一种基于反馈偏好的P2P信任模型

在P2P网络中,现有的信任模型中的反馈评价只是对服务质量简单评价,不能准确反映该节点的其他方面.不同的节点关注的反馈评价不同,信任模型不仅需要对服务可信度进行评价,还要对用户关心的其他问题进行评估.由于实验基于文件共享系统,我们将从文件可信度、带宽、服务过程、性能等方面计算出节点的信任值,在保证节点可信的情况下,根据用户的侧重点不同,其值有较大的不同.仿真实验表明,该模型不仅能够有效的抵御恶意攻击,而且能够满足用户的偏好需求,并且提供的评价更加客观、真实.     

基于隶属云理论的主观信任评估模型研究

根据主观信任的主观性、模糊性和随机性等特点,引入隶属云模型理论对信任进行建模,给出了信任云和信任等级云的定量描述,提出了信任评价云的计算、合成、综合运算和主体信任等级的评估.模拟实验结果表明该信任评估模型是有效的.     

P2P环境中的一种信任模型

提出一种新的信任模型,将用户提供服务和进行推荐的能力进行区分,用推荐信任表示用户的推荐能力,根据用户的历史交易经验和系统中其他用户的推荐来计算节点的综合信任度,用户依据计算结果选择交易对象,交易结束后给出信任评价,进行相应的反馈.仿真试验表明,模型能有效评估节点的信任度,提高系统的可用性.     

面向P2P网络应用的基于声誉的trust管理模型

提出一种新的面向P2P网络应用的信任管理模型,该模型根据节点的历史交易情况和系统中其他节点的推荐计算节点的信任度,给出了普遍适用的、具体量化和更新推荐者可信度的数学表达形式,同时考虑到节点的自主性动态行为影响信任度计算的不确定性,引入风险因素,提出了采用信息熵理论来量化风险,将实体之间的信任程度和信任的不确定性统一起来.仿真实验及分析表明,该信任模型能够有效识别恶意节点,相比已有的一些信任模型在系统成功交易率、模型的安全性等问题上有较大改进,可以使节点之间更有效的建立信任关系.     

基于推荐的信任链管理模型

针对目前信任链管理中信任计算时信任信息不完备的问题,提出了一种基于推荐的信任链管理模型RTCM(recommendation based trust chain management model)。通过构建信任网络,在信任链选择时考虑信任传递衰减、时间衰减、路径权重等因素,使得信任链搜索的时间复杂度大幅降低。充分包含了信任的相关信息,并利用加权紧密度对信任链上的推荐信任进行合并,同时给出了抑制节点恶意推荐的策略。仿真实验表明,该模型提高了网络的可信性。     

基于改进D-S证据理论的信任评估模型

针对已有的信任证据模型不能快速有效地处理分布式网络中存在的恶意攻击,且缺乏关于三元信任关系组的信任归一方法,提出了一种基于改进D-S证据理论的信任模型,在此基础上,提出了基于持续序列的基本可信度函数和基于评估函数的信任评估方法,使得模型能更快地抑制恶意节点,并且评估结果更贴近现实值。通过分析与仿真,验证了本模型具有抑制聚集信任攻击的有效性和健壮性,同时信任评估方法更具合理性和准确性。     

基于评价相似度及拍卖行为的信任模型

电子商务中如何选择可信卖家进行交易已经成为众多学者研究的对象,在此采用拍卖模式描述电子商务中的交易行为,并基于相似度、可信推荐者列表为买家选择可信的交易卖家。首先,计算买家与推荐者评价的相似度并选择较可信的一些推荐者加入到可信推荐者列表;然后,基于买家的可信推荐者列表,结合时间窗口等方面的因素,计算出卖家的私有信任度和公有信任度;最后,通过期望价值函数为买家选择出可信交易卖家。实验表明,该模型能够提高信任度计算的准确性,有效帮助买家选择可信卖家,提高交易成功率。     

认知无线电网络中基于信任的安全路由模型

针对认知无线电网络中路由受到的选择性转发攻击,提出一种基于信任的安全路由模型。通过监视节点的数据转发行为,构建节点的信任以识别恶意节点。基于路由选择必须同频谱分配紧密结合的特点,在路由发现阶段,请求信息中封装节点的可用频谱机会并避免发送给恶意节点。在路由选择阶段,利用节点信任构建可用路径信任,结合延时度量进行路由决定。同时,根据节点的信任划分对其服务请求采取不同的响应,对非可信节点的恶意行为采取更严厉的惩罚,从而激励节点在路由中的合作。仿真结果及分析表明,在选择性转发攻击下,该模型较已有模型在网络吞吐率、端到端延迟方面都有较好的表现。     

