CN113067785A - Channel selection method suitable for backscattering communication network - Google Patents
Channel selection method suitable for backscattering communication network Download PDFInfo
- Publication number
- CN113067785A CN113067785A CN202110341144.3A CN202110341144A CN113067785A CN 113067785 A CN113067785 A CN 113067785A CN 202110341144 A CN202110341144 A CN 202110341144A CN 113067785 A CN113067785 A CN 113067785A
- Authority
- CN
- China
- Prior art keywords
- channel
- selection method
- layer
- communication network
- channel selection
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000004891 communication Methods 0.000 title claims abstract description 30
- 238000010187 selection method Methods 0.000 title claims abstract description 19
- 238000001514 detection method Methods 0.000 claims abstract description 23
- 238000013528 artificial neural network Methods 0.000 claims abstract description 11
- 230000015654 memory Effects 0.000 claims abstract description 9
- 239000011159 matrix material Substances 0.000 claims description 14
- 230000006403 short-term memory Effects 0.000 claims description 3
- 238000012546 transfer Methods 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims 2
- 238000012795 verification Methods 0.000 claims 1
- 230000005540 biological transmission Effects 0.000 abstract description 2
- 238000005516 engineering process Methods 0.000 description 4
- 238000000034 method Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 2
- 239000000523 sample Substances 0.000 description 2
- 238000012163 sequencing technique Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L25/00—Baseband systems
- H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
- H04L25/0202—Channel estimation
- H04L25/024—Channel estimation channel estimation algorithms
- H04L25/0254—Channel estimation channel estimation algorithms using neural network algorithms
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B7/00—Radio transmission systems, i.e. using radiation field
- H04B7/22—Scatter propagation systems, e.g. ionospheric, tropospheric or meteor scatter
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Power Engineering (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention relates to a channel detection mode, a channel prediction and selection method suitable for a backscattering communication network. Specifically, channel detection mode selection is performed according to link burstiness and channel correlation, channel prediction is performed through a long-short term memory neural network (LSTM) based on a time sequence of detected channel indexes, and channel transmission information is selected according to a prediction result. It includes two detection modes, a passive detection mode and an active detection mode. The invention effectively utilizes the long-short term memory neural network (LSTM) to improve the precision of channel prediction, compared with the traditional wireless communication network, the invention greatly solves the problem of link burstiness in a backscattering communication network, and provides relatively accurate detection channel condition and channel prediction.
Description
Technical Field
The invention relates to a channel detection mode, a channel prediction and selection method suitable for a backscattering communication network.
Background
Backscatter communication networks, such as Radio Frequency Identification (RFID) technology, a short-range, ultra-low power consumption, low cost wireless communication technology, are widely used to construct computing platforms and large-scale deployed sensor networks. With the rapid development of the internet of things, more potential application scenarios are also greatly mentioned, such as: target positioning, gesture recognition, medical detection, and the like. Meanwhile, backscatter communication networks have also been extended to other wireless platforms, such as: Wi-Fi, Bluetooth, etc. It can thus be expected that in the near future, backscatter communications network-related devices will be used on a large scale in everyday life. Therefore, there is a need to tighten the research on backscatter communication networks.
Among them, channel prediction is one of the main problems of wireless communication networks, and the necessity for channel prediction is as follows: on one hand, accurate and real-time channel prediction can realize better adaptive modulation, thereby reducing unnecessary energy loss of the label and ensuring higher network throughput. On the other hand, better channel transmission information is directly selected after channel prediction, and compared with the method of changing a modulation mode, the frequency spectrum is better utilized. The problem of link burstiness in backscatter communication networks still exists, and channel prediction is more difficult due to the lack of pilots to accurately detect channel conditions, as compared to conventional wireless communication networks.
Disclosure of Invention
In order to solve the problem that the prior art lacks of utilizing a deep learning technology to predict a channel through a long-short term memory neural network (LSTM) and finally select a channel to transmit information according to a predicted result, the invention provides a channel selection method suitable for a backscattering communication network. Then, a channel prediction algorithm based on a long-short term memory neural network is introduced to accurately predict the channel index. And finally, selecting the channel with the best current channel condition according to the predicted channel index.
