Optimization Of Video Streaming In Wireless Sensor Networks

With the rapid development of CMOS technology, there have emerged a lot of vision sensors embedded withsingle-chip camera in the market. Wireless sensor networks consisting of visual sensor nodes can realize thetransmission of video streams. This new type of wireless sensor network can provide a more reliable andextensive technical support for video surveillance and other related fields. The site can be monitored in a morecomprehensive and intuitive way. Due to limited resources of traditional wireless sensor network and large dataand high business requirements of video transmission, it is a huge challenge to transmit video streams in thewireless sensor network. Therefore, the optimization of video streaming transmission in this new kind of wirelesssensor networks with limited resources is a highly realistic research.Based on the study of relevant literature about video streaming in wireless sensor networks at home andabroad, this paper achieves balance between video quality and network lifetime with the network optimizationmethod. The main research work is as follow:1) Video quality of transmission optimization based on improved dual-gradient algorithm. Taking linkcapacity constraint and link interference into consideration, this paper regards user’s perception of video as anetwork utility function. Since the problem is a non-smooth optimization problem, usually, dual sub-gradientalgorithm can be used to solve the problem. For the slow convergence or even no convergence caused byparameter sensitivity in dual sub-gradient algorithm, this paper proposes an improved dual sub-gradient algorithm.To avoid the step adjustment in the traditional dual sub-gradient algorithm, the proposed algorithm applies asmoothing technique to obtain a smooth approximation of Lagrangian function, then updates Lagrange multipliersby updating the two sequences. Matlab simulation results show that the improved method is convergent, and whencompared with the traditional algorithm, the improved method is faster in allocating video transmission rate.2) Wireless sensor networks is composed of a large number of sensor nodes with limited energy. When theenergy of a sensor node is exhausted, it may lead to paralysis of the entire wireless network, then shorten thenetwork lifetime. As for the video has more stream than the conventional scalar data, the visual sensor nodes needto more energy to finish computation. Therefore, the paper studies joint optimization problem of quality of videotransmission and network lifetime when link interference is strong and link capacity constraints are high andenergy is limited. In this model, a weight parameter to adjust relative importance of video quality of transmissionand network lifetime has been introduced, and a mapping parameter to ensure the two objective functions in thesame order of magnitude has been given. The distributed improved dual sub-gradient algorithm can be used tosolve the above problem. Matlab simulation results show that the algorithm can quickly converge to the globaloptimal solution of joint optimization problem.3) In wireless sensor networks, video transmission based on single path may cause network traffic loadimbalance, leading to depletion of vision sensor node energy and wireless link congestion before the computationis done, which would shorten the life of the network and result in transmission delay. For this drawback, the paperproposes a multi-path joint optimization models for video quality of transmission and network lifetime. The modelconsiders the interference among links, link capacity constraints and the node energy consumption constraints.The improved dual sub-gradient algorithm is used to solve the problem. Finally, Matlab simulation demonstrates that improved dual sub-gradient algorithm can quickly converge to the global optimal solution. Meanwhilesimulation results also show that when compared with single-path transmission, video quality of transmission andnetwork lifetime has been improved in the multi-path transmission.