Soutenance mémoire
Ad-hoc Networks Basics
Wireless ad-hoc networks are data networks that are deployed without a fixed infrastructure or central controllers such as access points or base stations . In these networks, data packets are forwarded directly to the destination, if it is in the transmission range of the sender, or indirectly, through a multi-hop path of intermediary nodes that act as relays.
This paradigm where a fixed infrastructure is not needed, is tolerant to topology changes and allows a fast deployment, is suitable for a large number of network implementations, such as mobile handheld devices, wireless sensors, disaster recovery networks, tactical networks and many others.
The versatility, low initial setup cost and ease of deployment allowed several manufacturers to enter into the ad-hoc networking field with different products and applications. Although they bring many advantages, ad-hoc networks face several issues particularly when the size of the network increases considerably, also know as scalability issues and when the velocity of the nodes increases from now on, mobility issues.
Design and Research Challenges
Security Issues and Challenges: To have secure communications among mobile nodes is a priority issue. During years security has been an active research theme. MANETs present several new challenges for researchers and developers like open network architecture, shared wireless medium and highly dynamic network topology. Nevertheless, the existing security measures do not cover MANETs and it is necessary the implementation of news security solutions. The routing protocols for wireless ad-hoc networks should be able to solve all these issues efficiently;
Bandwidth constraint: The broadcast condition is inherent to ad-hoc networks. Therefore, the available bandwidth per wireless link depends on the number of nodes and the traffic they handle. Thus, only a fraction of the total bandwidth is effectively available for transmission and reception in every node. On the other hand, mobility of the nodes makes established routes to broke thus route reparation procedures have to run, these procedures also consume part of the available bandwidth so they should be carefully designed to reduce the impact on the effective bandwidth available;
Location-dependent contention: The contention for wireless channel increases as the number of nodes increases. The high contention implies a high number of collisions and subsequent bandwidth wastage. A routing protocol for wireless ad-hoc networks should have mechanisms to distribute the network load uniformly across the network;
Other challenges: Computing power, battery power, and buffer size also limit the capability of a routing protocol for ad-hoc networks.
Smart antennas technologies
Smart antennas are most often realized with either switched-beam or fully adaptive array of antennas. An array consists of two or more antennas (the elements of the array) spatially arranged and electrically interconnected to produce a directional radiation pattern. In a phased array the phases of the exciting currents in each element antenna of the array are adjusted to change the pattern of the array, typically to scan a pattern maximum or null in the desired direction. Although the amplitudes of the currents can also be varied, the phase adjustment is responsible for beam steering. Smart antennas have the potential to increase the performance of wireless networks in general, but especially in ad-hoc networks as they can provide extended range coverage, better spatial reuse, lower energy consumption and increase system capacity.
Neighborhood Discovery
One of the distinctiveness of MANETs is their self-configuring capability, which means that the network does not need a centralized unit to perform the tasks required to keep the network functioning. Neighborhood discovery is one of the self-configuring tasks that is performed once nodes are deployed, and it allows each node to discover its surrounding neighbors. This information is then used by the upper layer protocols such as topology control, medium access control and routing protocol to perform their own tasks.
In recent years, a significant amount of research has been done in this area. For example, Vasudevan et al. (2005); McGlynn and Borbash (2001); Keshavarzian et al. (2004); An and Hekmat (2007) considered the type of antenna used. We gather from these articles that there are two types of neighbor discovery algorithms depending on which antenna type is used: om-nidirectional and directional neighbor discovery. The authors in McGlynn and Borbash (2001); Keshavarzian et al. (2004) presented two algorithms for neighbor discovery in wireless ad-hoc networks where nodes have omnidirectional antennas. While the algorithm proposed in McG-lynn and Borbash (2001) can operate in asynchronous mode, synchronization is a requirement for the algorithm described in Keshavarzian et al. (2004).
Previous research also proposed neighbor discovery algorithms using directional antennas. In Vasudevan et al. (2005), the authors developed several algorithms considering three approaches: directional transmission and omnidirectional reception (DO); directional transmission and reception (DD); and omnidirectional transmission and directional reception (OD).
