Deviation Correct Multiplier Angle Method MAM

1
Orthogonal Rendezvous Routing Protocol (ORRP) for
Wireless Mesh Networks Bow-Nan Cheng (RPI),
Murat Yuksel (UNR), Shivkumar Kalyanaraman (RPI)
ORRP Details
Analysis

• Motivation
• The growth of wireless technologies bring about
  major scalability issues as network size and
  density increases. One issue is the usage of the
  medium as a network becomes denser, the medium
  becomes more saturated. Yi et al. showed
  analytically in 2005 that by using directional
  antennas that limit the spread to 1/8th of the
  original omni-directional spread, capacity
  increases by a factor of 50x.
• Additionally, current RF-based ad hoc networks
  utilize omni-directional antennas, consume high
  power, are constrained by low bandwidth, and are
  highly error-prone. By contrast, free space
  optical (FSO) transceivers such as
  high-brightness LEDs (HBLEDs) are very low cost
  (2-5/transceiver package), highly reliable (10
  year lifetime) and low power (100 microW for
  10-100 Mbps), operate in license-free frequency
  bands, transmit at relatively higher bandwidth,
  and are more secure and spatially efficient due
  its transmissions directional nature.
• . Given the theoretical capacity gains and
  potential for higher bandwidth by using
  directional forms of communication, it becomes
  interesting to investigate how directionality can
  be used to route packets in wireless networks.

Reach Probability Analysis
Proactive Element (1) Paths from Rendezvous
Nodes-to-Destination Nodes are formed by
periodically sending announcement packets out
orthogonal directions
Reactive Element (2,3) Paths from Source
Nodes-to-Rendezvous Nodes are formed by sending
route request (RREQ) packets and waiting for
route reply (RREP) packets. These RREQ packets
are sent on demand
Deviation Correct Multiplier Angle Method (MAM)
Punreachable Pintersections not in
rectangle
The Multiplier Angle Method (MAM) is used to help
correct path deviations, route around voids, and
route along perimeters.
Stretch Analysis
Simulation Results
Trends From Flood-based to Unstructured Scalable
Hierarchy/Structured
Flood-based
Unstructured/Flat Scalable
Position-based protocols are considered highly
scalable but rely on coordinate space embeddings.
They often also do not consider overheads
associated with node localization (position to
coordinate mapping) and ID to location mapping
(via location services).
Research Areas in Position-based Schemes
(15,5)
Evaluation Metrics / Scenarios
(4,6)
D
(8,5)
S
D(X,Y)?
(0,4)
(12,3)
Metrics

• Reachability Percentage of nodes reachable by
  each node in network
• State Complexity The total state information
  maintained in the network
• Path Stretch Average ORRP Path vs. Shortest
  Path
• Goodput
• End to End Delay (Latency)

(5,1)
By removing node localization (coordinate space
embedding), can we still route packets
effectively?

• Conclusions
• ORRP achieves high reachability (98) in random
  topologies
• ORRP achieves N3/2 state maintenance much more
  scalable than traditional topology-based
  approaches
• ORRP achieves roughly 1.2 path stretch
  (Tradeoff for connectivity under relaxed
  information is very small!)
• Optimum number of interfaces for ORRP is roughly
  8
• ORRP end-to-end delay is much less even though
  paths are longer due to more efficient use of the
  medium
• ORRP achieves roughly 30x goodput vs. AODV, and
  10x goodput vs. OLSR

ORRP Introduction
Numerical Simulations (without MAM)
ORRP Design Considerations

• ORRP Primitive
• Local sense of direction
• leads to ability to forward
• packets in opposite
• directions

Rendezvous Points
A

• Reachability upper bound
• State information maintained at each node State
  distribution network-wide
• Average path stretch

• Advantages
• Connectivity under lessened/relaxed information
• Efficient medium reuse
• Heightened scalability (due to less state info)
• Even distribution of state information (no
  single point of failure)
• Trade-offs
• Path inefficiency (not always shortest path)

Packetized Simulations with NS2 Scenarios

• Effect of number of interfaces on density, voids,
  topologies
• Effect of TTL on density, voids, topologies
• Effect of number of lines on various topologies,
  latency, goodput
• Comparison vs. AODV (reactive), OLSR (proactive),
  and GPSR w/ GLS (position)

B

• Relevant Publications
• B. Cheng, M. Yuksel, S. Kalyanaraman,
  "Orthogonal Routing Protocol for Wireless Mesh
  Networks," Proceedings of IEEE International
  Conference on Network Protocols (ICNP), pp
  106-115, Santa Barbara, CA, November 2006.
• B. Cheng, M. Yuksel, and S. Kalyanaraman,
  Directional Routing for Wireless Mesh Networks A
  Performance Evaluation, Proceedings of IEEE
  Workshop on Local and Metropolitan Area Networks
  (LANMAN), Princeton, NJ, June 2007.
• B. Cheng, M. Yuksel, and S. Kalyanaraman,
  Rendezvous-based Directional Routing A
  Performance Analysis, To appear in Proceedings
  of IEEE International Conference on Broadband
  Communications, Networks, and Systems
  (BROADNETS), Raleigh, NC, September 2007.
  (invited paper)
• This material is based upon work supported by the
  National Science Foundation under Grant Nos.
  IGERT 0333314, ITR 0313095, and STI 0230787. Any
  opinions, findings, and conclusions or
  recommendations expressed in this material are
  those of the author(s) and do not necessarily
  reflect the views of the National Science
  Foundation.

• ORRP is based on two simple primitives
• Local directionality is sufficient to maintain
  forwarding of a packet on a straight line
• Two sets of orthogonal lines in a plane intersect
  with high probability even in sparse, bounded
  networks

State Complexity Analysis