InterRT is a routing table for storing routes to a destination out of its zone. The gateway nodes of the zone are used to find routes between zones.

3.2 ANTs

The defined ants in our scheme are same with HOPNET algorithm [19] that classified in 5 types: internal forward ant, external forward ant, backward ant, notification ant and error ant. The internal forward ant is the responsible for maintaining the proactive routing table continuously within its zone. The external forward ant performs the reac-tive routing to nodes beyond its zone. When an external forward ant is received at the destination, it is converted to a backward ant and sent back along the discovered route. If a new route is reactively discovered, then a notification ant will be sent to source node and to all nodes on the route to update their reactive routing table. The error ant is utilized to warn some changes in the network topology and to restart a new search by the destination if the source still needs a route.

3.3 Route Discovery

We use ACO algorithm for finding the shortest route between two nodes (Vi,Vj) in network. Each communication link has two values, , represents pheromone value per link and , represents time which the links may be in connection. The pheromone value gets updated by the ants as they move the links. The ants change the concentration of the pheromone value on their path to the destination and on their route back to the source.

Route discovery occurs by Intrazone and Interzone routing. The IntraRT basic structure is a matrix whose rows are its neighbors and the columns are all identified nodes within its zone. In route discovery within a zone (Intrazone routing), each node periodically sends internal forward ants to its neighbors to maintain the Intrazone routing table updated. When the source node wants to transmit a data packet to a node within its zone, it first searches the columns of its IntraRT to see if the destination exists in its zone. If it finds the destination in its IntraRT, then Route discovery phase is done. At the current node, the ant verifies the pheromone amount for each neighbor which has a route to destination. The neighbor which has the biggest pheromone amount is chosen to next hop. After selecting a node as next hop increase pheromone concentration selected link and along all other links the pheromone is decremented.

Pheromone concentration on a link (Vi,Vj) along consists considering the path from current node Vi to source node Vs, the pheromone value on link (Vi,Vs) in Vj’s


routing table is reinforced. The amount of pheromone on a link (Vi,Vs) is increased by following equation[19]: , , , , 1

That has to be chosen appropriately to avoid fast or slow evaporation and T (Vs,Vi) represents the total time required to traverse from Vs to Vi.

The pheromone concentration on all other entries not equal to Vi in the same col-umn Vs in Vj’s routing table is decremented using the evaporation equation below: , 1 , 2

Where is the evaporation coefficient provided by the user [19]. On its path back to the source, an ant again updates the pheromone concentration. The pheromone concen-tration update for entry (Vb, Vd) is [19]:

, , , , 3

If not found the destination in its IntraRT, then Route discovery between zones is done. In route discovery between zones (Interzone routing), When a node wants to send a data packet to a destination node, it verifies the Interzone routing table to discover an existent route. If the route exists and has not expired, then the node transmits the data packet. Otherwise, the node starts a search process to find a new path to destination. When a source node will to transmit a data packet to a node thither its zone, the node sends external forward ants to search a path to the destination.

The external forward ants are first sent by the node to its gateway nodes. The gate-way nodes check to see if the destination is within its zone. If the destination is not within its zone and the path has expired, the ants jump between the border zones via the other gateway nodes until an ant localizes a zone with the destination. This ant propagation through the border zones is called bordercast. At the destination, forward ant is converted to a backward ant and is sent to the source. Then, the data packet is transmitted. Use bordercast and routing tables process reduces the delay, because in-traRT proactively maintains all the routes within its zone and interRT stores the path to the destination that the ants recently visited. These tables contribute to fast end to end packet transmission since the paths are readily accessible.

An example of the route discovery between zones is given below using Fig 1. As-sume the source I want a route to the destination L. Since L does not belong to I’s zone, node I will send external forward ants to gateway nodes its neighbor zones, namely F and H. Nodes H and F look through the IntraRT table to check if L is within its zones. In this example, L will not be in the tables. Therefore, H will send the ant to its gateway node G. Node G will send external forward ants to gateway nodes of its neighbor zones, D and K. D cannot find L in its zone. Therefore, Node D sends the ant to its gateway nodes. Node K finds the destination node L within their zone. K then send forward ants with their attached addresses to node L via the path indicated in IntraRT table. The backward ant traverses in the reverse direction, for example, <L, K, G, H, I> to source I from destination T.