Routing Algorithms
Since the first development of routing protocol for mobile ad-hoc networks, researchers have shown their interest to support the routing decision based on the position of a node to increase the packet delivery ratio and reduce the end-to-end delay. In the literature, there are several implementations of both reactive and proactive routing protocols. In Ko and Vaidya (2000) the authors proposed a reactive routing protocol called Location Aided Routing (LAR) that uses the position information obtained by a GPS device to bound the search of a route to the defined request zone. This zone is defined by the expected location of the destination node at the time of the route discovery. Simulation results indicated that using location information resulted in significantly lower routing overhead, as compared to other algorithms that do not use location information. However, drawbacks of this approach are the need for GPS and the reactive nature of the protocol which increases the setup delay of a route.
Basagni et al. (1998) proposed a proactive protocol where the nodes need the position infomation from a GPS to calculate the route to destination. The entire network was divided into hierarchical zones within which the information about the position of the nodes is distributed.
As in Ko and Vaidya (2000), this solution also depends on the GPS devices, besides the mobility of the nodes between zones creates an unnecessary overload. In Quintero et al. (2007) the authors developed Location-Enhanced On-Demand (LEOD) routing protocol, a framework that uses the information provided by smart antennas to determine the position of the nodes in the network and using this information, discover and maintain routes. The positioning algorithm only considered one hop neighbors and the method used to estimate the position is based on the angle of arrival technique.
|
Table des matières
CHAPTER 1 INTRODUCTION
1.1 Problem Statement
1.2 Objectives
1.3 Methodology
1.4 Thesis Contributions
1.5 Thesis Outline
CHAPTER 2 LITERATURE REVIEW AND BACKGROUND
2.1 Introduction
2.2 Ad-hoc Networks Basics
2.2.1 Applications of Ad-hoc Networks
2.2.2 Design and Research Challenges
2.2.3 Requirements
2.3 Smart antennas technologies
2.3.1 Introduction
2.3.2 Benefits of Smart Antenna Technologies
2.3.3 Foundation of smart antenna technology
2.4 Literature Review
2.4.1 Modeling ad-hoc networks
2.4.2 Neighborhood Discovery
2.4.3 Routing Algorithms
2.4.4 Energy Models
CHAPTER 3 MODELING DIRECTIONALITY IN WIRELESS AD-HOC NETWORKS
3.1 Introduction
3.2 Graph theory and Ad-hoc Networks
3.3 System Model
3.3.1 Node Distribution Model
3.3.2 Network Topology
3.3.3 Antenna model
3.3.4 Wireless Channel Model
3.4 Simulation Setup
3.5 Analysis of Results
3.5.1 Link Probability
3.5.2 Hop count
CHAPTER 4 NEIGHBOR DISCOVERY AND ROUTING SCHEMES FOR AD-HOC NETWORKS
4.1 Introduction
4.2 Overview
4.2.1 Neighbor Discovery
4.2.2 Routing Algorithms
4.3 System Model
4.3.1 Antenna Model
4.3.2 Beamforming
4.4 Proposed Solution
4.4.1 Neighbor discovery
4.4.2 Directional Routing
4.4.3 Route Discovery
4.4.4 Route Establishment
4.5 Simulation Setup
4.6 Results Analysis
4.6.1 Average end to end delay
4.6.2 Packet loss fraction
4.6.3 Throughput
CHAPTER 5 ENERGY MODEL FOR DIRECTIONAL AD-HOC NETWORKS
5.1 Introduction
5.2 Proposed Model
5.2.1 Node Distribution Model
5.2.2 Network Topology
5.2.3 Antenna model
5.2.4 Wireless Channel Model
5.2.5 Graph Representation Matrix
5.2.6 Routing scheme
5.3 Energy Cost of Communication
5.3.1 Omnidirectional Energy Cost
5.3.2 Directional Energy Cost
5.4 Energy model for messages transmitted
5.5 Simulation Setup
5.6 Results Analysis
CONCLUSION AND RECOMMENDATIONS
Télécharger le rapport complet
