Printer Friendly

Wireless sensor network data collection method based on regionalized compressed sensing.

1. Introduction

With the widespread application of wireless sensor network technology, there are more and more usage and researches of wireless sensor technology, such as in the fields of environmental testing, geological exploration and even military information transmission, in which wireless sensor network technology is dominant (Gong, 2011). However, its performance in medical, radar, pattern recognition and image processing fields has been the focus of attention in many fields. Although it's theoretical research is still at the initial stage, its performance in the field of medical treatment, radar, pattern recognition and image processing has gained people's attention (Gu, 2015; Jiang, 2014). Wireless sensor network to meet all the ideas for the transmission of information, which has a low cost, wide range of radiation, equipment, small size and ease of deployment and disassembly and flexible features. Although the wireless sensor network technology has the advantages of this or that, but in the actual information transmission process, large-scale wireless sensor network technology, there are energy consumption problems, which restricts the application of wireless sensor network technology And promotion, which has become the shackles of the development of wireless sensor network technology, and its energy consumption is mainly caused by the central area of large consumption of energy consumption, and thus affect the wireless sensor network information processing results (Jiang, 2015). In order to ensure the information transmission is reasonable and accurate, based on regionalization using compressed sensing technology to solve the problem of network transmission process.


2. Employment Network Information Service Platform

The emergence of wireless sensor is the result of the development of wireless communication technology and electronic device production technology. Its main structure consists of RF transceiver, D/A converter, A/D converter and baseband processor and application program interface (Wang, 2015). Therefore, wireless sensor network has the sensor and micro-electromechanical and wireless communications in one of three technologies (Yang, 2016). The information is collected from the monitoring area and transmitted to the node through wireless communication. After many small nodes are gathered, the signal is transferred to the network or satellite, and the user can use the satellite as the relay station of signal transmission.

The following picture is the classic network structure of wireless network (WSN group).


It can be seen from the figure that the wireless sensor network is large in scale and high in density, which can provide guarantee for obtaining precise information and weaken the accuracy of single sensor node on the other hand. Wireless transmission node based on the collection and collation of information, but the nodes of the energy consumption of the superposition of information transmission, and it will cause information congestion and hysteresis. The sensor node is an important part of the wireless network, the node mainly has the battery, the sensor unit (responsible for monitoring the original object information collection), data processing unit (the original information processing analysis, based on infinite sensor network radiation range, information data of the processing task is large), transceiver devices (information receiving processor, the infinite sensor network information, information send and receive the most arduous work), GPS and mobile devices.


Because wireless sensor network has energy consumption and technical barriers of bottleneck nodes, large power consumption of nodes, limited communication capacity as well as computing and storage capacity.


It can be seen from the figure that sensor node equipment's power consumption is relatively serious, especially at the sending, receiving and idle stage.

Wireless sensor network architecture can be divided into planar routing and clustering routing. This kind of route operation is simple and good expansibility, do not maintain to the network structure. But because the network does not have the management node to the communication management cannot carry on the resource optimization. Cluster routing roommate cluster nodes in each cluster, by cluster nodes in the cluster of reasonable control, in order to achieve the purpose of saving network node energy. As the picture shows:


3. Function of Compressed Sensing Technology

3.1. The Concept of Compressed Sensing

Compressed Sensing (CS in short) is a kind of compressed data information based on the rise of computer network technology, which reduces data transmission and node energy consumption. The basic idea is to acquire the information by non-adaptive projection method, and to reconstruct the original signal with a little measurement value. Signal conditioning samples can be made in specific areas. Compression-aware technology can reduce the energy consumption of transmission and fusion mainly because it can reduce the huge energy consumption in the central area of the network. However, for the practical application effect of CS technology, the advantage of compression sensing is to realize energy-efficient data communication transmission strategy.

Although the wireless sensor network can reduce the energy consumption of the central area through some special transmission methods, the energy consumption phenomenon still appears in the actual operating environment, and the performance of the sensor has high dependence on the actual operating environment and conditions. In order to achieve the maximum effective transmission of energy, the research in this paper is mainly based on regional compression sensing wireless sensor network, to talk about data collection methods.

3.2. Comparison of CS Technology and Traditional Signal Collection and Processing

As can be seen from the chart, in the signal acquisition and processing and signal acquisition conditions, CS technology is superior to traditional signal processing and processing technology. To meet the actual environment of the signal acquisition and recovery needs.

