Abstract

Keywords

References

Abstract

At present, wireless sensor networks are developing rapidly with the support of the Internet of things. Wireless sensor networks can deliver the information people need at any time, free from the constraints of time and space. Wireless sensor network is widely used, which lays a solid foundation for the development of Internet of things. As the node deployment environment of wireless sensor networks is usually very complex, it is necessary to study the security of wireless sensor networks so as to reduce security threats and network attacks. In this paper, the security and application of the wireless sensor network are emphasized.

1. Introduction

Wireless Sensor Network (WSN) is perception, and the main tool is sensor. The use of sensors can effectively sense the external environment, and then with the help of wireless network for information transmission, to meet user needs. The security of wireless sensor network must be paid attention to, because it is supported by high technical content and complex structure, once there is a problem, the consequences are unimaginable. The development of the Internet of things is inseparable from wireless sensor networks, and the ability of people to communicate quickly is also closely related to it [1].

WSN can be used to collect all kinds of data and information and deal with all kinds of complex environments, the application of WSN covers all aspects. In the military field, it can be used to detect the deployment of troops in enemy terrain and other situations, and it has the function of detecting biological and chemical pollution and nuclear radiation. In terms of environmental monitoring and protection, it can be used for data collection in the field environment, tracking animal footprints, analyzing pollution situation, and predicting the explosion of forest fire and debris flow. In the field of industry and agriculture, the monitoring of crop growth and product flow intelligent production. In the field of medical care and health, it can realize the function of obtaining physiological data of patients, analyzing their conditions, and timely receiving medical treatment. In addition, wireless sensor network plays a crucial role in the field of smart home, smart transportation, smart city, cultural relic protection and business, etc. Wireless sensor network is changing people’s life [2].

The information security of WSN is concerned in many contexts, such as military medical disaster prevention and other fields. Wireless sensor networks use open wireless communication channel technology to transmit data, but without security protection means, data is very vulnerable to internal and external attacks [3]. However, the amount of computation based on the cryptographic defence method is not suitable for WSN network. Therefore, it is also a big problem to choose an appropriate encryption method to ensure the security of WSN information [4] .

Abstract

Keywords

1. Introduction

2. Cost-effective domain generalization

3. An implementation of cost-effective domain generalization

4. Experiments and results

5. Conclusion

CRediT authorship contribution statement

Declaration of Competing Interest

Acknowledgment

Research Data

References

Vitae

Abstract

Although deep learning has achieved remarkable successes over the past years, few reports have been published about applying deep neural networks to Wireless Sensor Networks (WSNs) for image targets recognition where data, energy, computation resources are limited. In this work, a Cost-Effective Domain Generalization (CEDG) algorithm has been proposed to train an efficient network with minimum labor requirements. CEDG transfers networks from a publicly available source domain to an application-specific target domain through an automatically allocated synthetic domain. The target domain is isolated from parameters tuning and used for model selection and testing only. The target domain is significantly different from the source domain because it has new target categories and is consisted of low-quality images that are out of focus, low in resolution, low in illumination, low in photographing angle. The trained network has about 7 M (ResNet-20 is about 41 M) multiplications per prediction that is small enough to allow a digital signal processor chip to do real-time recognitions in our WSN. The category-level averaged error on the unseen and unbalanced target domain has been decreased by 41.12%.

1. Introduction

Wireless sensor networks (WSNs) typically are designed to detect and identify neighboring objects in wild [1, 2, 3, 4] with sound or vibration sensors in the form of single [5] or microarrays [6]. The sound or vibration sensor has many advantages [7, 8, 5], such as low cost, low energy consumption, and relatively low in algorithm complexity. However, they are unsuitable for mixed objects detection because their spatial resolutions are usually too low to distinguish each person in a group of pedestrians. To overcome this shortage, we have employed cameras in our WSNs which has been proved to be effective for dense targets identification [9]. Unfortunately, images captured by WSNs are noisy, such as low in illumination, resolution and photographing angle, which are different from most publicly available datasets. Because the severe limitation in data and resources, despite the rapid development in deep learning [10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20], WSN-applicable deep-learning-based image classification algorithms evolve slowly. So, cost-effective dataset construction methods are needed urgently to build datasets that corresponding to specific WSN applications. Several random images of the target application (target domain) that was captured during our field experiments have been shown in Fig. 1, where targets like persons and cars are hard to identify.