3.4 Route Maintenance

In mobile ad-hoc network, the flexible mobility and communication interference will lead to the invalidation of some route. There are two reasons which an intermediate node will not be able to deliver packets: i) the pheromone concentration along the neighboring links is zero, in this case the ants cannot select any links to travel if all their links, up and down are zero and the data packet is failed at that node, ii) damaged route. If the damaged route is within a zone, it will recover after a period because the IntraRT is proactively maintained. If the damaged route is between zones, the up node of the broken link will conduct a local repair process and then search an alternative path to the destination while buffering all the packets it receives. If the node finds a new path to the destination, it will send all the buffered packets to the destination; then a notification ant will be sent to the source to allow the source node knows the change of route. If a new path cannot be found instead failed path, an error ant will be sent to the source node. Hence packet delivery ratio will be increased [19].

4 Simulation Results

Our algorithm is implemented in GloMoSim simulator. The simulation environment includes 200 mobile nodes working with IEEE 802.11, the area is 1000 m × 1000 m, they move according to the random way point mobility model (RWP). Each node moves with maximal 10 m/s, the whole time of simulation is 300s. The data rate is 2 packets per second (1024 bytes).

Fig. 2 shows the end to end delay of our proposed algorithm in comparison to AODV protocol and HOPNET algorithm. Our proposed algorithm produces better end to end delay results than AODV. This is attributed to the zone framework and the local intrazone routing table and interzone routing table. The intrazone table proac-tively maintains all the routes within its zone and interzone stores the path to the des-tination that the ants recently visited.

Fig. 2. End to end delay

120 M.K. Rafsanjani, S. A

These tables contribute readily accessible. By adjuants can traverse on the liconcentration. The evaporareasons allow our proposed

Fig 3 shows the deliverySince the network has densfrom multiple paths rather t

Fig. 4 shows the control ovAODV is a pure reactive prcontrol packets are periodizone. This is a major factor

Fig

Asadinia, and F. Pakzad

to fast end to end packet transmission since the paths usting the evaporation rate of pheromone on the links, inks or ignore the links by decrementing the pheromation rate helps in discarding links that are broken. Thd algorithm to produce better end to end delay results. y ratio for our proposed algorithm, HOPNET and AODe ants can find multiple paths, because the ants can chothan a single path like AODV.

Fig. 3. Packet delivery ratio

verhead of our proposed algorithm, HOPNET and AODrotocol. Proposed algorithm is proactive within a zone. Tically sent out within a zone to maintain the routes in r for the overhead in proposed algorithm.

. 4. Overhead per true received packets

are the

mone hese

DV. oose

DV. The the


5 Conclusion

In this work, Ant Colony Optimization algorithm and Zone-based Hierarchical Link State protocol are used for routing in MANETs. In fact, it is a hybrid routing algo-rithm that has the potential to provide higher scalability than pure reactive or proactive protocols. Our algorithm contains the proactive routing within a zone and reactive routing between the zones. The scheme that we presented in this paper only represents our initial effort for the development of routing algorithm for MANETs. In addition, although we have reasoned that our routing scheme is more advantageous over most previous schemes in terms of end to end delay and packets delivery ratio. As the net-work size increases, the overhead decreases and it is better than AODV.