4. Data Collection Based on Regionalized Compressed Perception

Regionalized compressed data is abbreviated as RCS. Firstly, the network topology is randomly divided into several regions. The partitioning principle is independent of the characteristics or relevance of the collected data. Then, each region is selected to select a regional center node to be used to compute the data. And then receives the sampled values of other nodes. Subsequently, the central nodes of the area use the CS method to obtain the regional measurements; finally, these nodes send them to sink nodes for data reconstruction.

4.1. Principle of Regionalized Compressed Sensing Data Collection

Regionalized compression-aware network technology has three regions, and any one region contains the central node and the regular node.


Figure: each node in data transmission, the packet transmission and distributed compressed sensing combination to reduce the transmission energy consumption of the transmitter. The whole network node using the non-CS method One divides into two. Half of the other part of the use of CS, still cannot avoid the huge areas of the energy center in the network data collection process, so it is better to be the traditional transmission mode and the advantages of CS based network life extension.

4.2. Principle of Regionalized Compressed Sensor Data Collection

1. Principle of Regionalized Measurement Matrix Design

From the CS theory, it can be seen that the random matrix has a large probability satisfying and the signal sparse base is irrelevant, and it is also used as a compressed perceptual measurement matrix. The characteristic of the measurement matrix is easily destroyed by the wireless network structure. Therefore, the area measurement matrix of the RCS method should reduce the required storage space by avoiding the repeated transmission of the regional measurement value when the regional center node makes the regional measurement value under the premise of satisfying the compression sensing theory.

2. Sampling-Stop Principle

In the actual information sampling process, once the signal recovery result reaches certain accuracy or reaches the preset threshold, the sink node can stop the sampling work, to avoid the network to produce redundant energy waste. In short, the regional measurement matrix design principles and sampling stop principle is practical, can guarantee the sampling accuracy, while effectively reducing network energy consumption. Especially when the network is large, it can reduce more unnecessary transmission energy consumption and prolong the whole network life cycle.

4.3. Basic Idea

In different regions, regional compressed sensing is implemented to disperse the load of the central area to save the energy consumption of the whole network.

4.4. RCS Method Data Transmission Process

* Initialization: Randomly regionalize the entire network so as to obtain several regions R, and then allocate central node C of corresponding region after determining several regions R.

* Collect Regional Data: Transmit the data collected within the region, and for each region and normal node, they can be directly transmitted to the central nodes to collect samples and for reading.

* Generate Regional Measurements: e regional center node generates the area measurement matrix, the data received in the same and the vicinity, and calculates the regional measurement value.

* Measurement Value of the Transmission Area: All the regional center node to the sink node transmission of other regional measurements and complete a complete measurement.

* While Number of samples, perform data reconstruction: Reconstruction accuracy does not meet the sampling stop rules to re-generate regionalized measurements.

In the regional division and the choice of the central node on its impact on network transmission energy consumption, as network data packets continue to node aggregation, CS transmission of the advantages of small scale gradually reflected. The design of regional measurement matrix should also consider the appropriate sparse matrix and then based on the matrix to make the theory of reconstruction accuracy after listening out the sampling stop principle.

4.5. RCS method guarantees

The regional center node needs to weight and sum the received data values using the area measurement matrix to produce an area measurement. As the network size increases, the measured value will be larger, increasing the storage load of the regional center node. At the same time, the generated packets may exceed the maximum payload allowed by the communication protocol, causing the packet to be split into two or more transmissions. Therefore, it is necessary to design an appropriate area measurement matrix to limit the storage and packet length of the central node.

In actual environments, the complexity of sampling objects leads to the number of measurements required is difficult to predict in advance, so the specific number of samples should be based on the original signal reconstruction accuracy. In other words, the sink node does not need to make any further measurements when the reconstruction accuracy is already high enough or the expected requirements are met. In order to avoid unnecessary network transmission of measured values, a decision rule needs to be considered to determine when the sampling process can be terminated.

5. Data Analysis of RCS Experiment

In the data collection process of wireless sensor network, the traditional transmission mode will make the relay node close to the sink node receive and send large amount of data, and form a sink node as the center area. Compared to other nodes, the energy of the sensor in this area will be exhausted faster, resulting in the entire network load imbalance.