Because of limited communication bandwidth, WSNs cannot run deep neural networks (DNNs) in a remote cloud (or fog) which is a common strategy for embedded devices [21, 22, 23, 24, 25]. To run DNNs in such devices locally [26, 27], a training strategy is wanted to cut computation costs without decreasing identification accuracy significantly. Fortunately, Han et al. [28, 29] have pointed out that only parts of weight parameters in neural networks are playing essential roles during predictions. Therefore, it is possible to train an efficient DNN for WSNs with fewer parameters if we can fully utilize key weight parameters.

Software Defined Networking (SDN) has focused enormous attractiveness in changing conventional network by means of offering flexible and dynamic network management. It has drawn important concentration of the researchers from together academia and industries. Mainly, integrating SDN in Wireless Body Area Network (WBAN) applications specifies capable results in terms of handling with the issues like traffic management, security, energy efficiency etc. Recent improvements in miniaturization and energy efficient physiological sensor designs in SDN based Wireless Body Area Networks (WBANs) paved the way for health monitoring systems for collection and processing the real-time physiological data. The collection of signals from different sensor allows reliable diagnosis in heterogeneous than in homogeneous WBANs. Inspired by the evolutions of heterogeneous WBANs, a study on Wireless Electroencephalography Sensor Networks (WESNs) is carried out under distributed signal processing. The distributed WESNs are designed under two different hierarchy i.e. Hierarchical Fully-Connected Topology (HFCT) and Ad-Hoc Nearest-Neighbor Topology (ANNT) to improve the energy-efficiency using distributed Multi-channel Weighted Weiner Filter design (MW2F). Here, each module transmits linear combination of local channels with other modules. The power efficiency is improved Journal Pre-proof in MW2F signal processing algorithm by avoiding centralization of EEG data. A case study is carried out to test the reduced energy consumption after the removal of eye blink artifacts and it is tested with centralized counterparts. The MW2F is evaluated in both topologies against centralized environments and significant reduction of eye blink artifacts improves the energy efficiency in HFCT than other topologies.

Introduction

The energy efficient and miniaturization is the most hot research area in health monitoring systems, where the body is placed with several sensor nodes. These nodes can collect the EEG data and process it in real time and measures are considered to improve the energy efficiency. This system is called as Wireless Body Area Networks (WBANs) [1], and it offers higher demand in the area of physiological monitoring in hospital environment. The WBAN system is composed of several nodes with physiological sensors, wireless communication devices and a signal processing unit. These nodes can communicate with other sensor nodes placed over the body or it can be connected to a fusion center or a storage device. The signals collected from various resources are analyzed jointly in heterogeneous WBANs in real time environment for medical diagnosis.

Opportunistic underwater sensor networks (OUSNs) are deployed for various underwater applications, such as underwater creatures tracking and tactical surveillance. However, the storage capacity of the sensor nodes in such networks may be insufficient, especially when a wealth of data messages are generated rapidly in some emergency response applications. The message dissemination in OUSNs therefore may differ significantly from those in wireless sensor networks or delay-tolerant networks, where network throughput should be taken as one of the primary objectives of network performance. To this end, the strategies for message storing, disseminating and discarding are investigated, and a Message Dissemination Approach for Storage-Limited (MDA-SL) OUSNs is proposed. In MDA-SL, the messages are preferred to be disseminated to the nodes with higher speed or larger residual storage. In addition, the copies of newer messages are inclined to be discarded when their message holders’ storage is full. Simulation results demonstrate the excellent performance of MDA-SL, showing that it can achieve satisfactory network throughput with propagation delay being restricted according to the diverse application requirements.

Introduction

With the broad deployments of mobile sensor nodes, opportunistic mobile sensor networks (OMSNs) [1,2] are introduced to conduct large-scale sensing at a lower cost compared to that of a ubiquitous static infrastructure of sensing devices. Because of the node mobility, however, the available contacts between nodes may be scarce and short, leading to some unstable communication paths. Opportunistic message dissemination techniques enable the network nodes to communicate in an environment where the contemporaneous end-to-end paths are unavailable or unstable, by allowing a data message to be transferred from source to destination in discrete hops even when an end-to-end communication path never emerges. As a derivative form of OMSNs, the opportunistic underwater sensor networks (OUSNs) [3] technology enables various underwater applications, such as underwater creatures tracking [4] and tactical surveillance [5]. However, compared with massive collected data, the storage capacities of nodes are usually insufficient [6,7]. What is worse, the network throughput (which is defined as the total number of data messages received per unit time by the des- ∗ Corresponding author. tinations of all the multi-hop flows in the network) is significantly restricted because the capacity of nodal storage modules is extremely limited (the space memory is even measured in KB [8], e.g., the chip CC2430 has a flash memory of 128 KB [9]).

