Research On The Provenance Schemes For Large-scale Wireless Sensor Networks

As large-scale wireless sensor networks(WSNs)have been established to cater to different demands,the amount of data collected by them has also exploded.Since different types of tiny sensor nodes composed of WSNs are easily affected by the unprotected physical environment,it is quite difficult to evaluate the trustworthiness of the data collected from WSNs.In many application areas of WSNs,there are many precedents that the use of the data with low trustworthiness for decision-making has resulted in significant losses.Hence,it is quite crucial for decision makers to assess the trustworthiness of the data collected from WSNs before making critical decisions.In view of this,it is of great significance to study the data trustworthiness evaluation schemes for WSNs to ensure highly trustworthy data based on which the critical decisions are made.In the context of WSNs,provenance,which summarizes the history of the data ownership,is as a key element to assess the trustworthiness of the data collected by WSNs.In largescale WSNs,when the data is forwarded and aggregated for many times,the data’s provenance size expands rapidly,which easily leads to provenance overload problem,i.e.,the size of the provenance exceeds the limit of the packet capacity.Since WSNs are severely resourceconstrained,such as computation,bandwidth and energy,it is challenging to study how to efficiently and securely transmit provenance,and meet the requirements of data trustworthiness assessment in large-scale WSNs,which is the significance of this dissertation and the core issue to be solved.In order to effectively alleviate the provenance overload problem in large-scale WSNs,this dissertation proposes the following four provenance schemes:(1)In order to mitigate the problem of the existing arithmetic coding based provenance scheme that the compression rate degrades and the computation increases when the WSN’s scale expends,a cluster-based arithmetic coding provenance scheme is proposed.In this scheme,a large-scale WSN is hierarchically clustered and managed,and the local probabilities of the nodes in each cluster are used to encode the provenance segment in the cluster.This scheme uses the local probabilities within the cluster to avoid excessive splitting of arithmetic coding intervals,thereby improving the provenance compression rate.At the same time,since the provenance on each cluster level can be encoded as an independent segment,this scheme can encode and decode the provenance in an incremental manner,which raises the efficiencies of the provenance decoding as well as the data trustworthiness evaluation.The simulation and experimental results show that the proposed cluster-based arithmetic coding provenance scheme outperforms the existing arithmetic coding based provenance scheme with respect to the average provenance size and energy consumption.(2)In order to mitigate the main drawbacks of the existing block provenance schemes that they are sensitive to block loss,and are not efficient at data trustworthiness assessment,a multi-granularity graphs based stepwise refinement provenance scheme is proposed.This scheme enables to decode the provenance segment on a coarse granularity level first,by which to assess the trustworthiness of the data efficiently.Then,this scheme incrementally refine the provenance through decoding the provenance segments arrived in succession until the precise result is retrieved.Provenance segments on different granularity levels can be used to assess the trustworthiness of the data with different strengths,and provenance decoding robustness and the evaluation efficiency are then raised accordingly.The simulation and experimental results show that the proposed multi-granularity graphs based stepwise refinement provenance scheme not only outperforms the known related schemes with respect to average provenance size and energy consumption but also drastically improves data trustworthiness assessing efficiency.(3)In order to mitigate the main disadvantage of the dictionary based provenance(DP)scheme that it is sensitive to the WSN’s topology changes,a packet path index differences based provenance scheme is proposed.This scheme first selects backbone paths for the WSN and then builds path indices and dictionaries for some selected backbone paths.After that,the provenance is encoded by a selected index from the backbone path dictionaries together with a deviation from the backbone path.This scheme,as an improvement on the DP scheme,can not only overcome the DP scheme’s sensitivity to network topology changes,but also achieve a higher provenance compression rate,thus enabling the DP scheme to be applied to large-scale WSNs with more complex network topologies.To further increase the provenance compression rate and decrease the computation complexity at the nodes on the packet path,a packet path hash value based provenance scheme is proposed as a complement of the packet path index differences based provenance scheme.In this scheme,provenance is encoded as the data source node ID together with a segment of the packet path’s hash value,which keeps the provenance size as a constant.The simulation and experimental results show that the proposed packet path index differences based provenance scheme and the packet path hash value based provenance scheme outperform the DP scheme with respect to both average provenance size and energy consumption even when a WSN’s topology is not stable.(4)In order to mitigate the main drawbacks of the covert timing channel based provenance scheme that its channel capacity is low and the amount of data transmitted is limited,a twodimensional covert timing channel based provenance transmission scheme is proposed.In this scheme,a two-dimensional covert timing channel by using pseudo packet IDs permutation and packet sizes variation is first designed,and the two-dimensional covert timing channel is then applied to provenance transmission in WSNs with no extra bandwidth and energy consumption.The simulation and experimental results show that the proposed two-dimensional covert timing channel based provenance transmission scheme not only improves channel capacity,but also conserves more energy for provenance transmission in WSNs.Furthermore,the two-dimensional covert timing channel paves the way for further increasing covert timing channel capacity through multi-dimensional approaches.