Passage Choice in Provincial Remote Cross section Systems Group: Lara Deek, Arvin Faruque, David Johnson.


58 views
Uploaded on:
Description
BATMAN has propelled a bit further. GW can publicize downlink speed ... ( Week 1 - Week 4) Have assessed BATMAN on 3 lattice UCSB MeshNet hubs. Need to move ...
Transcripts
Slide 1

Door Selection in Rural Wireless Mesh Networks Team: Lara Deek, Arvin Faruque, David Johnson http://www.octavetech.com/blog/wp-content/transfers/2008/03/long-go wireless.jpg

Slide 2

Introduction: Rural Wireless Mesh Networks (WMNs) ‏ A lattice system involved various, ware gadgets that gives Internet access to provincial regions Topology varies from center and-talked remote systems Applications: Education, human services Benefits: cost, power, base necessity

Slide 3

Introduction: Rural WMN Examples Digital Gangetic Plains (India) ‏ OLPC Project: Each XO-1 will work as a WMN hub Image from http://www.cse.iitk.ac.in/clients/braman/dgp.html Image from http://laptop.org/en/tablet/equipment/specs.shtml

Slide 4

Introduction: Mesh Network Gateway Selection Mesh systems associate with whatever remains of the Internet by means of entryways Rural and city WMNs have distinctive transmission capacity requirements Municipal: bottleneck is remote connections Rural: bottleneck is at passages Problem: Inefficiently used portals WMN can have extreme outcomes in country zones Our objective: alter a current cross section directing convention endeavor to ideally choose portals

Slide 5

B.A.T.M.A.N.(1)‏ A B F D G C X E A needs to achieve X

Slide 6

B.A.T.M.A.N. (2) ‏ Nodes communicate originator messages (OGM\'s) each second OGM\'s are rebroadcast Other hubs measure what number of OGM\'s are gotten in a settled time window A:10 A B F D A:9 G C X E

Slide 7

B.A.T.M.A.N. (3)‏ A B F A:8 D G C A:7 X E A:7 D BATMAN steering table TO VIA Q A B 8 A C 7 D Final directing table TO VIA A B

Slide 8

B.A.T.M.A.N. (4)‏ A B F A:0 A:6 D G C A:4 X A:7 E G BATMAN steering table TO VIA Q A D 6 A E 7 G Final directing table TO VIA A E

Slide 9

B.A.T.M.A.N. (5)‏ A B F D G A:5 C X A:6 E X BATMAN directing table TO VIA Q A G 5 A E 6 X Final steering table TO VIA A E

Slide 10

B.A.T.M.A.N. (6)‏ A B F D G C X E X BATMAN directing table TO VIA Q A G 5 A E 6 E BATMAN steering table TO VIA Q A C 7 A D 4 C BATMAN steering table TO VIA Q A A 9

Slide 11

Current GW choice methods Minimum jump number to entryways Used by directing conventions like AODV Creates single over congested portals GW2 A B F D G C X E GW1

Slide 12

Current GW choice procedures Best connection quality to GW Used by source steering conventions like MIT Srcr Link state conventions like OLSR Prevents congested connections to GW Not worldwide ideal of GW BW utilization GW2 2 2.2 A B F 1.5 1 3 D G C 1 X 1 E GW1

Slide 13

Current GW determination strategies BATMAN has propelled somewhat facilitate GW can promote downlink speed User can pick GW choice in light of GW with best BW Stable GW (need history) ‏ GW BW x LQ Can\'t trust publicized GW BW Doesn\'t accomplish reasonableness 256 kbps GW2 9 10 A B F 7 4 7 3 D G C 8 7 X E 10 512 kbps GW1

Slide 14

Proposed Solution: Introducing insight to the center of the WMN Introduce data about door execution into the system Nodes at "knowledge limit" have passage execution data, need to exchange this data to alternate hubs Transfer this data by means of: "Batsignal" bundles that are overflowed through the system