Effective energy control while maintaining reliable monitoring performance becomes a key issue in wireless sensor networks (WSNs) based surveillance applications. While importance difference of surveillance zone, limited energy and dynamic network topology pose great challenges to surveillance performance. It is necessary to adjust sensor nodes awakening frequency dynamically for information fusion. Thus an energy-aware scheduling with quality guarantee method named ESQG is proposed in this paper which considers sensor nodes residual energy, different importance degrees of the surveillance zone and network topology comprehensively. It first uses a Voronoi diagram to determine the effective scope of each sensor node and then calculates node importance according to its residual energy and the importance degree of the effective scope. Then ESQG utilizes the importance of individual sensing scope and current forwarding costs to further compute node importance and awakening frequency for information fusion. In this way, ESQG can dynamically adapts each nodes awakening frequency to its dynamic network topology and importance degree of each individual sensing scope. The nodes are then turned on stochasticlly via the node awakening probability and node importance based information fusion is conducted for target detection. Besides, an adaptive process of perception factor C is proposed to match actual situation, and automatically change according to the detected data. Experiments results demonstrate that the proposed method ESQG can reduce the number of awakening nodes to a large extent while maintaining high reliability via information fusion.

Introduction

With the rapid development of integrated circuits, digital signal processing and low-range radio electronics, wireless sensor networks (WSNs) have received great attention for wide application potentials. WSN based applications enlarge 5 human capabilities to remotely interact with the physical world. These applications include battlefield surveillance [1], target tracking [2], home appliances and inventory tracking in which information fusion [3] is needed. For example, in the monitoring of Blue-green Algae Bloom on Lake Tai [4], its physical information (eg., water temperature, water color) can be obtained to predict biolog10 ical growth and potential ecological disasters via information fusion. However, WSN enjoys some unique characteristics such as limited energy supply. That is because sensor node has a finite energy reserve supplied by a battery and is often unfeasible to recharge the battery, effective energy control while maintaining reliable detection performance is key problem to wireless sensor networks 15 surveillance. Thus minimizing energy consumption while maintaining system performance via information fusion remains a high priority in designing wireless sensor networks. To prolong the lifespan of a network, sensor nodes are often scheduled to sleep dynamically. While adjacent nodes share common sensing tasks, which 20 implies that not all sensor nodes are required to perform the sensing task in the whole system lifetime. That is to say, as long as there are enough working nodes, the function of the whole system will not be affected by some sleeping nodes. Therefore, if the sensors can be well scheduled, the system lifetime can be prolonged correspondingly; i.e. the system lifetime is prolonged by exploiting 25 redundancy [5]. In order to increase the node energy utilization and lengthen the lifetime of the wireless sensor network (WSN), a novel clustering node sleep scheduling algorithm based on information fusion is proposed in [6]. But these works mentioned above does not consider providing differentiated surveillance services for different importance degree areas while maintaining high enough 30 detection probability to targets. All of them do not consider link fluctuations and other network dynamics. What is more, sleep scheduling mechanism is the most widely used technique for efficiently managing network energy consumption. The author of [7] provides a survey on energy-efficient scheduling mechanisms in wireless sensor networks that has different network architecture 35 than the traditional wireless sensor networks. However, although much more energy can be saved by reducing the working time of sensor module [8], duty cycling usually wastes energy due to unnecessary wakeups, and low-power radios are used to awake a node only when it needs to receive or transmit packets while a power-hungry radio is used for data transmission and information fusion.

1- Introduction

2- Layouts

3- Sampling techniques

4- Communication technology

5- Prediction models

6- Decision support systems

7- Conclusion and future work

References

Abstract

Integration of Wireless Sensor Networks (WSN) in the field of agriculture has provided a new direction for the production of crops. The same can be applied particularly to greenhouses. Greenhouses provide a protected environment for plants and crops. This paper reviews the role of WSN in monitoring of greenhouses. The paper presents an end-to-end survey; starting from the layout of crops in greenhouses, wireless technology used for communication of sensors, techniques to choose transmission interval or rate. Layouts and sampling techniques are also classified and compared. Other than these, the paper also studies various prediction models and decision supporting techniques adopted in greenhouses for efficient integration and management of WSN. These prediction models are also comprehensively analyzed in this work.

