CarrierAssisted Routing in Mobile Networks

1
Carrier-Assisted Routing in Mobile Networks

• Jie Wu
• Dept. of Computer Science and Engineering
• Florida Atlantic University

2
Overview

• Introduction
• Current state of networking
• Mobility as a foe
• Mobility as a friend
• Some Existing Routing Schemes
• Random mobility
• Controlled mobility
• Carrier-Assisted Routing
• Logarithmic store-carry-forward
• Poly-logarithmic store-carry-forward
• Some Final Thoughts
• GENI wireless

3
Current State of Networking

• Scale Internet, cellular, WiMax, WPAN, WLAN,
  Bluetooth
• Types MANETs, sensor nets, vehicular nets,
  underwater nets

4
Wireless Devices

• Current
• Different types PDA, blackberry, IPoD
• Internet connections more wireless
• Node mobility
• (Near) future
• 1 billion vehicles
• 5 billion RFID
• 10-15 billion sensor/embedded devices
• Mobility is at the heart of wireless communication

(a) Edge of the Internet
(b) General way of data transmit
5
2. Mobility as a Foe

• Node mobility is considered to be undesirable in
  MANETs using a connection-based model
• Recovers from and tolerates bad effects caused
  by mobility
• Nodes are assumed to be relatively stable

6
Two Mobility-Mitigation Schemes

• Recovery Scheme
• If a routing path is disrupted by node mobility
  it can be repaired quickly
• E.g., route discovery and route repair
• Tolerant Scheme
• Mask the bad effect caused by node mobility
• E.g., transmission buffer zone and view
  consistency

7
Mobility as a Serious Threat (Wu and Dai, 2004)

• Mobility threatens localized solutions that use
  local information to achieve certain global
  objectives
• Bad decisions occur because of
• Asynchronous sampling of local information
• Delays at information aggregation/propagation
• Movement of mobile nodes
• (Localized solutions are least sensitive to
  mobility compared to global and distributed
  solutions)

8
Local Information

• 1-hop information
• 2-hop information
• 3-hop information

• k-hop information
• Discovered via k rounds of Hello exchanges
• Usually k 1, 2, or 3
• Neighborhood vs. location info.

9
Snapshot

• Snapshot a global state in time-space view
• Energy saving applications
• Virtual backbone sleep mode
• Topology control adjustable transmission range

Virtual backbone (CDS)
Topology control
10
Two Technical Issues

• Link Availability
• How protocols deal with imprecise neighborhood
  information caused by node mobility and delays
• Inconsistent Local Views
• How each node collects and uses local information
  in a consistent way

y
y
y
w
w
w
x
x
x
11
Tolerant Scheme I (link availability)

• A buffer zone is incorporated into existing
  protocols without having to redesign them

12
Tolerant Scheme II (inconsistent local view)

• Consistent Local View
• Each view keeps a version by using a timestamp
• Conservative Local View
• Maintaining a window of multiple views
• New-view(i) F(view(i), view(i-1), view(i-k))
• where F union, max, min,
• (More information on tolerant schemes Wu and
  Dai, IPDPS 2004, INFOCOM 2004, IEEE TMC 2005,
  IEEE TPDS 2006)

13
3. Mobility as a Friend

• Movement-Assisted Routing
• Views node movement as a desirable feature
• Store
• Carry
• Forward

14
Challenged Networks

• Assumptions in the TCP/IP Model are Violated
• Limited End-to-End Connectivity
• Due to mobility, power saving, or unreliable
  networks
• Disruption, Intermittency, and Large Delays
• DTNs
• Delay-Tolerant Networks
• Disruption-Tolerant Networks

15
Sample Networks

• Military networks
• Vehicular networks
• Underwater networks
• DTNs built by sending physical media
• DakNet project in India
• Wizzy Digital Courier project in South America

16
Infostation (Goodman et al, 1997)

• Drive-through walk-through

17
Other Applications

• SWIN (Small and Haas, 2003)

18
Epidemic Routing (Vahdat and Becker, 2000)

• Nodes store data and exchange them when they meet
• Data is replicated throughout the network through
  a random talk

19
Message Ferrying (Zhao and Ammar, 2003)

• Special nodes (ferries) have completely
  predictable routes through the geographic area

20
Some Design Choices

• Contact predictability
• Random
• Implicity (human movement)
• Approximate (highway mobility)
• Precise (underwater)
• Information
• Zero
• Partial
• Complete

• Replication
• Single copy
• Multiple copies
• Copy per node (epidemic)
• Mobility control
• Random
• Partial control
• Total control
• Hop counts
• Constant (2-hop)
• Uncontrolled

21
3. Carrier-Assisted Routing

• Importance of mobility control
• Reduce delay
• Control the number and location of contacts (hop
  counts)
• A model for loose trajectory control of mobile
  nodes
• Compromise of total control (message ferrying)
  and no control (epidemic routing)
• Graasfuaser and Tses seminal work mobility
  increases the capacity of ad hoc wireless
  networks

22
Dual-Control Planes

• M(mobile)-plane mostly store-carry-forward
• S(stationary)-plane store-and-forward
• Two types of nodes carriers and keepers

23
Controlling Hop Count

• Traditional routing (with spatial-diameter-hop-cou
  nt)
• Movement-assisted routing (with
  constant-hop-count)

