HAZY SIGHTED LINK STATE ROUTING PROTOCOL Eleonora Borgia

HAZY SIGHTED LINK STATE ROUTING PROTOCOL Eleonora Borgia Pervasive Computing & Networking Lab. Per. Lab IIT – CNR eleonora. borgia@iit. cnr. it Mobile. MAN Project - Helsinki – June 7/8, 2004

Main Target Identify a routing protocol suitable for a cross-layer architecture: • in term of scalability, performance and efficiency; • providing more advantages to the other protocols. q MANET standpoint: reactive protocols are more efficient than proactive ones for the introduced overhead q OUR standpoint: ü The protocol overhead cannot be computed in isolation; ü New cross-layer metrics must be applied to compute it; ü It’s useful having a knowledge (at least partial) of the network topology, also to provide it to other protocol layers. proactive protocols are more suitable

Overview of Link State Routing In literature there are different approaches aimed to reduce the overhead introduced by protocols based on LS: • Efficient Dissemination : updates sent throughout the network more efficiently (e. g. OLSR, TBRF, STAR) • Limited Dissemination : restriction of the scope of routing updates in space and time (e. g. hierarchically LS, FSR, GSR, HSLS)

Total overhead (1) TRADITIONALLY: OVERHEAD CONTROL OVERHEAD Amount of bandwidth require to construct and maintain routes • PROACTIVE OVERHEAD: number of packets exchanged between nodes in order to maintain node’s routing table • REACTIVE OVERHEAD : consumed bandwidth for Request/Reply messages • As N increases, keeping route optimality is much expensive particularly in LS; • SUB-OPTIMAL paths may be used. The impact of SUB-OPTIMAL routes and its overhead must be taken into account.

Total overhead (2) NEW DEFINITION * : TOTAL OVERHEAD of protocol X is the sum of: 1. PROACTIVE OVERHEAD; 2. REACTIVE OVERHEAD; 3. SUB-OPTIMAL OVERHEAD : Difference between the bandwidth consumed using the sub-optimal paths and that eventually consumed if the data had followed the shortest available path(s). S A D B Ex: A = 3 hops B = k hops SUB-OPT ov = (k-3)*Packet_length (*) C. Santivanez, I. Stavrakakis et al, “On the scalability of Ad Hoc Routing Protocols”, INFOCOM 2002

Hazy Sighted Link State (HSLS) (1) • Proactive protocol with Limited Dissemination; • Scope’s restriction: LSU packets are broadcasted periodically over the network with frequencies fi that decreases with the distance from the node itself; • Each node has a partial knowledge of the network (i. e. not real-time uploaded): ü detailed view of 1 -hop neighborhood; ü “hazy” knowledge of distant nodes. • HSLS is based on the observation that nodes that are far away not need to have a complete topological information in order to make a good next hop decision

Hazy Sighted Link State (HSLS) (2) ü te: : period with which a LSU is sent through the network ü TTL = 2 i 1. A node wakes up every te sec and sends a LSU with TTL=2 if there has been a link status change in the last te sec; 2. Every 2 i-1 * te sec (with i=1, 2, 3. . ) a node wakes up and sends a LSU with TTL= 2 i if there has been a link status change in the last 2 i-1 * te sec; 3. Every tb sec (tb>te) a global LSU (TTL= ) is sent in the entire network to give a complete overview of the network topology, even if there’s no link changes. High mobility scenario:

HSLS Example Network size: 5 hops G F T V W X A B H N S E C R D M Z Q U L LSU packet O Originator ID ID TTL 2 4 TTL == Neighbor ID 1 J K T = 42 te : P Neighbor ID 2 Broadcast transmission

Comparative study* Routing Protocol Total_overhead SLS (proactive) (N 2) DSR (reactive) (N 2+N 2*log 2 N) ZRP (hybrid) HSLS (N 2) (N 1. 5) (*) C. Santivanez, I. Stavrakakis et al, “On the scalability of Ad Hoc Routing Protocols”, INFOCOM 2002

Mobile. MAN project § HSLS software architecture design n Ad Hoc Framework (HUT) Interface with the cross-layer architecture and cooperation with other protocol layers Possible integration with HUT Ad Hoc framework HSLS Subsystem Proactive Subsystem Hybrid Subsystem Reactive Subsystem Test and Performance evaluation Common Modules

Software Architecture HSLS Initialization Packet Mangement Garbage Collector LSU Socket Management Processing NE ST Communication Hello