The technical scheme of the invention is as follows:
a channel selection method suitable for a backscatter communication network, comprising a passive sounding mode and an active sounding mode;
the passive detection mode is a light-weight overhead detection mode which is carried out when the current channel index change of a target tag is small, and other tags do not exist or the channel index changes of other tags are also small;
the active detection mode is a comprehensive detection mode which is carried out when the current channel index of the target label is changed greatly or the indexes of other labels are changed greatly.
Further, the channel index of each channel is predicted by a long short term memory neural network (LSTM) of channel indexes obtained by the passive probing mode and the active probing mode.
Further, the LSTM prediction model structure comprises an input sequence X ═ { X ═ X1，x2，x3，...，xn}, timing step and corresponding input xtAnd a forgetting gate f for controlling information transfertAnd input gate itAnd an output gate ot. Input h at time t-1t-1And memory cell status ct-1And the model output of the moment is obtained after the moment is reached and combined with the input of the moment, passes through the input gate and is filtered by the forgetting gate f.
And further, selecting the channels by adopting hierarchical structure model analysis according to the predicted channel indexes of each channel, firstly establishing a hierarchical structure model, then establishing a comparison judgment matrix, and finally calculating the weight and sequencing through the comparison importance judgment matrix of pairwise comparison, namely calculating the geometric mean value of the jth column, calculating the weight of the geometric mean value, and finally checking.
Furthermore, the hierarchical structure model comprises a target layer, a criterion layer and a scheme layer;
the target layer: often the desired goal that the user ultimately wishes to achieve, typically at the top-most level in the hierarchical model.
The criterion layer: the various influencing factors, which are typically intended to achieve a certain goal, are usually located in the middle layers of the model, and the criteria layers may also adopt a nested structure.
The scheme layer: often the actual decision of the final problem, so the general solution layer is often at the very bottom of the model.
Further, a comparison judgment matrix is constructed:
s1) the criterion layer includes all n decision attributes, denoted as P { P }1,p2,...}；
S2) the plan layer includes m candidate plans, and the judgment matrixes a ═ a (a) of n judgment attributes are obtained by comparing each two plans with each otherij)(n×n)Wherein a isijIndicates the importance degree of the attribute i relative to the attribute j, and satisfies aii＝1,aij＝1/aji；
S3) the user gets the value of the element in the decision matrix by comparing the two attributes "which is more important", "to what extent of importance".
Further, the geometric mean of the jth column is:
further, the geometric mean of the jth column is weighted:
the invention effectively utilizes the long-short term memory neural network (LSTM) to improve the precision of channel prediction, compared with the traditional wireless communication network, the invention greatly solves the problem of link burstiness in a backscattering communication network, and provides relatively accurate detection channel condition and channel prediction.
Drawings
Fig. 1 is a main flow chart of a channel selection method suitable for a backscatter communication network according to the present invention.
Fig. 2 is a diagram of detection model selection for a channel selection method of a backscatter communication network according to the present invention.
Fig. 3 is a schematic diagram of an AHP hierarchical model of a channel selection method for a backscatter communication network according to the present invention.
Fig. 4 is a flowchart illustrating a channel selection method for a backscatter communication network according to the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
As shown in fig. 1-4, the present invention provides a channel selection method for a backscatter communication network, including a passive sounding mode and an active sounding mode; based on the real-time feedback and channel correlation of the current channel index, we have designed a lightweight channel sounding scheme that includes two sounding modes, passive and active. The passive detection is a light-weight overhead detection mode which is carried out when the current channel index change of a target label is small, and other labels do not exist or the channel index changes of other labels are also small; the active detection is a comprehensive detection mode which is carried out when the current channel index of the target label changes greatly or other label indexes change greatly. The detection mode selection is shown in fig. 2.
Further, the channel index of each channel is predicted by a long short term memory neural network (LSTM) of channel indexes obtained by the passive probing mode and the active probing mode.