Trust evaluation method for clustered wireless sensor networks based on cloud model

Trust Management has been proved to be an effective method to detect malicious nodes and ensure security in Wireless Sensor Networks (WSNs). While, most existing trust management methods are not good at dealing with uncertainty of trust relationship such as randomness, fuzziness in WSNs, which leads to inaccurate trust metric. In this paper, a trust evaluation method for clustered wireless sensor networks based on cloud model is proposed and evaluated, which implements the conversion between qualitative and quantitative of sensor nodes’ trust metrics in order to achieve better trust evaluation. Firstly, the method considers multi-factors including communication factor, message factor and energy factor and builds mathematical model for each trust factor to get factor trust cloud. Secondly, immediate trust cloud is calculated by assigning adjustive weights for each factor trust cloud and combining them. Thirdly, recommendation trust cloud and immediate trust cloud are synthesized according to time sensitive factor in order to get final trust cloud. Furthermore, the final trust cloud of sensor node is converted to trust grade by trust cloud decision-making. Verification Experiments manifest that the proposed method has feasibility and accuracy in the aspect of evaluating sensor nodes’ trust. Moreover, comparison experiments under different attacks show that our method is sensitive to multiple attacks, it outperforms other trust evaluation methods not only in the accuracy of detecting malicious nodes, but also in the tolerance of abnormal conditions.     

Trust based negotiation bargaining game model for bandwidth management algorithms

Nowadays, Multi-hop Relaying Network (MRN) has gained wide acceptance as a next step towards future radio networks. MRN can extend the service area as well as improve the performance of wireless networks. To exploit the multi-hop relaying operation, an important issue is how to properly control wireless bandwidth. In this paper, a new bandwidth management scheme is proposed for MRNs. By integrating the random arrival rule and Nash bargaining model, the proposed scheme adaptively controls the wireless bandwidth to maximize network efficiency. In our scheme, trust value and bargaining powers are decided according to the Bayesian inference and real-time negotiation process, respectively. This approach can make the network system be close to the optimized network performance. To prove the effectiveness of the proposed scheme, a simulation has carried out. The results demonstrate the effectiveness of the proposed scheme in comparison with other existing schemes.     

基于模糊Petri网的信任评估

针对网络实体间信任关系的不确定性,论文提出了一种基于模糊Petri网的信任评估模型,该模型给出了信任评估的模糊规则,在此基础上利用模糊Petri网与矩阵运算相结合的推理方法,通过并行处理关于一个实体的多个推荐来评估实体间的信任关系。经信任关系评估实例计算可以看出,基于模糊Petri网的推理方法使得信任评估灵活、简单、有效。     

Nanobiosensors based on individual olfactory receptors

The animal olfactory system represents the gold standard of biosensors, due to its ability to identify and discriminate thousands of odorant compounds with very low thresholds. Using olfactory receptors (ORs) as sensing elements instead of chemical sensors, biosensors would benefit the naturally optimized molecular recognition of odorants to develop a new generation of bioelectronic noses. The purpose of SPOT-NOSED European project was the development of nanobiosensors based on single ORs anchored between nanoelectrodes, to mimic the performances of natural olfactory system. Nanobiosensors arrays could then increase odorant sensitivity or widen the odorant detection spectrum. ORs were expressed in yeasts plasmic membrane, and their functionality tested in whole yeasts. Then, nanosomes bearing the ORs were prepared from S. cerevisiae, and Surface Plasmon Resonance was performed on nanosomes for quantitative evaluation of OR response to odorant stimulation. ORs retain full activity and discrimination power in immobilized nanosomes, thus allowing their use in the fabrication of the nanobiosensors. Nanoelectrodes were fabricated using conventional photolithography and focused ion beam milling, with sizes in adequation with the nanosomes. ORs borne by nanosomes were specifically immobilized onto conducting substrates via mixed Self Assembled Monolayers, neutravidin and specific antibody to the ORs. The process was optimized by microcontact printing, and the anchored nanovesicles visualized by Atomic Force Microscopy. A transimpedance preamplifier suited for low-noise wide-bandwidth measurements was designed to be directly connected to the nanoelectrodes. Electrochemical Impedancemetric Spectroscopy detected significant changes upon electrodes functionalization, grafting of ORs carried by nanosomes, and ORs conformational change induced by odorant binding.