Research On Multi-node Cooperative Image Acquisition And Compression Technology In Wireless Multimedia Sensor Networks

As a special application and extension of the traditional Wireless Sensor Networks, Wireless Multimedia Sensor Networks (WMSNs) expand human's horizon to any phsical space. It has shown its potential in many areas, such as future information warfare, space warfare, unmanned combat and civil applications. Besides the common features of the traditional Wireless Sensor Networks, WMSNs has many unique characteristics, such as high computational complexity and equal energy consumption of image acquisition, computation and transmission. Because of these characteristics, solving the problem of high energy and high cost and prolonging the lifetime of WMSNs becomes the key factor in the practical process of WMSNs. Supported by the National Science Foundation of P.R.China"Image compression algorithm and its multi-node cooperative transmission strategy for wireless multimedia sensor networks"(Grant No. 60872151), this dissertation concentrates its attention on the research of multi-node cooperative image acquisition and compression technology in WMSNs with the purpose of minimizing the energy consumption and maximizing the lifetime of WMSNs.The main contents and contributions of this dissertation are as follows:1. For the purpose of maximizing the lifetime of the network while still ensuring the monitoring quality, a two-layer cooperative WMSNs architecture is proposed. Based on the three-level cooperation model between the camera-equipped node layer and the normal sensor node layer, the main design idea of the proposed architecture is to share the camera-equipped node's energy consumption burden of image acquisition and compression with the normal sensor nodes which are deployed densely. The proposed network architecture is low-cost, and suitable for large-scale deployment.2. In order to save energy of the camera-equipped node layer by reducing the acquisition amount of the useless background images, the event-driven camera-equipped node wakeup problem in WMSNs is brought forward and studied. According to the directional sensing model of the camera-equipped node, a region-division-based distributed camera coordinative wakeup mechanism is proposed. In this mechanism, all the normal sensor nodes need to estimate the position relationship with the camera-equipped nodes nearby and record the results before the running of the network in order to omit the process of event-covering judgement when the network is running. Then the sensing field is divided into regions based on the types of the normal sensor nodes. The events occurred in the sensing field are managed by the cluster heads in the regions when the network is running in order to reduce the computation energy consumption of the camera-equipped nodes and the communication energy consumption among the nodes. What is more, the complexity of the decision-making process and the communication energy consumption are reduced by running a camera neighbor list in the memory of the camera-equipped node.3. For further reducing the energy consumption of the camera-equipped node layer by decreasing the acquisition amount of the redundant event images, the problem of event image acquisition with minimum energy consumption is brought forward and studied. Considering the characteristic of event image acquisition in WMSNs, a local camera coordinative energy-saving strategy is proposed. After the implementation of this strategy, a redundant active camera-equipped node distributed self-examination method is used to let the redundant active camera-equipped nodes with lower residual energy sleep. What is more, a local centralized coordination strategy is running periodically to further reduce the number of the active camera-equipped nodes in the process of the image acquisition and balance the energy consumption of the camera-equipped nodes.4. When the camera-equipped node compresses the images or the cluster head compresses the images in the traditional WSNs cluster structure, its energy will be exhausted quickly. To avoid local paralysis of the network, a two-hop clustered image transmission scheme is proposed. In this scheme, many redirectors are used to compress and forward the images for the purpose of reducing the energy consumption of the camera-equipped node and the cluster head. With adaptive adjusting of the transmission radius in the transition cluster and tasks allocating based on the residual energy of the normal sensor nodes by the camera-equipped node, the energy consumption of the nodes in the network is balanced which greatly prolongs the lifetime of the network.5. In order to solve the problem of high computation complexity and high energy consumption of JPEG2000 and poor image quality in the case of multi-node distributed implementation in WMSNs when the bit rate is low, a Lapped Biorthogonal Transform (LBT) based multi-node cooperative image compression scheme is proposed. Because of the time-domain pre-processing of DCT, the block effect is reduced efficiently. The proposed scheme greatly lowers the complexity of computation and reduces the required memory, while it still achieves better image quality than the distributed JPEG2000 method.The above-mentioned research work of this dissertation provides fundamental theory and technique support to the establishment of the battlefield visualization system based on WMSNs, which is of great practical and strategic significance to the building of the integrated battlefield situation monitoring platform of land, sea, air, and space.