User-Oriented Hybrid Opportunistic Network Design  

作　　者：ZHANG Yong LI Jin SONG Mei YANG Ming 

机构地区：[1]Beijing Key Laboratory of Work Safety Intelligent Monitoring,Beijing University of Posts and Telecommunications [2]Beijing Information Science & Technology University

出　　处：《China Communications》2015年第11期119-133,共15页中国通信（英文版）

基　　金：supported by the 863 project (No.2014AA01A701);the National Natural Science Foundation of China under Grant No.61171097 and No.61372117

摘　　要：To improve the end-to-end Successful Delivery Probability(SDP) of mobile nodes in opportunistic network,hybrid Opportunistic Network(OppNet) architecture using Fix Infrastructure Nodes(FINs) is proposed.As enhanced OppNet serving for special users,the network planning and design should take user's mobility behavior into account.We adopt collected users' mobility records,analyze users' stay pattern,and establish Message Exchange Points(MEPs) where FINs should be installed.Our solution includes MEPs detection algorithm,Maximum end-to-end Success Delivery Probability(MSDP) detection algorithm,and simulated annealing algorithm.After data preprocessing,MEPs detection algorithm detects feasible locations from users' trajectory records where not more than one user visit frequently.End to end SDPs of these feasible MEPs are evaluated using MSDP detection algorithm.To reduce the computation complexity,a simulated annealing algorithm to select optimal FIN locations in a large number of feasible MEPs is given.The evaluation experiments based on reallife trajectory dataset verify our proposal can achieve better E2 E SDP gain than "most visitors" strategy and "max residence probability" strategy.To improve the end-to-end Successful Delivery Probability(SDP) of mobile nodes in opportunistic network,hybrid Opportunistic Network(OppNet) architecture using Fix Infrastructure Nodes(FINs) is proposed.As enhanced OppNet serving for special users,the network planning and design should take user's mobility behavior into account.We adopt collected users' mobility records,analyze users' stay pattern,and establish Message Exchange Points(MEPs) where FINs should be installed.Our solution includes MEPs detection algorithm,Maximum end-to-end Success Delivery Probability(MSDP) detection algorithm,and simulated annealing algorithm.After data preprocessing,MEPs detection algorithm detects feasible locations from users' trajectory records where not more than one user visit frequently.End to end SDPs of these feasible MEPs are evaluated using MSDP detection algorithm.To reduce the computation complexity,a simulated annealing algorithm to select optimal FIN locations in a large number of feasible MEPs is given.The evaluation experiments based on reallife trajectory dataset verify our proposal can achieve better E2 E SDP gain than 'most visitors' strategy and 'max residence probability' strategy.

关 键 词：OPPORTUNISTIC NETWORK NETWORK ARCHITECTURE USER MOBILITY 

分 类 号：TN915.02[电子电信—通信与信息系统]