Introduction

Modern day scenario of agriculture has changed. Over the years, population has increased. According to UN’s world population prospect 2017 report, population is estimated to reach 9.8 billion by 2050 (United Nations, 2017). Along with population, demand for food is also multiplying. According to Food and Agriculture Organization of the United Nations (FAO), although there are 30,000 types of edible plants available, humans cultivate only 4% of it (FAO, 1999). The agricultural land has considerably reduced due to various factors like urbanization and industrialization. To cope up with these situations, modern-day technological solutions are required. Other than this, water scarcity, increase in fertilization and dynamic climate changes have also strained to integrate technology to achieve required production in agriculture with minimum wastage of resources. WSN is one prominent technology that has developed over past few years and has found its path in many applications like military, defense, healthcare, agriculture, etc. In agriculture, WSN is particularly used to achieve a state called Precision Agriculture (PA). PA works on the phenomenon of observing (using sensors) and responding (using actuators) and tries to achieve parametric values and conditions required for optimum health and yield of crop. PA also focuses on optimizing resources used for production. Fig. 1 shows the complete process of monitoring and control in a field. Sensors are placed at required stations in suitable and optimal deployment. Sensed information is sent to central repository through available wireless channel.

The lifetime of underwater sensor networks (USNs) can be prolonged significantly thanks to wireless power transfer technology. In this paper, we first proposed a Shortest Path Partial Charging based on Charging Curve Scheme (SPBS) to increase the survival rate of nodes in 3D underwater networks, and then we proposed a concept of secondary charging stations for mobile charging ships to reduce the traveling cost and improve charging efficiency. We first use k-means clustering algorithm to divide our network with k clusters, and then we place our secondary stations at k clustering centers, in this way, mobile charging ships can be charged at secondary stations quickly. Based on secondary stations, we proposed Hamilton Charging Scheme (HCS) using the Hamilton ring, and then we proposed a temporal and spatial collaborative charging algorithm (mCS-TS) for USNs with multiple mobile charging ships and secondary charging stations, which also takes the cluster factor and deadline time into consideration. Simulation results show the effectiveness of our proposed algorithms.

Introduction

With the development of emerging information and communication technologies, such as intelligent Internet of things, 5G, cloud computing, artificial intelligence and machine learning, great changes have taken place in people’s lifestyles. In traditional networks, sensor nodes are powered by batteries with limited capacity, which limits the working hours of sensor nodes. Battery exhaustion means the end of the sensor node lifetime. Therefore, the battery capacity becomes a main factor that restricts the lifetime of the whole sensor network. Fortunately, the advancement of wireless communication and microelectronic technology furthermore contributed to the emergence of wireless sensor networks and bring a new choice for extending batteries lifetime. Wireless Sensor Networks (WSNs) are composed of a large number of sensors that are deployed in the monitoring area in a self-organizing and multi-hop manner. Characterized by low cost, low power consumption and multiple functions [1]. WSNs are widely used in important fields such as environmental detection, military, smart home and telemedicine [2], [3]. In remote areas or areas where human intervention is not suitable, replacing the battery for the sensor nodes are troublesome and costly. Therefore, it is necessary to extend the lifetime of WSNs, and the energy consumption of sensor nodes in the area must be equalized and then study the energy supply technology of sensor nodes.

The use of enhanced clustering techniques and multiple mobile sinks can improve the data-collection efficiency and reduce the energy consumption of wireless sensor networks. Most of the existing approaches use static sinks and multihop routing, which can cause data loss and the early death of sensor nodes, leading to energy-hole problems and inefficient data collection. Nowadays, sensors produce time-sensitive data, and hence, lossy data-collection approaches must be avoided. In order to overcome the aforementioned issues and improve the network lifetime, we propose an enhanced clustering methodology with multiple mobile sinks for efficient data collection. The performance of the proposed method is tested and compared with that of the existing algorithms LEACH, GA, and PSO. The simulation results show that the proposed approach improves the network lifetime significantly and reduces data loss.

Introduction

Sensors are small electronic devices used to monitor or measure changes in real-time environments [1]. They are usually inexpensive and low-power-consuming devices, which are capable of processing and transferring data to a base station. After successful deployment, the sensor nodes form a network, by selforganization, which is called the wireless sensor network (WSN). The primary work of a sensor node is to log the analog information of environmental changes and convert it into digital information with the help of analog-to-digital converters. In addition to logging information, the sensor nodes also act as repeaters to forward the data from other sensor nodes to the base station through single or multihop communication. Generally, the nodes are operated with limited battery resources, and the process of recharging or replacing the battery after deployment is tedious. Thus, the limited lifetime of battery is one of the fundamental issues in WSNs [2–4]. Sensor technology has various applications, which include industries, automated homes, military applications, modern health care, environmental monitoring, and security, to name a few [5–9]. In WSNs, information from the sensor nodes is usually transferred to the sink through multihop communication, in order to reduce the energy consumption of the nodes. Though such a data- ∗ Corresponding author. E-mail address: yoonmojung@skku.edu (Y.M. Jung). forwarding approach improves the overall lifetime of the network, it also causes energy-hole problems and the premature death of sensor nodes [10,11]. In particular, with this approach, the sensor nodes close to the sink forward not only their own data to the sink but also the data from their neighboring nodes.