Slide 15

Proposed Solution: What does the limit hub measure? At the point when hubs will choose passages, they should evaluate the measure of transfer speed they will get: Example: Hence, limit hubs must transmit current aggregate entryway transmission capacity and current # of VPNs Total portal limit is the entirety of Measured additional transfer speed (measured through dynamic tests) The total of the present transmission capacities of the VPNs

Slide 16

Field Description GWID Gateway ID (0-255) ‏ TS Time stamp DB Total download data transmission VPNs Number of VPNs on door TTL Packet time to live Proposed Solution: Batsignals A hub at the knowledge limit occasionally Record portal estimation If the estimation is not definitely not quite the same as a past worth, then transmit a Batsignal bundle just in the event that we have not as of late transmitted a batsignal parcel If the estimation is radically not quite the same as a past worth, instantly transmit a Batsignal parcel All different hubs Forward a got bat-sign to its neighbors (on the off chance that it has not expired)‏ Update their own entryway inclination tables Batsignal Packet Node Gateway Preference Table

Slide 17

Proposed Solution: Using Batsignal information to pick a door Gateway Preference Table To pick a passage, the accompanying metric in view of table information and connection quality (figured just when current_time - timestamp is underneath an edge) is utilized Gateway fluttering: When a door comes up and goes down every now and again, a substantial number of clashing Batsignal\'s will be communicated to the WMN hubs. The VPN won\'t change to another door until every one of the streams inside it have ended (Srcr)

Slide 18

Evaluation: UCSB Meshnet status

Slide 19

Evaluation: The huge lattice in South Africa 7x7 matrix of 49 remote hubs utilizing 802.11 a/b/g radios Each hub system boots off a focal server Makes utilization of 30dB attenuators on radios to accomplish different bounces in little space Has been utilized for broad cross section system convention benchmarking Complete remote control of tests conceivable

Slide 20

Evaluation Environment I Parameters at the Gateway and Mesh Nodes Technologies Used Load: activity/blockage. Misfortune: signal shortcoming, impediments. Delay: . Data transfer capacity: of the accessible correspondence channels between lattice hubs or between cross section hubs and doors. Throughput: between cross section hubs and a test server outside the lattice system. tc: linux movement control. iperf : TCP/UDP data transfer capacity estimation apparatus. iptables : characterizes bundle preparing plans.

Slide 21

Evaluation Environment II Metrics Measurement Methodology Gateway productivity: measures how adequately we coordinate the throughput created by the VPNs to the limits of the passages. Portal decency: measures how reasonably the total door throughput is disseminated among VPN streams. Passage Flapping: measures the recurrence a cross section hub switches between use of different portals. Measure VPN streams at each GW Have limit of all GW\'s. Measure VPN streams. What is the time window? Normal after some time. Parse BatSignals for every hub and record the timestamp for each GW utilization. What amount of hysteresis?

Slide 22

How are we utilizing advances to decide basic parameters? Dynamic Probing to decide GW throughput utilizing a decentralized, appropriated approach by means of trusted web network hubs that shape the insight limit {B1, B2} .

Slide 23

Current Progress (from Proposal) We are in Week 4 . Figure an arrangement of preparatory assessment measurements for the convention. (Week 1 - Week 3). Done Formulate an estimation strategy to test the adequacy of the convention. (Week 1 - Week 2) Done Emulate a portal on a UCSB MeshNet hub utilizing Linux devices, for example, tc and iptables. (Week 2 - Week 3) Have created scripts to control TC and iptables. Need to create remote control for this script. Run and assess the most recent designers arrival of B.A.T.M.A.N. on the UCSB MeshNet. (Week 1 - Week 4) Have assessed BATMAN on 3 network UCSB MeshNet hubs. Need to move monstrous cross section (has been done some time recently). Actualize answers for Goals 1, 2, 3, and 4 and measure execution utilizing the estimation procedure portrayed as a part of (2) and assessment measurements depicted in (1) (Week 3 – Week 6) In advancement, examining code.

Slide 24

Nifty Animations

Recommended
View more...