A Study On Routing Algorithms For Opportunistic Networks Based On Bargaining Game

Opportunistic Networks are a kind of wireless self-organized networks which have been used in many areas,such as the military,disaster relief and remote wilderness areas on account of tolerating delay and network division.The end-to-end links maybe do not exist between network nodes in Opportunistic Networks.The method to communicate is to utilize meeting opportunity coming from nodes movement.It can realize data forwarding by “store-carry-forward” routing method.Recent years,it is an important trend to study routing algorithms based on Game Theory.Firstly,a research background and summary about Opportunistic Networks is introduced.We assort the existing Opportunistic Networks algorithms on the basis of the Game Theory,and analyze some classical algorithms,especially the algorithms based on Bargaining Games.Then,we propose two kinds of efficient routing algorithms named LPR-DS and RSON and analyze these algorithms in detail.To solve the problems existing in the present probabilistic routing based on Bargaining Games,including redundant control packets exist in the exchange process of SV,redundancy exists in the messages interaction process,and unnecessary overhead from using special data structure to transmit matrix of delivery probability,an routing algorithm based on delivery probabilities piggybacked by SV,LPR-DS,is proposed.By designing improved exchange mechanism of SV,improving the existing interaction mechanism to reduce the times of games,and proposing mechanism of piggybacking matrix of delivery probability with SV,LPR-DS can reduce the control overhead and improve the network throughput.A seller need to pay transaction fee to CCC when an exchange is finished,with the passage of time,which results in declining purchasing power because of the lack of virtual currency in the networks.A buyer doesn't distinguish the difference of valuation of messages when requests messages for a seller,which leads to a raising delay for messages of high valuation.In addition,when buffer space is full,a node need to delete messages to receive new messages,but it's possible to delete some messages that other nodes would just like to request for this node,obviously,which brings unnecessary overhead and delay.To address the problems above,this thesis proposes an routing algorithm based on subsidy for opportunistic networks(RSON).By making up virtual currency for a buyer to settle the problem of the lack of virtual currency in the networks,sorting the message request vector on the basis of valuation of messages,and improving congestion discarded strategy when buffer space is full,With the new mechanisms above,RSON can increase throughput,decrease messages delay,and improve the network performance.Finally,we give a summary of this thesis,then directions for further research of OppNet based on Bargaining Game were presented.