References

1. Haas, Z.J., Gerla, M., Johnson, D.B., Perkins, C.E., Pursley, M.B., Steenstrup, M., Toh, C.K.: Mobile Ad-Hoc Networks. IEEE J. on Selected Areas in Communications, Special Issue on Wireless Networks 17(8), 1329–1332 (1999)

2. Mauve, M., Widner, J., Hartenstein, H.: A Survey on Position-based Routing in Mobile Ad-hoc Networks. IEEE Network J. 16, 30–39 (2001)

3. Abolhasan, M., Wysocki, T., Dutkiewicz, E.: A Review of Routing Protocols for Mobile Ad hoc Networks. Ad Hoc Networks J., Elsevier Computer Science 2, 1–22 (2004)

4. Dorigo, M., Di Caro, G., Gambardella, L.: Ant Colony Optimization: A New Meta-heuristic. In: IEEE Congress on Evolutionary Computation, Washington, DC, vol. 2, pp. 1470–1477 (1999)

5. Perkins, C.E., Watson, T.J.: Highly Dynamic Destination Sequenced Distance Vector Routing (DSDV) for Mobile Computers. In: ACM Conference on Communications Archi-tectures SIGCOMM 1994, London, UK (1994)

6. Murthy, S., Garcia-Luna-Aceves, J.J.: A routing protocol for packet radio networks. In: The 1st ACM/IEEE Annual International Conference on Mobile Computing and Network-ing, Berkeley, CA, pp. 86–95 (1995)

7. Das, S., Perkins, C., Royer, E.: Ad hoc on Demand Distance Vector (AODV) Routing. In-ternet Draft, draft-ietf-manetaodv-11.txt, work in progress (2002)

8. Johnson, D.B., Maltz, D.A.: The Dynamic Source Routing Protocol for Mobile Ad hoc Networks. Internet Draft, draft-ietf-manet-dsr-07.txt, work in progress (2002)

9. Toh, C.K.: Associativity-based Routing for Ad-hoc Mobile Networks. Wireless Personal Communications 4(2), 103–139 (1997)

10. Park, V.D., Corson, M.S.: A Highly Adaptive Distributed Routing Algorithm for Mobile Wireless Networks. In: The IEEE Conference on Computer Communications, Kobe, Ja-pan, pp. 7–11 (1997)

11. Hass, Z.J., Pearlman, R.: Zone Routing Protocol for Ad-hoc Networks, Internet Draft, draft-ietf-manet-zrp-02.txt, work in progress (1999)

12. Joa-Ng, M., Lu, I.T.: A Peer-to-Peer Zone-based Two-level Link State Routing for Mobile Ad Hoc Networks. IEEE J. on Selected Areas in communications, Special Issue on Ad-Hoc Networks, 1415–1425 (1999)

13. Nikaein, N., Laboid, H., Bonnet, C.: Distributed Dynamic Routing Algorithm (DDR) for Mobile Ad hoc Networks. In: 1st Annual Workshop on Mobile Ad Hoc Networking and Computing, MobiHOC 2000 (2000)


14. DiCaro, G., Dorigo, M.: AntNet: Distributed Stigmergetic Control for Communications Networks. J. on Artificial Intelligence Research 9, 317–365 (1998)

15. Schoonderwoerd, R., Holland, O., Bruten, J., Rothkrantz, L.: Ant-based Load Balancing in Telecommunication Networks. Adaptive Behavior 5, 169–207 (1996)

16. Gunes, M., Sorges, U., Bouazzi, I.: ARA – The Ant Colony Based Routing Algorithm for MANETs. In: the International Conference on Parallel Processing Workshops (ICPPW 2002), Vancouver, BC, pp. 79–85 (2002)

17. Hossein, O., Saadawi, T.: Ant Routing Algorithm for Mobile Ad hoc Networks (ARA-MA). In: 22nd IEEE International Performance, Computing, and Communications Confe-rence, Phoenix, Arizona, USA, pp. 281–290 (2003)

18. DiCaro, G., Ducatelle, F., Gambardella, L.M.: AntHocNet: An Adaptive Nature Inspired Algorithm for Routing in Mobile Ad hoc Networks. European Transactions on Telecom-munications (Special Issue on Self-Organization in Mobile Networking) 16(2) (2005)

19. Wang, J., Osagie, E., Thulasiraman, P., Thulasiram, R.K.: HOPNET: A Hybrid Ant Colo-ny Optimization Routing Algorithm for Mobile Ad hoc Network. Ad Hoc Network J. 7(4), 690–705 (2009)

A Hybrid Routing Algorithm Based on Ant Colony and ZHLS Routing Protocol for MANET

Documents