FSR isan inferred proactive routing protocol. It uses the “fisheye” techniqueproposed by klein rock & Stevens FSR4 (where the technique was used toreduce the size of information required to represented graphical data) fishesdo have 3600 vision eye ofthe fish captures with high details the distance near focal point grows.
Based on these concept a table drivenproactive protocol. Fish eye routing generates accurate routing decisions bytaking advantage of the global networks information. It takes the shorterroutes, class sensitive to traffic load, less sensitive to topology.
It iseasier to debug & Account for routes since the entire network topology& route tables are stored at each node. For the above stated reasons tabledriven scheme is chosen for routing. Reactive protocols are distance vectorbased where as proactive protocol like FSR is link & state based and hasfaster speed of convergence & starter lived routing loops. In link statetopology information is disseminated in special link state packages where eachnodes receives a global view of the network rather than the view seen by nodesneighbor. Fisheye routing takes advantages of this feature by implementing anovel updating mechanism to reduce control overhead traffic.
FSRis chosen to be the appropriate protocol for the present study; FSR is a linkstate proactive protocol. Why fish eye protocol is the best?1) Fish eye is a proactive routing protocolwhich maintains routes to all destinations, instead of whether or not theseroutes are needed. The main advantage of this category of protocols is thathosts establish a session.2) Through updating link state informationwith different frequencies depending on the scope distance, FSR scales well tolarge network size & keeps overhead low without comprising routecomputation accuracy when the destination is near.3) As mobility increases, routes to remotedestinations become less accurate. However, when a package approaches itsdestination, it finds increasingly accurate routing instructions as it enterssectors with a higher refresh rate.
4) The control overhead is greatly reducedin FSR average no of neighboring nodes is independent from network size sincenode density is kept constant. The reason why FSR reduces O/H is that only a fraction of the entriesare updated each time. In a two level fisheye hierarchy, the smaller radius,the smaller fraction of entries updated in the “fast” interval & the lowerthe control O/H.5) FSR is more desirable for large mobilenetwork where mobility is high & the BW is low.
By choosing proper no ofscope levels & radius size, FSR proves to be a flexible solution to thechallenge of maintaining accurate routes in ad-hoc network.6) In a wireless environment, a radio linkbetween mobile nodes may experiencefrequent disconnects & reconnects. The Ls protocol release a link stateupdate for each such change, which floods the network & causes excessiveoverhead. FSR avoids this problem by using periodic, instead of event driven,exchange of the topology map greatly reducing the control message overhead.7) FSR scales well to large network size & keepsoverhead low without compromising route computation accuracy when thedestination is near. By retaining a routing entry for each destination, FSRavoids the extra work of finding the destination & thus maintains lowsingle package transformer latency. (As mobility increases, routes to remotedestinatios become less accurate. However, when a package approaches itsdestination, it finds increasingly accurate routing instructions as it enterssectors with a higher refresh rate)8) Throughput of FSR also increases as ituses multilevel fish eye scope.
This technique results in lower overhead &less consumption of BW which is a major plus point for throughput. FSR ifhighly scalable as it uses different frequencies for different scope that is atdifferent time intervals. Anothercharacteristic of FSR is it uses different frequency in exchanging link stateinformation.
(One hop neighbors are classified as scopes) scope is defined interms of the nodes that can be reached in a certain no of hopes. Thecenter nodes has most accurate information about all nodes in the circle &because less accurate with each outer circleEventhrough a node does not have accurate information about distance nodes, thepackage are routed correctly because the route information becomes more &more accurate as the package moves closer to destination The reduction ofrouting messages is achieved by updating the network information for nearbynodes at a higher frequency & remote nodes at lower frequency. As a resultsconsiderable amount of LSPs are suppressed.FSRreduces significantly the consumed BW as the link state updating packets areexchange only among neighboring nodes.
The routing overhead is also reduced dueto different frequencies of updates among nodes of different scopes. FSR manages to reduce the message size of thetopology information due to removal of topology information concerned far awaynodes.