A Fuzzy Content Centric Network Architecture for Real-time Communications in MANETs Niaz Morshed Chowdhury Dr. Lewis M. Mackenzie School of Computing Science.

A Fuzzy Content Centric Network Architecture for Real-time Communications in MANETs

Niaz Morshed Chowdhury

Dr. Lewis M. Mackenzie

School of Computing Science

University of Glasgow

Mobile Ad-hoc Networks (MANETs)

• MANETs:– Can be formed on the fly– Do not require fixed infrastructure– Node can communicate wirelessly

• Practical applications of MANETs include:– Military operation– Disaster recovery– Tactical operation– Conference room

Practical Application of MANETs

• In most practical applications:– One sender – many receivers– Many senders – many receivers

• Effectively makes it…– Group-based communication– Real-time communication– Content sharing system

Multi-constraint Problem

• Conventional approach– Congestion

• Appropriate approach– Congestion– Distance– Mobility– Battery

Content Sharing System

• Who shares?– A sender node to a group of nodes.

• How does it share?– By supplying real-time data to a group of nodes

• Who receives?– An interested node.

• How does it receive?– By notifying sender node about its interest.

Content Sharing System

A Different View…

• Application– stream of real-time data/contents

• Originating node– that starts an application

• Sender node– that can supply an application

• Local node– that receives an application

The way we see it…

The way we see it…

The way we see it…

The way we see it…

The way we see it…

Running on low power…

Gets congested…

Moved away…

Data Structure

• Suitability heap– It holds Sigma for each potential sender– It’s a max-heap

• Node-to-application matrix– It keeps track availability of application at each node

• Requested-application list– Lists all allocation requested by the local node

Calculation of Sigma

Fuzzy System

d

c

mp

For Node ‘X’ in relation to ‘Y’

Sigma

‘X’ is a potential sender node

‘Y’ is the local node

Sigma is a weight

Calculation of Sigma: Distance

Calculation of Sigma: Congestion

Calculation of Sigma: Mobility

Calculation of Sigma: Power/Battery Life

Rules

Calculation of Sigma

Fuzzy System

d

c

mp

For Node ‘X’ in relation to ‘Y’

Sigma

‘X’ is a potential sender node

‘Y’ is the local node

Sigma is a weight

Operation

• When a local node receives request for an application from the user:– It triggers CC-AODV– CC-AODV sends RREQ for specific contents, instead of specific

node (address)– Each node having requested application sends back RREP to the

local node– Based on d, c, m and p, local node calculates Sigma for those

nodes and inserts into the suitability heap.– Finally local node picks root as its ‘sender node’

Maintenance

• If any negative change in root-node’s Sigma occurs,– Root node will be pushed down in the suitability heap– Local node will cross-check current status of new root– If new root is found suitable, local node switch receiving content

from the old root to new root (we call it hand-off)

• A local node,– Can act as sender for other nodes that receives content via it.– Can act as sender upon receiving new request.

Moved away…

Re-structuring

Re-structuring

Evaluation

• 21 node scenario• Custom-built C++ simulation

• Effort– The number of hop a packet needs to travel to reach its

destination. – For example, to transfer a segment over a 5 hop path, 5 times

effort is required.

Performance of Individual Transmission

Performance of Conference Communications

Frequency of Packets on Hops

Future Study

• Extending functionalities for VANET (primary)– Disseminating warning and safety information

• Introducing Reliability (secondary)– Adding transport functionalities

Thank You