This may lead to false positives, which in turn increases some of the sounding overhead, but we do not want to miss the possibility that any channel may change. Based on the correlation of adjacent channels, the passive sounding mode uses a dichotomy of a single probe to sound a currently unused channel, while the active sounding mode uses a cyclic traversal of four probes to sound.
By detecting two different modes, we can obtain a time series form channel index, and then use long-short term memory neural network (LSTM) to predict each channelThe channel indicator of (2). The neural network can memorize data for a long time and a short time, and a channel prediction model which changes along with the channel condition is obtained through the neural network and a back propagation technology, so that on one hand, a large amount of expenses are reduced, and on the other hand, the prediction precision is also ensured. The LSTM prediction model structure comprises an input sequence X ═ { X ═ X1，x2，x3，...，xn}, timing step and corresponding input xtAnd a forgetting gate f for controlling information transfertAnd input gate itAnd an output gate ot. Input h at time t-1t-1And memory cell status ct-1And the model output of the moment is obtained after the moment is reached and combined with the input of the moment, passes through the input gate and is filtered by the forgetting gate f.
Finally, the channel is selected by adopting an analytic hierarchy process according to the predicted channel index. The analytic hierarchy process for solving the problem of multi-attribute decision generally comprises the following steps: analyzing the problem needing decision making, and establishing a hierarchical structure model, wherein the hierarchical structure model is shown in figure 3. The model of the hierarchy typically comprises:
(1) target layer: often the desired goal that the user ultimately wishes to achieve, typically at the top-most level in the hierarchical model.
(2) A criterion layer: the various influencing factors, which are typically intended to achieve a certain goal, are usually located in the middle layers of the model, and the criteria layers may also adopt a nested structure.
(3) Scheme layer: often the actual decision of the final problem, so the general solution layer is often at the very bottom of the model.
2. Constructing a comparison judgment matrix: after the hierarchical model is established, the positions of the elements in the hierarchical model can be determined, and the user's knowledge of the importance degree of the final expected result in the decision can be expressed by a matrix.
(1) The criterion layer includes all n decision attributes, denoted as P { P }1,p2,...}。
(2) The scheme layer comprises m candidate schemes, and a judgment matrix A of a judgment attribute is obtained by comparing every two schemes with each other (a)ij)(n×n)WhereinaijRepresenting the importance of the attribute relative to the attribute j, satisfying aii＝1,aij＝1/aji。
(3) The user obtains the values of the elements in the decision matrix by comparing the two attributes "which is more important" and "to what extent.
3. And calculating the weight of each element relative to the upper-layer element through the relative importance judgment matrix compared pairwise. Namely, the elements of each layer are subjected to level single sequencing, and the consistency of the decision matrix is checked.
The geometric mean of the column is:
4. calculating the weight according to the geometric mean of the obtained first column
The weighted value of the decision index from the subjective point of view can be expressed as formula (3.7), and the weighted value is calculated by an arithmetic mean method.
5. Finally, a consistency check is performed as shown in FIG. 4 to verify the correctness of the selection.
Claims (8)
1. A channel selection method for a backscatter communications network, comprising a passive sounding mode and an active sounding mode;
the passive detection mode is a light-weight overhead detection mode which is carried out when the current channel index change of a target tag is small, and other tags do not exist or the channel index changes of other tags are also small;
the active detection mode is a comprehensive detection mode which is carried out when the current channel index of the target label is changed greatly or the indexes of other labels are changed greatly.
2. The channel selection method for a backscatter communications network of claim 1, wherein the channel indicator for each channel is predicted by a long short term memory neural network (LSTM) of channel indicators derived from a passive sounding mode and an active sounding mode.
3. The channel selection method of claim 2, wherein the LSTM prediction model structure comprises: input sequence X ═ { X1，x2，x3，...，xn}, timing step and corresponding input xtAnd a forgetting gate f for controlling information transfertAnd input gate itAnd an output gate ot. Input h at time t-1t-1And memory cell status ct-1And the model output of the moment is obtained after the moment is reached and combined with the input of the moment, passes through the input gate and is filtered by the forgetting gate f.
