Documents
Resources
Learning Center
Upload
Plans & pricing Sign in
Sign Out

A_Pullin-am_4

VIEWS: 20 PAGES: 15

									Using Ship Movement is the Irish
Sea for MANET Evaluation

                  Adrian J Pullin
           Liverpool Hope University
               pullina@hope.ac.uk
                Dr. Steve Presland
           Liverpool Hope Unviersity
               preslas@hope.ac.uk
              Prof. Colin Pattinson
         Leeds Metropolitan University
             c.pattinson@lmu.ac.uk


                                         1
     Outline
Mobile Ad Hoc Network (MANET) overview
Marine Automatic Identification System (AIS)
The model
The experiments
Results
Conclusions
Future developments
                                                2
      Mobile Ad Hoc Networks
 Wireless nodes acting as hosts and routers
 No fixed infrastructure
 Nodes are mobile and can join and leave at will
 Network is self configuring
 Applications include disaster zone, vehicular and
  battle field
 Issues include dynamic nature of network, frequent
  route changes, limited power at nodes
 Most research uses simulations that require real
  world models

                                                       3
         Automatic Identification
         System (AIS)
 Legal requirement for commercial ships
 Transmits information on marine VHF band between every 2
  seconds (moving) to 3 minutes (stationary)
 Data includes:
      Ship identification
      Location
      Speed
      Direction
      Destination
      Etc.
 Publicly accessible with low cost receiver and software
 Web site (www.shipais.com) has displays for UK

                                                             4
www.shipais.com




                  5
       AIS Data
 Mr Ian McConnell, webmaster of www.shipais.com supplied data
  (thanks)
 339300130;under way; 250’; 14.5kt; 53.988500N; 3.005500W; 220.0°;
  215°; 57s; 070303 000001;serial#1(B)[1]
 Maritime Mobile Service Identity
 Status
 Rate of turn (° per minute)
 Speed over ground
 Location
 Course over ground
 True heading
 Time stamp
 Channel (A or B)


                                                                      6
      ns-2 movement files
 Simulator (ns-2) uses movement file
 $ns_ at 15 "$node_(4) setdest 266.738403 158.395828
  0.000051”
 At time 15 (seconds from start of simulation)
 For node 4 (each ship is mapped to a different node)
 Start moving to this (x,y) location
 At this speed
 Translation from AIS to ns-2 by some very dirty C
  programming
 Scaled down by 1000 because ns-2 has problems with very
  large areas


                                                            7
    World Modelled
North Irish Sea N52° to N55°, W2° to W7°




                                            8
    Model Verification
Results from ns-2 runs are fed into the
 Network Animator (nam) that comes with ns-
 2 and compared with the same period on
 www.shipais.com
Confirms that ship movements are correctly
 reproduced in ns-2 simulation



                                              9
        Simulation Experiments
 Trying to establish contact through MANET to each of four
  Vessel Traffic Service (VTS) stations
     Elsam Traffic in Dee estuary, Snaefell (Isle of Man), South Stack
      (Anglesey) and Crosby
   Using constant bit rate traffic
   Using one routing protocol (DSDV)
   Using a number of different radio transmission ranges
   Question “Can a MANET be established in this scenario?”
   Question “How much does it cost?”
     Routing overhead


                                                                          10
       Results – The Good News
       Packet Delivery Ratio
         Elsam     Snaefell   South        Crosby
         Traffic   node 1     Stack node   node 3
         node 0               2
5km      3%        0          0            77%
10km     97%       0          0            97%
15km     98%       0          97%          99%
20km     99%       0          99%          99%
25km     99%       0          99%          99%
30km     99%       99%        99%          99%
                                                    11
       Results – The Bad News
       Normalised Routing Load
         Elsam     Snaefell   South        Crosby
         Traffic   node 1     Stack node   node 3
         node 0               2
5km      3913%     0          0            172%
10km     153%      0          0            153%
15km     174%      0          174%         168%
20km     144%      0          146%         145%
25km     135%      0          132%         133%
30km     128%      128%       128%         128%
                                                    12
      Conclusions
 It is possible to build a model based on AIS data for
  use in ns-2 simulator
 It is possible to establish MANET connections in the
  North Irish sea using DSDV as the routing protocol,
  providing the radio transmission range is high
  enough
   30km for Snaefell
   15km for other VTS
   Within the ranges available to these VTS
 The routing overheads using DSDV are prohibitive
   Usually expect in the order of 1%-2%, not over 100%

                                                          13
     Future Work
More reliable and robust AIS data sets
Different geographical areas
Repeater masts
Use model to evaluate different MANET
 routing protocols



                                          14
Questions




            15

								
To top