Research On Fog-based Trustworthy Data Collection In Sensor-cloud

Due to the limitations of computing power,storage capability,and finite energy,traditional wireless sensor networks can't take full advantages of sensing ability in larger fields.Fortunately,the emergence of cloud computing makes up these weaknesses and expands its application space.A Sensor-Cloud is a combination of wireless sensor networks and the cloud computing.Therefore,the sensor-cloud inherits the cloud computing operation model,equips with real-time sensing ability and provides users remote and efficient on-demand services.However,some urgent problems have arisen with the integration of wireless sensor networks and cloud computing.On the one hand,traditional data collection methods are not applicable to the sensor-cloud.Collecting and uploading mass of data to cloud servers have become a huge challenge.Besides,a limited communication bandwidth of wireless sensor networks results in a delivery delay which is not friendly to the delay sensitive applications.On the other hand,data collection in sensor-cloud faces with some secure problems.Some features of sensor-cloud provide the convenient condition for malicious attacks,such as physical compromised attacks,influencing data confidentiality and integrity.Based on the above problems,this paper mainly focuses on the following points.(1)Research on minimizing data collection delay.With the combination of sensor networks and cloud computing,upload sensing data by static sinks is of long delivery delay and it is also a challenge for sensors' energy.In order to alleviate influence caused by physical transmission bottleneck,this paper proposed a data collection strategy based on the multiple mobile sinks.Utilizing mobile sinks balance energy consumption of sensors.Moreover,through the cooperation of multiple sinks,the bandwidth is broaden,which improves the data collection efficiency and reduces the collection delay.(2)Research on the trustworthy data collection strategy.The sensor networks are often deployed at the unattended environment with bad weather,so sensors suffer from malicious capture and natural damage.Besides,attackers can obtain the communication key by capturing some sensors.Then they place the fake sensors to destroy the confidentiality and authenticity of sensing data.For improving the security of data collection,we designed a trust-based collection strategy to insulate sensing data from malicious nodes.The sensor with bigger trust value is more likely to relay data.Furthermore,a monitoring mechanism is operated at cloud servers to prevent the fake sink attack.(3)Research on fog-based data collection security.When studying the security of sensor-cloud,we found that most current security researches neglected the fact that the physical limitation of sensor-cloud has an important impact on the trust evaluation.For example,the security monitoring frequency and trust computation complexity are restricted by the limited energy and computing power of sensors.Therefore,we proposed a fog-based universal trust evaluation method.Fog computing is between the sensor networks and cloud computing.Equipping with local computing and storage abilities,fog computing makes up the weaknesses of sensor-cloud and offers the strong technical support for protecting sensor-cloud data security.In this thesis,we primarily design a data collection method with multiple mobile sinks which can achieve the effect of self-adaptive delay.Then,to improve the security of data collection,we take sensor's reliability into consideration when design the data collection strategy,and propose a trust evaluation model to compute sensor's reputation.After that,to improve the availability and accuracy of trust evaluation,we propose a fog-based trust computing method.Finally,we conclude the study and discuss some future works.