Changing Paradigms in Mobile Agenda Ad Hoc Networks : MANET - PDF document

Changing Paradigms in Mobile Agenda Ad Hoc Networks : MANET • Computer Network Classification, IEEE 802 project • Ad-Hoc Networks(MANETs)  Advances in MANET Presented By:  Routing Dr. Shailendra Mishra M. Abdul Rahim Khan  Areas of current research College of Computer & Information Sciences  Research focus Majmaah University Saudi Arabia 4 Evolution Computer Network -Drivers are capacity, data speeds, lower cost of delivery for revenue growth Expected market share TDMA TDMA EDGE EDGE EDGE EDGE Evolution Evolution 3GPP Core GSM GSM GPRS GPRS 90% Network WCDMA WCDMA HSDPA HSDPA PDC PDC CDMA2000 CDMA2000 1x EV/DV 1x EV/DV 10% cdmaOne cdmaOne CDMA2000 CDMA2000 CDMA2000 CDMA2000 1x 1x 1x EV/DO 1x EV/DO 2G First Step into 3G 3G phase 1 Evolved 3G Ad-hoc Networks Ad-hoc Networks • Infrastructured network • Two types of wireless network: • Infrastructured • the mobile node can move while communicating • the base stations are fixed Infrastructure • as the node goes out of the range of a base station, it gets (Wired line) into the range of another base station • Infrastructureless or ad-hoc Radio tower Desktop computer • the mobile node can move while communicating Radio tower • there are no fixed base stations PDA • all the nodes in the network need to act as routers Laptop computer • In Latin “ad-hoc” literally means “for this purpose only”. Then Laptop computer an ad-hoc network can be regarded as “spontaneous network” Pen computer 1

Ad-hoc Networks Ad-hoc Networks  Classification of ad-hoc networks • Infrastructurless (ad-hoc) network or MANET (Mobile Ad-hoc NETwork)  Multi hop – som e nodes are far and • Single hop – nodes are cannot com m unicate directly. The in their reach area and traffic has to be forwarded by can communicate other interm ediate nodes. directly PDA PDA Laptop computer Laptop computer Pen computer Mobile Ad Hoc Networks Fundamental Concepts (MANET) • Ad hoc networks are autonomous networks operating either in isolation or as “stub networks” Backbone connecting to a fixed network Mobile nodes Access points • Do not necessarily rely on existing infrastructure • No “access point” • Each node serves as a router and forwards packets for other nodes in the network • Topology of the network continuously changes MANET Wireless Mobile Network 9 10 Ad-hoc Networks Ad-hoc Networks  Why we need ad-hoc networks? M obile Ad Hoc Networking is a multi-layer problem ! - Security  Setting up of fixed access points and backbone - Service Discovery infrastructure is not always viable - Location-dependent  Infrastructure m ay not be present in a disaster Application Application Layer area or war zone - TCP  Infrastructure m ay not be practical for short- - Quality of Service Transport Layer range radios; Bluetooth (range ~ 10m ) - Routing  Do not need backbone infrastructure support Network Layer - Addressing  Are easy to deploy - Location Management  Useful when infrastructure is absent, destroyed Physical/Link Layer or im practical - Power Control - Multiuser Detection - Channel Access 2

Problems for both Infrastructure Problems and Ad hoc Mode • Communication is only possible between nodes which are directly in range of each other • If nodes move out of range of the access point (Infrastructure Mode) • OR nodes are not in direct range of each other (Ad Hoc Mode) • Then communication is not possible!! What if ?? How can this be done? • ROUTING!! • Wired Networks: Multi-hop Infrastructure Access Multi-hop Ad Hoc Network • Hierarchical Routing • Network is divided into subnets • Nodes look at netmask and determine if the address is directly reachable. If not, just forward to OR the default gateway. • Different protocols for different levels of the hierarchy •RIP, OSPF, BGP Wireless Routing Motivation • Flat routing • Avoid single point of failure typical of centralized • You can’t assume that since a node is in your subnet systems that it is directly accessible • Often unable to rely on existing communications • Node must maintain or discover routes to the infrastructure destination • Desire for a rapidly deployable, self-organizing • All nodes are routers network • Multi-hop packet routing used to exchange messages between users 18 3

Applications Characteristics • Military • Dynamic topology • Rapidly deployable battle-site networks • Heterogeneity • Sensor fields • Bandwidth-constrained variable-capacity links • Unmanned aerial vehicles • Disaster management • Limited physical security • Disaster relief teams that cannot rely on existing infrastructure • Nodes with limited battery life and storage • Neighborhood area networks (NANs) capabilities • Shareable Internet access in high density urban settings • communications among groups of people • Meetings/conferences • Automobile communications (more on this later) 19 20 Standardization ADVANCES IN MANET • Internet Engineering Task Force (IETF) MANET working group (http://www.ietf.org/html.charters/manet-charter.html)  Areas of current research “ The primary focus of the working group is to  Routing develop and evolve MANET routing specification(s)  Cluster management and introduce them to the Internet Standards track. The goal is to support networks scaling up to hundreds of routers. (…) The working group will also serve as a meeting place and forum for those developing and experimenting with MANET approaches.” 21 22 Research focus to date Routing in MANETs • Routing protocols • Why is it different from routing in other types of network? • Reactive, proactive, hybrid • Because both end nodes i.e routers are mobile • Cluster management • Rate of link failure can be high if mobility is high • T o reduce overhead, to facilitate network management, to enable • Unicast and multicast routing problems are being treated QoS, etc. • Quality of service (QoS) • No protocol has been standardized yet (but several • Differentiating among different types of applications under consideration as Internet Drafts at the IETF) • Medium access • Need new metrics to assess the effectiveness of the • Closing the link, recognizing neighbors, scheduling protocol transmission, etc. • Route stability • Other • Control overhead • TCP performance in MANETs, etc. • Data rebroadcast overhead (for multicast) 23 24 4

MANET Routing Protocols MANET Routing Protocols cont.. • Proactive Ad Hoc Routing Protocols • Establish routes in advance Reactive Proactive • Example: Optimized Link State Routing Protocol (OLSR) Table-Driven • Reactive Demand-Driven • Establish routes as needed • Example: Dynamic Source Routing (DSR) DSDV WRP • Less routing overhead, but higher latency in establishing the path AODV LMR DSR ABR • Hybrid CGSR • Proactive within a restricted geographic area, reactive if a packet must traverse several of these areas TORA SSR • Example: Zone Routing Protocol (ZRP) 25 Problems with Routing Distance Vector (Loops)  Is it possible to use standard routing protocols?  Distance-vector protocols  Slow convergence due to “Count to Infinity” Problem  Creates loops during node failure, network partition or congestion  Link state protocols (D, 2) (D, 2)  Use flooding technique and create excessive traffic 1 1 1 and control overhead D A B C  Require a lot of processor power and therefore Dest. Next Metric … Dest. Next Metric … Dest. Next Metric … … … … … … … … … … high power consum ption 3 D B 3 D C 2 D B Distance Vector Distance Vector (Count to Infinity) • DV not suited for ad-hoc networks! • Loops • Count to Infinity • New Solution -> DSDV Protocol (D,5) (D,4) (D,4) (D,3) (D,2) (D,2) 1 1 1 D A B C Dest. Next Metric … Dest.c Next Metric … Dest. Next Metric … … … … … … … … … … 3, 5, … D B 3, 5, … D C 2, 4, 6… D B Metric = Number of Hops to Destination 5