Research On Key Technologies Of Data Dissemination For Selfish Opportunistic Networks

Opportunistic networks(Opp Nets)turn the obstacle,i.e.,mobility,into an opportunity and implement communications between nodes without connectivity path by a “store-carryforward” paradigm.Opp Nets are more suitable for the demand of Internet of Things and ubiquitous computing and have become an important research direction of mobile ad hoc networks.In the realistic applications of the content-centered Opp Nets,data transmission is implemented by data dissemination because the data has no explicit destination address.Data dissemination,which decouples the data producers and consumers in time,places,and data streams,is more suitable for the intermittent connectivity of Opp Nets and becomes the current hotspot of Opp Nets.Most of the existing data dissemination protocols in Opp Nets obviously requires the cooperation among nodes.However,the node may be reluctant to cache data for other nodes due to safety or resources.The selfishness has seriously affected the performance of data dissemination in Opp Nets.Therefore,it has more significant for practical applications to design incentive-aware data dissemination protocols in Opp Nets.The current data dissemination protocols are mainly driven by senders.However,it is difficult to fully acquire the status of receivers for senders,such as resources and willingness of forwarding,which maybe degrade the performance of data dissemination since the data was dropped by receivers.The main incentive mechanisms take the forwarding behavior of the next hop as a service.However,in intermittently connected Opp Nets,it is difficult to monitor the forwarding behaviors of next hop.In this paper,therefore,we employ the receiver-driven data dissemination,regard the data as the service in the incentive mechanism and use the barter trade incentive mechanism.Based on the above considerations,we propose an incentive-aware and efficient data dissemination protocols.In order to improve the performance of data dissemination,a utility-based cache policy is proposed to optimize the cache efficiency of receivers and a data transmission model based on opportunistic network coding is proposed to reduce the cost of data forwarding.Because the receiver-driven cache policy and the barter trade incentive mechanism are mutual promotion,it is necessary to combine the two methods to put forward the utility-based barter trade incentive mechanism.Based on the above analysis,the specific research contents and contributions are as follows:(1)In view of the fact that simple barter trade(SBT)incentive mechanism reduces network performance due to the rigid exchange principle,we propose a debt-based barter trade(DBT)incentive mechanism.In DBT,we employ the debt relationship to relax the equivalent exchange principle so as to increase the amount of exchange data and improve network performance.In order to prevent the debt relationship from being out of control,a trust-based punishment mechanism is proposed.Simulation results show that DBT,compared with SBT,can increase the delivery ratio and reduce the delay at the cost of the increase of network cost.(2)In view of the fact that DBT incentive mechanism degrades network performance due to the blindness of caching data,we propose a utility-based barter trade incentive mechanism based on DBT(UDBT)for data dissemination in Opp Nets.In UDBT,we propose a method of propagation node interest in aggregation manner and a calculation method of data delivery probability.We also design utility value and cache policy based on node interest,data deliver probability and debt relationship for the purpose of caching data.Simulation results show that UDBT,compared with DBT,can obtain higher delivery ratio and lower delay at the lower network cost.(3)Opportunistic network coding(ONC)can effectively improve the throughput of wireless networks.However,ONC is rarely used in Opp Nets because it is difficult to exist encoding opportunities in Opp Nets.In this paper,we design an ONC-based data transmission models(ONCTM)for Opp Nets.In ONCTM,we establish encoding opportunities by adding new communication opportunities to enable more nodes to participate in one communication.We also propose a simple Markov-based delay performance analysis model and a model for estimating neighbor number based on birth death process.The two models were verified by simulation experiments.Simulation results show that ONC-based routing,compared with Epidemic routing,can obtain higher delivery ratio and lower delay at the lower network cost.(4)The most existing data dissemination protocols suppose that nodes are cooperation.In realistic Opp Nets,however,nodes are often selfishness due to safety or resources,which seriously affected the performance of data dissemination.In this paper,we design an interestand-incentive-aware and ONC-based data dissemination protocol(I2ONC)in Opp Nets.In I2 ONC,we employ DBT incentive scheme to encourage node cooperation,use ONC to redecethe forwarding cost of senders,and propose an interest-aware buffer management strategy(IBMS)based on the utility in UDBT to optimize the cache efficiency of receivers.However,DBT and IBMS are not fully suitable for ONCTM,so we design the extended version DBT and IBMS.Simulation results show that I2 ONC can not only encourage node cooperation but also obtain higher delivery ratio and lower delay at the lower network cost in selfishness scenarios.