| With the expansion of the application scope of multimedia wireless sensor network,it has become a reality to deploy multimedia wireless sensor network in pig house environment to monitor pig behavior and breeding environment.However,there are many problems in the deployment and information transmission of wireless sensor networks in piggery environment,especially in large pig farms,which need to deploy many nodes,resulting in large amount of data,complex networking,and difficult management.In this paper,a hierarchical topological structure suitable for large-scale farms is optimized for the network topological structure of piggery environment.On the basis of hierarchical topological structure,aiming at the transmission collision problem of MAC layer,the network collision is reduced by adjusting the minimum competitive window value of nodes at different levels.This paper designs a multimedia wireless sensor network node based on embedded technology.Compressive sensing theory is applied to the collection and compression of pigs’color images,which can greatly reduce the amount of network data,reduce transmission collision and prolong the lifetime of the node.The following tasks are accomplished.1)Designing and optimizing the WSN topology structure in pigsty environment,selecting an improved hierarchical topology generation algorithm,improving the parameters of the algorithm,selecting the optimal cluster head to establish the optimal cluster,and verifying the feasibility of the algorithm through experimental simulation.The experiment shows that the improved algorithm prolongs the survival time of the classical LEACH algorithm by more than 400 times,and nearly doubles the time length.Compared with LEACH algorithm,it has great advantages.From the comparison experiment between the improved algorithm and the LEACH algorithm,it can be seen that the LEACH algorithm consumes all energy after transferring less than 1*104 data packets.The improved algorithm can continue to transmit after transferring 5*104 data packets,but it can not reach 6*104 data packets.2)MAC layer conflict avoidance based on hierarchical topology structure in pigsty environment.Select the MAC layer optimization algorithm based on hierarchical topology structure,and implement the minimum competitive window value algorithm based on the minimum competitive window calculation method.The simulation results show that the proposed algorithm is lost compared with 802.11 protocol when the transmission rate is less than 5 b/s(sending 5 packets per second).Packet rate is reduced by 20%.In the actual experiment,the loss rate is reduced by only 10%because of the small number of nodes and layers.Although it is lower than the theoretical value,it greatly improves the network performance.3)Compression and reconstruction of pig color image based on compressed sensing technology,RGB color image compression and reconstruction based on CS.Through the contrast experiments of different sparse bases,different measurement matrices,different reconstruction algorithms and different compression rates,the R,G and B components of pig color image are compressed and transmitted separately.After receiving,three components are reconstructed separately,and then fused into color images.It is confirmed that the discrete Fu Liye sparse basis,the Fu Liye observation matrix and the OMP reconstruction algorithm can restore the pig color image without affecting visual observation when the compression ratio is 60%.4)Design and implement compression-aware WMSN node based on embedded system.Cortex－A72 core board is used to design the hardware of intelligent image acquisition node,including the transplantation of the whole system,the design of camera control circuit,the design and implementation of data acquisition and transmission protocol.The program of transplanting openCV dependency library,camera control,data acquisition and transnission,and compression sensing algorithm in openCV environment are completed.Finally,the experiment is carried out.Under the condition of 60%compression rate,the peak signai-to-noise ratio of reconstructed image after transmission can still be maintained at more than 30,which guarantees the normal observation demand for pig activity and pig house condition.In the experiment of node transmission,it is verified that the conflict avoidance method based on the minimal contention window adjustment of the MAC layer of pigsty topology reduces the network packet loss rate by 10%compared with 802.11 protocol cluster,and greatly improves the network performance. |
PDF Full Text Request