基于移动无线传感器网络的植株图像监测系统设计与测试

针对静态无线传感器网络(static wireless sensor network,SWSN)在图像监测中功耗分布不均、传输不可靠等问题,设计了基于移动无线传感器网络(mobile wireless sensor network,MWSN)的农田植株图像监测系统。选用JN5139模块搭配摄像头采集和编码图像,利用无人机(unmanned aerial vehicle,UAV)搭载协调器收集信息。通过仿真和测试可得,节点在MWSN中的最小功耗为10μW,工作功耗为133 m W,在SWSN中为133 m W,路由节点功耗为普通节点的2倍;在10~35 m范围内,MWSN的信号强度为-68~-86 dbm,SWSN为-83~-85 dbm;在10~80 m范围内,MWSN的误码率范围是0~9.2%,SWSN是0~38.6%。试验时UAV在15 m高空悬停接收地面设备发出的图像数据,测得获取一张图片平均需135 s,图片分包平均为22次,解码后的图像可以较好的反映植株生长状态。上述结果表明,工作时间相同,MWSN中节点的功耗差异性小,呈均匀分布;在一定距离范围内,MWSN的传输能力和抗干扰能力整体上优于SWSN,能保证数据传输的可靠性,因此,基于移动无线传感器网络技术的图像监控系统能够满足大范围农田中植株图像的监测需求。

Design and test of plant image monitoring system based on mobilewireless sensor network

Wireless sensor network (WSN) has become a common basic component in agricultural information collection and monitoring technology. However, with the rapidly increasing agriculture monitoring requirement, the limitations of traditional static wireless sensor network (SWSN) have emerged, including the heterogeneous network power consumption, the weak performance of network and the lack of node’s computing. In order to solve the SWSN’s unbalanced energy consumption problem and improve the node’s transmission capacity, an agricultural crop image monitoring system based on mobile wireless sensor network (MWSN) technology is proposed in this paper, which contains end-device nodes (ENs), coordinator node (CN) and unmanned aerial vehicle (UAV). The system architecture is divided into 3 layers: the information collection layer (ENs and cameras), the information relay layer (CN and UAV) and the information concentration layer. In the information collection layer, the ENs combined with cameras are used to detect, and convert JPEG picture into coding message. The gateway consists of CN and 3G module in the information relay layer. It takes the message from EN while UAV flies over EN. All messages are stored into the information concentration layer. This layer is also called management center, which can decode messages and show plant images. It is important to note that the device used in MWSN does not include route nodes (RNs). Instead of using multi-hop scheme by RNs to transmit image message, a single-hop mode is used in this paper, and all data can only be issued from ENs and entry into CN. SWSN’s energy consumption is reduced by this way. On the other hand, the MWSN can ensure the data transmission performance and anti-interference ability and improve the network reliability effectively. We test the system from 2 aspects, the performance test of network via energy consumption modeling with MATLAB simulation and the network transfer capability test. Firstly, we test JN5139 power consumption under different conditions, and then create a model using “first order radio”. The simulation shows that the energy consumption of the RNs is twice as much as EN (133 mW); in contrast, node consumption in MWSN is 133 mW (work mode) and 10μW (sleep mode). Secondly, we test the performance of 2 kinds of networks. We put one EN on the ground, and test the received signal strength indication (RSSI) and package error rate (PER) of CN in different spatial locations. The results show that for the RSSI within the 35 m, the MWSN’s range is from -68 to -86 dbm and the SWSN’s range is from -83 to -85 dbm, for the RSSI over 35 m, the MWSN’s range is from -86 to -88 dbm and the SWSN’s range is from -85 to -92 dbm, and for the PER between 10 and 80 m, the MWSN is from 0 to 9.2% and the SWSN is from 0 to 38.6%. The results show universal phenomenon of unbalanced energy consumption of SWSN. But the MWSN not only balances energy consumption, but also has low energy consumption level. These results indicate that the performance of MWSN is better than SWSN within a certain range. A set of experiments are carried out in the small potato plantations in the campus of Qiqihar University in July 2015. Three lower power JN5139 modules (ENs) are arranged at a distance of 100 m. It should be noted that, ENs can’t transmit data interactively, and only send data to CN. Thus, the four-rotor UAV flies into the area at a slow speed, and hovers 15 m high from the ground to receive the message. At any moment, only one EN can communicate with the CN. The experiment shows that, the average time which is cost in acquisition and transmission of JPEG picture is about 135 s, and the number of package sent is 22 on average. The decoded images can reflect the growth status of plants well. In conclusion, the results of experiment show that the method is effective and applicable.