Research On Data Transmission And Processing Mechanisms In Delay Tolerant Mobile Sensor Networks

To meet the development requirements of emerging technologies in today’s world, the equipment is equipped with intelligent sensors that no longer keep static, but change with the mobile objects. Composed of the moving sensors above, the topological structure of the whole network keeps changing dynamically. It is very difficult to establish a fixed link between nodes. End-to-end data transmission schemes that applied in Traditional Wireless Sensor network(WSN) become no longer available. Herein, it is an urgent demand for studying and analyzing this new type of network, seeking distinctly new methods of data transmission. Therefore, Delay Tolerant Mobile Sensor Networks(DTMSN) arises at the historic moment.DTMSN is one special kind of Delay Tolerant Networks(DTN) that developed in WSN, by the means of ubiquitous data gathering, aiming to acquire a large sum of data information. DTMSN not only inherited some characteristics of DTN and WSN, but also possesses its unique network characteristics.Based on the theory of linear dimension reduction, this paper improves the data analysis ability of traditional k-means algorithm, and completes for rapid analysis and processing tasks of a large amount of data collected in DTMSN. According to node activity of different nodes, this paper improves the probability estimates of probabilistic routing algorithm. RVNS meta-heuristic algorithm is introduced to select the optimal value for multiple hops relay transmission, achieving the balance of different nodes’ energy expenditure in the network.Main innovations of our research are concluded as follows:(1) To eliminate the phenomenon of data accumulation in DTMSN, an improved k-means clustering algorithm is proposed based on linear discriminant analysis(LDA), namely LKM algorithm. LKM firstly introduces LDA to complete the process of linear dimension reduction, from the high-dimension dataset into two dimensional dataset. Afterwards, LKM takes advantage of k-means algorithm for clustering analysis of the dimension-reduced data. As a result, LKM algorithm cuts down the time of feature extraction, and improves the clustering analysis accuracy of traditional k-means algorithm, thus enhancing the performance of k-means clustering algorithm to analyze and process vast data.(2) In the probabilistic routing algorithms of DTMSN, the prediction of probability is not reasonable and the communication link between nodes is intermittent. A Node Activity-based Probabilistic Routing algorithm(NAPR) is proposed by mixing a node activity factor into the prediction of delivery probability. NAPR algorithm takes the encountering records and the nodes’ active level into consideration. Besides, NAPR adopts the TimeTo Live(TTL)-based discarding strategy to manage nodes’ buffer space.(3) Because of the limited sensors’ resources in DTMSN, unbalanced energy consumption will exhaust the batteries of active sensors soon, which shortens the network survival cycle significantly. In this paper, an energy-balanced multi-hop relay transmission scheme based on Reduced Variable Neighborhood Search(RVNS) in DTMSN is proposed. RVNS is introduced to select the optimal sensor(a sensor that contains the maximum remaining energy) in transmission path as the next hop relay, aiming at the balance of power consumption and prolonging the survival time of the network.