Node-Selfishness Management And Analysis Of Network Performance In Wireless Networks

Given the proliferation of smart devices in intelligent networks,each node is expected to be endowed with smart autonomic functions for network communications.By instinct,the individual network nodes would prefer to act selfishly rather than altruistically in intelligent network scenarios.For instance,while forwarding the packets of other nodes at the cost of sacrificing their own limited resources,they expect to satisfy some of their own objectives,such as maximizing their own transmission rate and/or minimizing their own resource con-sumption.A wireless network which consists of nodes exhibiting a selfish behavior is hence referred to as a selfish wireless network(SeWN).In such network scenarios,the selfish be-havior of network nodes,referred to as "node-selfishness",may reduce the throughput of the nodes and/or their integrity,thus potentially leading to the degraded network performances,i.e.,network connectivity,the reliability of both routing and delivery,and the effectivity of resource utilization.In this thesis,we mainly explore the effects of the node-selfishness on these network performances.The effect of the node-selfishness on the network connectivity of SeWNs constituted by selfish nodes(SeN)is investigated.The SeN's degree of node-selfishness(DeNS)is used for characterizing the effects of its energy resources and the benefits of the incentives provided for enhancing its transmission willingness.Furthermore,the SeNs' signal to in-terference plus noise ratios(s-SINR)are defined in terms of both their DeNSs and their interference factors(InF).We then continue quantifying the effect of node-selfishness on the grade of network connectivity and derive both the upper and lower bounds of the critical DeNS.Explicitly,the network is deemed to be connected when the DeNS is below the lower bound and unconnected for a DeNS above the upper bound.This allows us to quantify the asymptotic critical DeNSs for our SeWNs.Additionally,we develop an energy-conscious node-selfishness model for characterizing the relationship between the SeN's residual energy and its DeNS.Based on this model and on the asymptotic critical DeNS derived,the critical amount of residual energy required for maintaining a specific grade of network connectivity is determined,which is verified by our simulation results.The multi-service delivery between the source-destination pairs is also investigated in distributed SeWNs,where selfish relay nodes(RN)expose their selfish behaviors,i.e.,for-warding or dropping multi-services.Owing to the effect of the RNs' node-selfishness on the multi-service delivery,a distributed framework of the node-selfishness management is con-structed to manage the RN's node-selfishness information(NSI)in terms of its available re-sources,the employed incentive mechanism and the quality-of-service(QoS)requirements,and the other RNs' NSI in terms of their historical behaviors.In this framework,the RN's NSI includes the DeNS,the degree of intrinsic selfishness(DeIS)and the degree of extrin-sic selfishness(DeES).Under the distributed node-selfishness management,a path selection criterion is designed to select the most reliable and shortest path in terms of RNs' DeISs af-fected by their available resources,and the optimal incentives are determined by the source to stimulate forwarding multi-services of the RNs in the selected path.Our simulation re-sults demonstrate that this framework effectively manages the RNs' NSI,and the optimal strategies of both the path selection and the incentives are determined.We investigate the cross-layer resource optimization for SeWNs with dynamical NSI.The DeIS and DeES of the selfish RN are defined as the effects of its own residual energy re-source and the incentive in an incentive mechanism,respectively.A model of two-timescale node-selfishness dynamics is constructed to describe the time-varying characteristics of the RN's DeIS and DeES.Based on the two-timescale node-selfishness dynamics of the RNs,a two-timescale resource-optimization dynamical algorithm(TRODA)is employed to control the transmit powers of both RNs and end users(EU),as well as the flow rates of the data packets injected by all EUs.Meanwhile,the stability and the tracking error of the TRODA under the two-timescale node-selfishness dynamics are analyzed by Lyapunov theory.Our simulation results demonstrate that the two-timescale optimization algorithm is stable and has trivial tracking error.We investigate the dynamic packet delivery through a specific path in SeWNs with cascaded selfish RNs under the unknown node-selfishness dynamics.A dynamic node-selfishness model is designed to formulate the variation of the RN's DeNS with both its own resource and the incentive controlled by the source.By using the neural network(NN),we identify the unknown functions of the node-selfishness model,maximize the source's infinite-horizon utility,and approximate the optimal incentives,for finally obtaining the tradeoff between the path reliability and the incentive cost.Simulation results demonstrate the effectiveness of the proposed scheme.