Link quality is an important factor for nodes selecting communication links in wireless sensor networks. Effective link quality prediction helps to select high quality links for communication, so as to improve stability of communication. We propose the improved fuzzy C-means clustering algorithm (SUBXBFCM) and use it to adaptively divide the link quality grades according to the packet reception rate. The Pearson correlation coefficient is employed to analyse the correlation between the hardware parameters and packet reception rate. The averages of the received signal strength indicator, link quality indicator and the signal to noise ratio are selected as the inputs of the link quality estimation model based on the XGBoost (XGB_LQE). The XGB_LQE is constructed to estimate the current link quality grade, which takes the classification advantages of XGBoost. Based on the estimated results of the XGB_LQE, the link quality prediction model (XGB_LQP) is constructed by using the XGBoost regression algorithm, which can predict the link quality grade at the next moment with historical link quality information. Experiment results in single-hop scenarios of square, laboratory, and grove show that the SUBXBFCM algorithm is effective at dividing the link quality grades compared with the normal division methods. Compared with link quality prediction methods based on the Support Vector Regression and 4C, XGB_LQP makes better predictions in single-hop wireless sensor networks.

Introduction

Wireless sensor networks (WSNs) are multi-hop selforganizing networks that are formed by wireless communication, and they consist of large numbers of inexpensive micro sensor nodes that are deployed in a monitoring area [1]. Their purposes are to collaboratively perceive, collect and process the information of the area and then send it to the observer. Sensors, perceptual objects, and observers form the three elements of a wireless sensor network. WSNs currently have broad application prospects in the fields of smart home, urban transportation and so on. In WSNs, nodes communicate by radio frequencies, and links have characteristics such as asymmetry, irregularity, and directionality [2], [3]. The WSN links are susceptible to multipath effect, loss and adjacent channel interference, which result in unreliable links, low channel quality and frequent topological changes [4]. The quality of the communication links has a significant impact on the performance of the wireless sensor network, such as the life cycle of the network, the throughput of the network and the reliability of transmissions [5]. Selecting a link with poor link quality for communications is likely to cause a large amount of packet loss. The retransmission due to packet loss recovery is the main cause of the increase of network energy consumption [6]. Effective link quality prediction helps to select good links for data transmissions, so as to improve network throughput and to maintain normal network operations and topological stability of networks, thereby improving the performance of wireless sensor networks [7]. It can be seen that the link quality prediction for WSNs is of great significance.

A wireless sensor network (WSN) is deployed in an open and sensitive environment like health monitoring, military surveillance, industrial monitoring, landslide detection and so on. It consists mainly of two types of nodes: sensor nodes and sink nodes. The sensor nodes are similar to that of a small computer, but they have limited processing capability, memory and battery power. Due to the resource constraint nature of sensor nodes, the direct transmission of raw data from the sensor nodes to the sink nodes consumes more energy and leads to a lot of congestion in the network. To reduce energy consumption, data aggregation is introduced in a WSN, as shown in Figure 1. More powerful sensor nodes act like aggregators (cluster head) or, in some cases, regular nodes act as aggregators (Castelluccia et al., 2009). These aggregators collect and process the data coming from the other sensor nodes of the network. For example, the aggregation functions like sum, min and max are useful in calculating temperature readings. The aggregator collects the data from n sensor nodes, calculates the sum of all n values and forwards it to the sink node. The sink node, after receiving the sum value, calculates the average temperature by dividing the value by n. The data aggregation saves the energy of sensor nodes and, thereby, increases the lifetime of the network (Papadopouloset al., 2012). Security of data aggregation is a challenging task in WSNs. False aggregation may lead to a large damage in the sensor network. For example, in the case of health-related applications, the attacks on sensor nodes may affect the person’s life, and in the case of military applications, it may lead to a national security threat. Various attacks like node compromise attack, eavesdropping, collision attack, Sybil attack, blackhole attack, sinkhole attack, selective message forwarding and attacks on data aggregation occur. (Xing et al., 2005). So, the secure data aggregation needs the following key properties: confidentiality, integrity, authentication and freshness of data (Papadopouloset al., 2012). The objective of the present study on Secured Data Aggregation (SDA) in WSNs is to securely transmit the aggregated data to the base station by reducing the energy consumption of sensor networks. The Paillier additive homomorphic cryptosystem is used to encrypt the data, and Boneh et al.’s aggregate signature method is used to generate the signature at each sensor node (Boneh et al., 2003).