| Objective: Bioterrorism and biological warfare agents attack, which have features of large hazards and wide effects, not only have a huge impact on national security, social stability and people’s living, but also threaten the army security and military situation seriously in the war. Response, prevention, monitoring, early warning of bioterrorism and biological agents attack are of great value that subject to extensive research around the world since the 21 st century. Biological warfare agents are typically used in droplets or powder state, spread in the air in the form of biological warfare agents aerosol, which have low difficulty to release, high hazard efficiency and wide range destruction. At the same time, due to colorless, odorless and extremely small in particle size, bioaerosol are difficult to discover and detect for human senses and instruments detectors, this greatly increases the difficulty of detection and early warning. Therefore, in the event of bioterrorist attack or biological warfare agents threat, early rapid identification and authentication disposal can greatly reduce casualties and biological agents dissemination. Bioaerosol detection technology has a very important strategic significance to the country and the army.For a long time, the method of combining culture, counting, immunology and genetics for the detection of biological warfare agents aerosol are widely used internationally, which could perform qualitative and quantitative analysis, with advantages of stable and reliable. However, the relatively long inspection cycle and requirement of complex human labor make the method above severely constrained in the field real-time bioaerosol detection. With the rapid development of optical technology and electronic technology, bioaerosol detection technology has also been greatly improved, such as laser-induced breakdown spectroscopy and laser-induced fluorescence spectroscopy, are able to analyze biological properties of bioaerosol and achieve fast real-time detection.The current bioaerosol detection technologies have advantages of fast detection speed, high sensitivity and low cost, on this basis, the development of detection devices and testing equipment has been put into practical use in many areas of military and civilian. Nevertheless, detecting bioaerosol in the form of single piece of equipment limits control range and has low efficiency in large scale monitoring applications. In order to achieve continuous monitoring of large areas, multiple devices need to run simultaneously, each piece of equipment must be unified scheduling and centralized operation to get meaningful monitoring results. In order to reduce costs of operation and management, separate operation and controlling should be avoided, which can realize rapid and effective warning of unforeseeable harm incidents. Accordingly, bioaerosol monitoring requires a technology of centralized network management and intelligent information processing, in order to achieve joint operation, cooperative sensing, data sharing and quick warning of multiple detection devices.In summary, an intelligent monitoring system with characteristics of high accuracy, low power consumption and fast warning is required in wide range of on-site real-time monitoring of bioaerosol, which should also has the ability of multi-hop communication and ad hoc network self-organizing. All nodes in this system operate with each other in a collaborative manner, realize information sensing and data gathering of the sensor network, and send the information and data to the observers.Methods and contents: Based on cluster monitoring technology and wireless sensor network technology, with the application background of bioaerosol real-time detection and wide range on-site monitoring, this research explored and studied relevant protocols and specific algorithms of wireless sensor networks layer by layer according to the structure of the network hierarchy. Along with the order from bottom to the top of network architecture, this paper carried out a detailed study of key technologies in five stages to realize ultimately all system function of cluster bioaerosol monitoring wireless sensor networks, including:(1) Research on topology construction and routing algorithm: For the wireless sensor network of bioaerosol monitoring applications, a long-path connected clusters topology was proposed, which is a common non-uniform topology and can well describe the network form of practical application. Considering the problem of node isolate caused by sensor nodes unevenly distributed and hotspot nodes excessive energy consumption, an IGSHMR algorithm was proposed, which is composed of path connected groups and inner group segment hybrid multi-hops. The network is divided into cluster groups and long paths, two areas are classified in groups and paths, each adopts a different multi-hop routing algorithm, namely flat multi-hop routing and cluster multi-hop routing. With rigorous mathematical analysis of IGSHMR, this research modeled its power consumption, selected the optimal segment location and optimized its various parameters. Computer simulation demonstrated the superior improved performance of IGSHMR in terms of network connectivity, energy consumption, node lifetime and other aspects compared with other conventional algorithms.(2) MAC protocol for complex monitoring network: A large number of collection data in wireless sensor networks are transmitted through routing path after the topology has been built. As the operating environment of wireless sensor networks is the actual environment with complex background, it cannot form an effective network system if there is no related controlling and scheduling in transmission control layer, even though the routing layer has high performance. In order to respond effectively to the energy consumption of network operation and frequent changes in the amount of communication data, reduce the seriously affect to network caused by wireless interference and jamming attacks which are existed anywhere in the wireless background, avoid the system appears completely paralyzed in extreme cases, an MDPEE-MAC protocol was proposed based on specific requirements of cluster bioaerosol monitoring, which utilizes the mechanisms of adaptive receiver-initiated multichannel rendezvous and scheduling of predictive wake-up, improves the efficiency and transmission of wireless channel. During the operation, each node is able to dynamically optimize and select the wireless channel, resist wireless interference and jamming. Meanwhile, the optimization is based on the perceived channel conditions, without using a control channel. Additionally, computer simulation demonstrated the superior improved performance of MDPEE-MAC compared with other MAC protocols.