5.1. Network node settings

Because RCS uses regionalized transmission, DT and CS are adopted in each area. With the expansion of network size, the performance advantage of this method is more obvious. When the number of nodes is greater than 800, the number of RCS transmission curves begin to become smooth, which means that the number of transmissions based on the RCS method will not increase drastically as the network nodes continue to increase.

5.2. Node forwarding energy consumption

As the data packets of each node are increasing, resulting in the energy consumption of the intermediate nodes to increase dramatically. According to the compression sensing theory, as long as the sensor serial number is set in advance, the sink node can receive enough measurements to accurately recover the sampled value of each sensor.

5.3. Node and Data Transmission Amount

As the number of nodes increases, the growth trend of RCS traffic volume tends to be flat, and its growth rate is stable. Suggesting that nodes and data transmission have a certain relationship may show positive growth.

5.4. Matrix Design

Since the number of transmissions needed for one measurement is not less than the number of nodes in the network, to further reduce the transmission energy consumption, we can consider increasing the matrix value to reduce the length of the packet. By comparing the number of Gaussian random matrix, uniform random matrix, Bernoulli random matrix and other matrices, it is concluded that the larger matrix value may not produce sufficient projection, resulting in insufficient reconstruction accuracy. Therefore, in the choice of matrix, we must take into account the projection and reconstruction of clarity.

6. Conclusion

As information technology and network technology has provided great convenience for people's life, the development of wireless sensor technology and electronic devices makes wireless sensor network technology develop rapidly. After introducing compression sensing and wireless network sensing technology, this paper establishes the principle of information sampling by sensing node resource and compression sensing technology. Analyze the experimental data to explain the theory, to solve the wireless transmission network technology information transmission dilemma.

Recebido/Submission: 05/07/2016

Aceitacao/Acceptance: 18/10/2016


Achievements of scientific research project of Ningxia higher education in 2014 (ningjiaogao[2014]222hao); This research was supported by the Information Engineering Research Center of Ningxia Teachers'University.


Gong, J. (2011). Data Compression Based on Compressed Sensing Technology in Wireless Sensor Networks. Southwest Jiaotong University, 2011.

Gu, Y. (2015). Wireless Sensor Network Energy Saving Data Collection based on Compressed Sensing. East China University of Science and Technology, 2015.

Jiang, L., Wu, D. (2014). The scheme of wireless sensor network data gathering based on compressed sensing. Modern Science and Technology, 18, 48-51, 55.

Jiang, S. (2015). Research of Wireless Sensor Network Data Collection based on Compressed Sensing. Guangdong University of Technology, 12-15.

Pereira, C., Ferreira, C. (2015). Identification of IT Value Management Practices and Resources in COBIT 5. RISTI - Revista Iberica de Sistemas e Tecnologias de Informacao, (15), 17-33.

Wang, H. (2015). Data collection algorithm for wireless sensor networks based on compressed sensing. Guangxin Normal University, 20-25.

Wang, Y. (2015). Data collection method based on compressed sensing in low duty cycle Wireless Sensor Networks. Shanghai Jiao Tong University, 15-17.

Wang, C. (2012). Research on data collection in Wireless Sensor Networks. Beijing University of Posts and Telecommunications, 42-46.

Yang, H., Wang, X. (2016). Data collection method for wireless sensor networks based on Regionalization compressed sensing. Computer Journal, 2, 1-14.

Guoqi Li *

School of Mathematics and Computer Science, Ningxia Normal University, Guyuan, Ningxia, China

* Guoqi Li,
Table 1--The traditional technique of signal processing and CS
contrast figure

                   Traditional signal             CS technology

Steps           Traditional signal          Raw signal, compressed
                collection and processing   sampling, finding sparse
                can be divided into five    solution, signal
                steps: original signal,     reconstruction and
                sampling processing,        recovery signal
                quantization coding,
                transmission storage
                coding and signal

Sampling Rate   The traditional sampling    CS technology is a sparse
                stage signal broadband in   sampling method
                2 times the purchase
                guarantee signal
                acquisition is perfect
COPYRIGHT 2016 AISTI (Iberian Association for Information Systems and Technologies)
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2016 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Li, Guoqi
Publication:RISTI (Revista Iberica de Sistemas e Tecnologias de Informacao)
Date:Dec 1, 2016
Previous Article:Analysis of industrial organization reform based on network economy.
Next Article:Application of computer-aided English teaching mode in college English education based on multimedia online platform.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters