社交网络中基于节点影响力和累积效应的数据转发算法

Data Forwarding Algorithm Based on Node Influence and Accumulative Effect in Social Networks

目的社交网络是DTNs网络的一种,该网络是一种特殊的无线自组织网络,缺乏端到端的持续连接,具有较高的丢包率和传输延迟。研究目的是克服上述缺点,实现数据的高效转发。方法针对社交网络中节点与社区以及节点与目的节点的关系进行研究分析,提出一种基于节点影响力和累积效应的数据转发算法:DFNS算法。结合C++软件进行实验模拟,与DTNs网络中的Epidemic算法和Label算法进行对比分析。结果随着网络中发包数目的增加,DFNS算法的传递率略低于Epidemic算法,高于Label算法,其平均延迟却高于其他2种算法。拷贝数目方面,DFNS算法比Epidemic算法最高减少44.38%,相比于Label算法最高减少39.47%。随着网络中数据包的存在时间的改变,3种算法的传递率先迅速增长,之后基本不发生改变,且3种算法平均延迟的变化趋势基本相同。在拷贝数目方面,DFNS算法比Epidemic算法平均少了45.16%,比Label算法平均少了40.25%。结论 DFNS算法在传递率方面优于Label算法,略低于Epidemic算法,相比其它2种算法能明显降低网络中数据包的拷贝数目,减少资源的能量消耗。在提高网络性能,降低网络成本方面优于其它两种算法。

Objective Social network is a kind of DTNs network,which is a special wireless ad hoc network.It lacks of end-to-end continous connection,and has high packet loss rate and transmission delay.However,the social network is one of DTNs.To overcome these shortcomings,efficient forwarding of data was achieved.Methods The relationship between nodes and communities,and nodes and destination nodes in social networks were studied and analyzed,and a data forwarding algorithm based on node influence and cumulative effect was proposed.The experiment was simulated with C++software,and the algorithm was compared with Epidemic algorithm and Label algorithm in DTNs network.Results With the increase of the number of contracts in the network,the transmission rate of DFNs algorithm in this paper was slightly lower than that of the Epidemic algorithm,and higher than the Label algorithm,But the average delay was higher than those of the other two algorithms.In terms of copy number,the DFNS algorithm reduced the maximun by44.38%compared with the Epidemic algorithm,and by39.47%compared with the Label algorithm.With the changes of packet existence in the network,the transmission of the three algorithms took the lead in rapid growth,and then basically unchanged.And the average delay trend of the three algorithms was basically the same.In terms of copy number.the DFNS algorithm in this paper was45.16%less than the Epidemic algorithm,and compared with the Label algorithm,the average was less than40.25%.Conclusion The DFNS algorthm in the transfer rate is better than that of Label algorithm,and slightly lower than that of the Epidemic algorithm;and compared with other two algorithms it can significantly reduce the number of copies of the packets in the network and the energy consumption of resources、Therefore,this algorithm plays a certain role in improving network performance and reducing network cost.