(3) Research on compressed sensing and fault detection: Wireless sensor networks will generate a huge amount of collecting information and gathering data during the cluster monitoring. In order to reduce repeat information, control command information and redundant information, this paper optimizes the relevant data in the data processing, compressed sensing is used in multi-hop wireless networks, which combines routing and aggregation. On the other hand, fault detection for sensor of WSNs based on adaptive genetic algorithm and support vector machine is proposed in the study, the output residual error between the actual output value and the standard output value of sensor of WSNs is applied as diagnosis parameter, the training parameters of support vector machine are selected by adaptive genetic algorithm, and the selected parameters of support vector machine are used to create diagnostic model for sensor of WSNs with high performance.Therefore, a high efficiency and intelligent dynamic fault detection of large-scale sensor networks is accomplished. Simulation results showed the efficiency of data compression and fault detection, which not only effectively reduced the amount of communication data and energy consumption in sensor network, improved the evaluation capabilities of environmental parameters, but also maintained a high performance of diagnostic capabilities on various fault status of sensor nodes.(4) Hardware and software design of cluster monitoring wireless sensor network system: Researches of wireless sensor networks in this paper are used for cluster bioaerosol monitoring, therefore, sensor nodes, relay nodes, sink nodes and monitoring terminal are designed respectively depending on the different functions and purposes. Meanwhile, power supply module, processor module, communication module and data collecting module of each sensor node are designed, also with human-computer interaction and database management of monitoring terminal. According to the steps of divided layer and concept of package designing, each module realizes its various functions internally, only keeps fewer key interfaces externally. After designing and realizing different function parts of each node individually, combine and unify all of the modules into a whole system. Design of the whole system includes hardware design, software design and structure design.(5) Performance tests and experimental verification: To validate protocols and algorithms in each layer of the network in this paper and detect the overall performance of the actual system, a real wireless sensor network system and its detection platform were built, which consist of sensor nodes, sink nodes, monitoring terminal that have been designed and equipments such as network analyzers and computer terminals. The network performance was tested in detail at multiple layers, and the correlation experimental results were analyzed and discussed, including basic communication test, ad hoc network and routing test, MAC protocol test and running experiment of cluster monitoring wireless sensor network system.Result: In order to verify the performance of ad hoc network, data transmission and energy efficiency in cluster bioaerosol monitoring wireless sensor networks, this paper built a series of experiments based on the computer simulation model and experimental platform of actual network.(1) Computer simulation: In simulation experiment of routing algorithm, the proposed protocol and algorithm were able to build a network topology successfully, which not only reduced energy consumption, improved network connectivity, prolonged survival time of wireless sensor network nodes, but also alleviated the problem of isolated nodes, avoided the impact caused by network hotspots. By comparing the experimental performance between different MAC protocols, the MAC protocol used in this research could achieve a 100% data delivery ratio, simultaneously with a relatively low duty cycle and transmission delay. The research also conducted computer simulation of compressed sensing based on greed inspired and fault detection based on support vector machine and adaptive genetic algorithm, which proved the efficiency of these methods and demonstrated the great improvement on energy efficiency.(2) Performance tests of actual network systems: The various components of the network system were designed according to the requirements of the research, and actual monitoring network system was built, which showed that performance tests results and simulation experiments were identical. Communication performance experiment of wireless communication capabilities tested the RF module, identified the physical layer functions of cluster monitoring network. Based on simulation results, experiments of topology construction and routing efficiency tested performances of ad hoc organization and data transmission through routing paths in actual wireless sensor networks. MAC protocol experiments not only demonstrated advantages of the proposed method in this research in basic performance compared to other MAC protocols, but also proved the robustness of this method in complex environments through experiments of wireless interference, jamming attacks and other tests. Finally, the whole network system experiment verified the performance and stability of the cluster wireless sensor network in bioaerosol monitoring.In conclusion, computer simulation and performance experiments of practical network system showed that the method and technology of each network layer proposed in this research are able to fully apply to the specific needs and have great advantages compared to other methods. Meanwhile, system performances of the whole network were stable and reliable in a wide range, which achieve a high efficiency of cluster bioaerosol monitoring.Conclusion: In on-site real-time detection of bioaerosol and biological terrorist warning, cluster bioaerosol monitoring wireless sensor network in this research is capable of monitoring tasks including reliable collection, data analysis and wireless transmission in a variety of complex situations, which has a huge improvement in aspects of detection range, collaborative work, warning speed and energy efficiency compared to the conventional single-form detection device and equipment. Meanwhile, all the network layers are studied and designed in detail based on specific requirements of bioaerosol detection in this paper. Considering the complexity of operating environment and variety of influencing factors, optimization is performed on protocols and algorithms of network layers. Wireless sensor network technology provides intelligence and information solutions for bioaerosol detection and monitoring of wide range and complex areas. Under the support of database server, in-depth analysis, processing and evaluation can be conducted on big data of sensor network through systems of data mining and pattern recognition. Cluster bioaerosol monitoring wireless sensor network has ability of effective on-site detection, rapid assessment and hazard warning in the event of bioterrorist incident and biological warfare agents attack, which is not only of great significance in national security and social stability, but also plays an important role in guaranteeing forces security and monitoring battlefield environment.
|
PDF Full Text Request