24
Logarithmic Store-Carry-Forward

• Multiple tracks for carriers
• Laid out in a hierarchical structure
• Contacts are predefined in specific locations
• Trajectory for a carrier
• Follows mostly a circular track
• It can span multiple tracks depending on the need
• Routing processing
• Moving from one carrier to another
• Logarithmic hop count
• The number of carrier changes

25
Eyes and Tracks

• Tracks
• Eye theory (Cang and Wu, 1998)
• Optimal total comm. in time-optimal broadcast
  (with recursive doubling)
• Eye square
• Hierarchical tracks
• Solicited carriers follow tracks (with home)
• Contacts eyes
• Routing
• Mapping 2-D to eye space
• Routing track transitions (log m)

26
Two-Phase Routing

• Up phase
• Moving up the level
• Down phase
• Moving down the level
• Properties
• Forwarding number
• O(log m)
• Distance stretch
• O(m/d) (worst case)
• O(log(m/d)) (average)
• d source and destination distance

27
Routing in Dense Mode

• Features
• On-demand solicitation of carrier
• Fixed trajectory (on tracks)
• Notion of home
• Carrier sharing
• Example
• S1 m1 to D1
• S2 m2 to D2

28
Extensions in Sparse Mode

• Unique Features
• Role changes between carriers and keepers
• Dynamic home
• Dynamic trajectory
• Example
• m1, m3 to D1
• m2 to D2
• m1 and m2 go together

29
Extensions with Random Tracks

• Small-World Model
• Area m m
• D(s,t) 2m
• Short link four grid neighbors
• Long link probability
• cd(u,v)-2 (c is a constant)
• Phase intermediate node u in phase i if 2i lt
  d(u,t) 2i1

30
Navigable Small-World Network

• Long jump (u,v) is used if d(v,t) 2i
• Properties
• Long link O(log m) and short link O((log m)2)
• Total distance long link 5d(s,t) and short
  link d(s,t)

31
Routing with Mobile Destination

• Distance effect
• The precision destination position depends on the
  distance of the current node to the destination
• Lazy location update
• Time and location of its last encounter with
  other nodes.

32
Extensions to High-Dimensional and Non-Euclidian
space

• High-dimensional (s1, s2, , sk)
• Non-Euclidian Frequency space (f1, f2, , fk)
• Mobile node that have similar frequency vectors
  have a higher chance of being neighbors.
• Mapping between two spaces
• Euclidian to Non-Euclidian
• Non-Euclidian to Euclidian

33
Extensions to Other Communication

• Collective Communication
• Broadcast one to all
• Multicast one to many
• Geocast one to all (in a region)
• Anycast one to any (nearby)
• Messaging System
• Publish/subscribe asynchronous messaging
• Mobile pub/sub

34
Integrating the S-plane and M-plane

• Information
• S-plane link state information
• M-plane time-based or frequency-based
• Control Plan vs. Data Plan
• Mobile nodes can carry both data and control
  message (including state information)
• (More information on carrier-assisted routing
  J. Wu, S. Yang, and F. Dai, IEEE TPDS 2007)

35
4. Future of Wireless Networks

• Future world is more wireless
• Convergence to global Internet
• Telecom (PSTN for voice)
• IT (Internet for data)
• Cable TV (Hybrid fiber coax for video)
• New challenges for architecture and protocol
  design
• From top more demand from the end user
• From bottom emerging technologies

36
Some Key Issues Mobility

• Challenges in information dissemination
• Routing and pub/sub
• Naming and addressing
• Location services
• Opportunities in increasing system performance
• Routing capability
• Network capacity
• Security
• Sensor coverage
• Information dissemination (location and data)
• Reducing uncertainty in reputation systems
  (INFOCOM07)

37
Information Service

• Coverage Diversity
• Long-range WiMax
• Medium-range WiFi
• Short-range Bluetooth, Zigbee
• Anywhere, Anytime
• 4Gs MAGIC
• Many-where, Many-time
• Infostation and mobile infostation (e.g. 7DS)

• Coverage vs. Capacity
• Cellular
• High availability, but slow and expensive
• Infostation/WiFi
• Low availability, but fast and low cost
• Information Locality
• Time and space
• Tagged message based on locality

38
Manageability

• Heterogeneity
• Radio technology
• Software radio, MIMO,
• Deployment styles
• Land, air, underwater,
• Devices
• Laptops, PDA, BlackBerry,
• Scales
• Billions of sensors and RFID tags

• Process Orchestration
• Real-time interactions
• Between numerous network entities
• Real-time orchestration
• Computing and network resources

39
Multiple Layers and Cross-Layer

• Multiple Layers
• Radio technology
• Software radio
• Spectrum sharing
• Protocols
• MAC/discovery/routing
• Middleware and overlay
• Mobile P2P
• Mobile pub/sub
• Systems applications
• MANETs
• Sensor nets
• Vehicular/underwater nets

• Cross-Layer Issues
• Mobility management
• Power management
• Dependability and QoS
• Security and privacy
• End-to-end security for mobile devices
• Incentive mechanisms
• Cooperative mechanisms
• Decentralized management
• Self-organization
• Localized solutions

40
NSF FIND Wireless

• Integration with the Internet
• Edge (subset net with border routers)
• Overlay (subset net with a global overlay
  network)
• Integrated (wireless and wired Internet)

41
NSF GENI Wireless

• Different Tools
• Simulators (ns-2, GloMoSim, QualNet, and OPNET)
• Emulators (Emulab and TWINE)
• Testbeds
• PlanetLab (for wired and wide-area)
• ORBIT (for wireless)
• Small vehicular networks (for mobility)

42
Questions