Research On Data Transmission Strategies In Opportunistic Networks

Wireless ad hoc network is constructed by the self-organizing and multi-hop manner and does not rely on pre-existing infrastructure,which has been used in many practical environments due to flexible networking and rapid expandability.However,in some application environments,the transmission path of wireless ad hoc network may appear intermittent connectivity because of low density,node mobility and limited communication capability of node.In order to achieve data transmission in these scenarios,opportunistic network is presented.Opportunistic network is a type of special wireless ad hoc network,which exploits encounter opportunities of mobile nodes and follows “store-carry-and-forward” manner to achieve data communication.This communication manner has significant impact on routing design.Hence,this thesis focuses on several key issues about data communication in opportunistic network,and its main contents are listed as follows.First,this thesis proposes an opportunistic routing with admission control for different types of flows.In the traditional routing of wireless ad hoc network,if one node at the optimal path is broken because of congestion or exhaustion of available energy,the whole path will be broken,and the source must re-route again.Hence,how to design an efficient data forwarding scheme that can provide better Qo S for different user requirements is a challenging issue,espically for different types of flows in resource-constraint opportunistic networks.Therefore,in this paper,we present a novel OR scheme which considers admission control of nodes for the different types of flows,namely ORAC.ORAC scheme first exploits a new flow admission control scheme which is based on bandwidth,backlog traffic and residual energy of nodes to select forwarding candidates.And then,ORAC uses a novel forwarding scheme to provide services for multiple different types of flows.In this way,ORAC scheme can reduce the network congestion,and protect from any changes of traffic load for established routes.Thus,it can provide better Qo S for different types of flows.Our extensive simulation results demonstrate that the proposed scheme ORAC can achieve better network performance compared to existing OR schemes in terms of transmission delay,and bandwidth-aware OR with admission control in terms of system throughput,flow acceptance ratio,average delay,overhead,energy consumption,and network lifetime.Second,a novel routing scheme based on Nash bargaining solution in resource-constraint opportunistic networks is proposed.Existing routing schemes in opportunistic networks can be classified into two categories: single-copy based and multi-copy based.Single-copy based schemes can save network resources,however,they could not guarantee transmission delay and packet delivery ratio.Multi-copy based schemes can increase the robustness of networks and reduce average delay,while they consume a large amount of network resources and induce large overhead.Hence,how to design an efficient routing scheme which can make a better tradeoff between the performance of multi-copy and single-copy in resource-constraint opportunistic networks is a challenging issue.Inspired by this,in this paper,we propose a novel routing scheme based on Nash bargaining solution in resource-constraint opportunistic networks,named Game R,which is a cooperative scheme.In Game R,we map the distributed message dissemination in opportunistic networks to a multi-player bargaining model,by constructing a proper utility function to achieve efficient message transmission.Our scheme can efficiently utilize the network resources,achieving higher delivery ratio and lower overhead in resource-constraint opportunistic networks.Extensive simulation results show that Game R performs better compared to existing classical single copy schemes,multi-copy schemes,and quota-based schemes in terms of message delivery ratio,average delay,overhead,goodput,and the overall performance.Third,this thesis proposes an incentive aware routing based on game theory for selfish opportunistic networks,named as IAR-GT.Nodes in the networks may be selfish due to limited resources or poor social tie,and the nodes' selfishness can be classified into individual selfishness and social selfishness according to whether they consider the social relationship between nodes.The existing works only consider how to solve the individual selfishness or social selfishness for data transmission respectively.But,the individual selfishness and social selfishness may co-exist in OPPNETs,which have a significant effect on network performance.Inspired by this,in this paper,we propose an incentive aware routing based on game theory for selfish opportunistic networks,IAR-GT,which maps the message transmission between two selfish nodes as a Rubinstein-Stahl bargaining game.The proposed scheme employs virtual currency and constructs proper price function which not only considers the nodes resources but also considers the nodes' social ties for stimulating selfish nodes cooperation in data transmission.Finally,we use trace-driven simulations in selfish opportunistic networks,and compare IAR-GT with three classical routing schemes and an incentive based routing in terms of message delivery ratio,average delay,overhead,and the overall performance.The simulation results show that IAR-GT can achieve better performances in delivery ratio,average delay and overall performance.Finally,a novel secure routing based on social trust is proposed for opportunistic networks attacked by the malicious nodes,named as SSNCR.There is no a complete path between source and destination nodes,which has some unique features,e.g.,link changes frequently and delay tolerant,and utilizes the contact opportunities of the mobile nodes to achieve data forwarding in OPPNETs.In this scenario,the network may be attacked by the malicious nodes,resulting in degrading the whole network performance.To solve this problem,we propose a social trust based secure routing protocol,SSNCR.The scheme exploits a new metric which is based on states and social relationship of nodes for secure routing,and then forwards messages among trust nodes selected according the metric.In this way,the scheme can protect against promise-then-drop,trust-boosting and defamation attacks.In addition,by using network coding in data transmission among trust nodes,the scheme can reduce message redundancies,and enhance data forwarding processing.Finally,we compare SSNCR with two classical routing schemes and a secure routing in terms of message delivery ratio,average delay,overhead,and average trust value.When there are a large number of misbehaving nodes in OPPNETs,the simulations results show that SSNCR scheme can achieve better performance in terms of delivery ratio,delivery delay and delivery cost as compared to existing solutions.