Research On Key Technologies Of Data Aggregation In Wireless Sensor Networks

The wireless sensor networks(WSNs)is becoming a premier research area,especially with recent advances in embedded systems and wireless communications.A WSNs is composed of one or multiple base-stations(or sink)and a set of sensor nodes deployed in an area called sensing field.In WSNs,the sensor nodes sense events and generate event data,and then send these data to the base-station(or sink)using hop-by-hop communication.WSNs have numerous applications in areas as security surveillance,real-time target tracking,environmental monitoring and so on.However,resource limitation is the main concern of any WSNs application.One important mechanism to solve this bottleneck is the in-network data aggregation,which aims to eliminate redundant packet transmissions by filtering repeated and unnecessary data readings and thus cut down the energy used in communication.Generally speaking,an aggregation protocol should achieve two main objectives: Quality of Service(QoS)and Privacy-preserving,and this dissertation is based on the study of them.First of all,optimal scheduling strategies satisfying different performances were designed via analysis of aggregation scheduling algorithm;secondly,privacy protection is realized in the process of data aggregation by introducing Homomorphic encryption and Source coding technologies.The main research work and contributions of this dissertation are as follows:1.A low latency data aggregation scheduling algorithm is proposed aiming to optimize single QoS performance.This algorithm actualizes TDMA data aggregation scheduling through construction of approximate maximum collision-free set.Meanwhile,it can also be used in sleep scheduling by determining the awake or asleep state of each node according to scheduling sequence.The results of analysis and testing show that this algorithm can greatly cut down delay of data aggregation and save energy compared with the present algorithms.2.The main task in designing a data aggregation schedule is to get a good trade-off between Qos and weighted fairness guaranteed.In order to address this problem,a novel data aggregation scheduling algorithm,MISS,is proposed in this dissertation,which can produce a collision-free TDMA schedule for data aggregation based on twice Maximum Independent Set(MIS).In the first phase,we constructed a tree structure based on MIS,and then adjusted it according to the energy consumption forecast of each node,and the balanced data aggregation tree was formed.In the second phase,we utilized the approximate Maximum Weighted Independent Set(WMIS)of all selected links to allocate time slots for each communication link.The results of simulation demonstrate that MISS not only improves the aggregation delay and lifetime of the network,but also gives weighted fairness guarantee.3.Based on source coding technology,recoverable data aggregation privacy preserving scheme is proposed.Firstly,this scheme realized hiding and compression of data by operating source coding to privacy data according to the correlation of sensing data.Secondly,this scheme converted and concatenated the coded data so that the base station can recover all sensing data even these data has been aggregated.Meanwhile,the base station verified the integrity of data in accordance with aggregate signature.Experiment results demonstrate that the proposed source coding based recoverable data aggregation scheme can protect data privacy as well as decrease redundant transmitted data.4.A data aggregation privacy-preserving scheme guaranteeing confidentiality and data integrity of data is presented.HASH function is used to update the key of each time slot by using present aggregation round and private key as an input.The application of homomorphic encryption makes intermediate node perform aggregation operation on ciphertext directly.HMAC enables base station to verify whether the aggregation data has been modified during transportation.Moreover,plaintext is coded before being encrypted so as to satisfy multiple applications.Through theoretical analysis and simulation,we verify that the proposed algorithm could preserve data privacy with lower communication assumption and higher data aggregation accuracy.