1 AntHocNet Protocol for Achieving QoS in MANET Usha Devi Gandhi, Kasa Sreenivasa Reddy, Dino Cherian and Avinash Mishra School of Information Technology & Engineering, VIT University, TamilNadu, India ushadevi.g@vit.ac.in, sreenukasa@gmail.com Abstract--Mobile Ad-hoc networks are common in the world after the phones, PDAs, laptops. In the mobile/wireless ad-hoc networks (MANET) management, Quality of service (QoS) and network reliability are important issues. Existing routing algorithms are not sufficient to achieve these properties due to the dynamic nature of the ad-hoc networks. AntHocNet is a biological algorithm based on ant colony optimization (ACO) which can be used to achieve above properties in a MANET. In this paper, we proposed some modifications in existing ant based routing algorithm to achieve QoS with the simulation results. Key words: ANT agents,MANET,ACO,QoS 1. INTRODUCTION A mobile ad–hoc network is infrastructureless, self configuring network of mobile devices like smart phones, laptops, tabs and so on. This devices are connected by wireless connection. The ad-hoc networks are dynamic, which means any existing devices in the network can be passive at any time and also any new devices can be added to the network at any time, so the design of efficient routing algorithm is required for route discovery and updation of these ad-hoc networks. In ad-hoc networks, QoS [1] became more important due to the increase in number of applications. In the routing, QoS are mainly data rate, end to end delay and jitter. In this paper, AntHocNet routing algorithm [2] has been proposed which is a combination of both reactive and proactive routing algorithms. 2. LITERATURE REVIEW The efficiency and robustness of specific behaviors of ants are the key aspects of their bilological success. Ants are using a hormone known as pheromone [3] for updation and recovery of their route. Due to this property they became an inspiration for design of many routing algorithms for dynamic networks. Ant based algorithms are working on how pheromone is updated, how routing table probabilities are calculated and how often ants send for route request. Many ACO algorithms were proposed for adaptive routing. The first proposed one is ant based control (ABC) [4]. ABC routing was developed for wired networks and it has symmetric path cost in between the nodes. Ants move from one node to another to reach the destination by probability of the pheromone presence in their destination table of the node. When ants reach the destination, they die. Ants may get delayed on a system which is heavily used. It can add noise to avoid freezing of pheromone trails. Routing by ants algorithm (ARA) [4] uses virtual circuits and supports both unicast and multicast routing, which has a similarity with ABC. Difference between them are the parameters which define routing decisions that are obtained from pheromone carried inside the ants, so they can be differenciated from each other. These parameters are added to ants table in their source node and can be calculated using a genetic algorithm. The optimized link state routing(OLSR) is a proactive routing protocol [5]. The optimization of this algorithm is done by the links to diffuse its topology information. In a classic link state, its information is spreaded all over the network. It uses this aproach, but running in wireless enviornment to preserve the bandwidth of the messages. Its optimization is the multipoint relaying. Its main aim is to update the tables. The data obtained in these tables are updated by the traffic which is controlled information retrieved from these tables. It is based on hello, topology, many interface declaring messages. Ad-hoc networking with swarm intelligence (ANSI) [5] is a hybrid routing protocol for ad-hoc networks. It uses both proactive and reactive routing componets. Nodes in ANSI uses only reactive routing and choose routes deterministically, while the nodes belonging immobile networks use a combination of both reactive and proactive routing to perform random probability analysis routing when multiple routes are available. Ad-hoc on demand distance vector (AODV) [6] is a reactive routing protocol and routes are determined when needed. Hello messages may be used to detect the neighbour links. An active node broadcasts a hello message to all its neighbour so that they can receive. If a node fails to receive