4. The channel selection method applicable to the backscatter communication network of claim 2, wherein the channels are selected by adopting a hierarchical structure model analysis according to channel indexes of each channel, the hierarchical structure model is firstly established, then a comparison judgment matrix is constructed, and finally the weights are calculated and sorted by comparing the comparison importance judgment matrix two by two, namely, the geometric mean value of the jth column is calculated and the weight calculation is carried out on the geometric mean value, and finally the verification is carried out.
5. The channel selection method for a backscatter communications network of claim 4, wherein the hierarchical model comprises a target layer, a criteria layer, a scheme layer;
the target layer: is often the desired goal that the user ultimately wishes to achieve, typically at the top-most level in the hierarchical model;
the criterion layer: the various influencing factors which are generally used for achieving a certain target are usually positioned in the middle layer of the model, and the rule layer can also adopt a nested structure;
the scheme layer: often the actual decision of the final problem, so the general solution layer is often at the very bottom of the model.
6. The channel selection method for a backscatter communication network of claim 4, wherein the comparison decision matrix is constructed by:
s1) the criterion layer includes all n decision attributes, denoted as P { P }1，p2，...}；
S2) the plan layer includes m candidate plans, and the judgment matrixes a ═ a (a) of n judgment attributes are obtained by comparing each two plans with each otherij)(n×n)Wherein a isijIndicates the importance degree of the attribute i relative to the attribute j, and satisfies aii＝1，aij＝1/aji；
S3) the user gets the value of the element in the decision matrix by comparing the two attributes "which is more important", "to what extent of importance".
7. The channel selection method for a backscatter communication network of claim 4, wherein the geometric mean of the jth column is:
8. the channel selection method for a backscatter communication network as set forth in claim 4, wherein the geometric mean of the jth column is used for weight calculation:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110341144.3A CN113067785A (en) | 2021-03-30 | 2021-03-30 | Channel selection method suitable for backscattering communication network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110341144.3A CN113067785A (en) | 2021-03-30 | 2021-03-30 | Channel selection method suitable for backscattering communication network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113067785A true CN113067785A (en) | 2021-07-02 |
Family
ID=76564693
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110341144.3A Pending CN113067785A (en) | 2021-03-30 | 2021-03-30 | Channel selection method suitable for backscattering communication network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113067785A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140337263A1 (en) * | 2013-05-07 | 2014-11-13 | Iotelligent Technology Ltd Inc | Architecture for implementing an improved neural network |
| CN108964815A (en) * | 2018-07-30 | 2018-12-07 | 太原理工大学 | A kind of channel selection and bit rate adaptive approach based on BP neural network |
| US20200234130A1 (en) * | 2017-08-18 | 2020-07-23 | Intel Corporation | Slimming of neural networks in machine learning environments |
| US20200293729A1 (en) * | 2019-03-16 | 2020-09-17 | Nec Laboratories America, Inc. | Tracking rfid groups with spatio-temporal proximity |
| CN111951556A (en) * | 2020-08-20 | 2020-11-17 | 重庆大学 | Bus travel time prediction method considering truck mixing rate based on RFID data |
-
2021
- 2021-03-30 CN CN202110341144.3A patent/CN113067785A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140337263A1 (en) * | 2013-05-07 | 2014-11-13 | Iotelligent Technology Ltd Inc | Architecture for implementing an improved neural network |
| US20200234130A1 (en) * | 2017-08-18 | 2020-07-23 | Intel Corporation | Slimming of neural networks in machine learning environments |
| CN108964815A (en) * | 2018-07-30 | 2018-12-07 | 太原理工大学 | A kind of channel selection and bit rate adaptive approach based on BP neural network |
| US20200293729A1 (en) * | 2019-03-16 | 2020-09-17 | Nec Laboratories America, Inc. | Tracking rfid groups with spatio-temporal proximity |
| CN111951556A (en) * | 2020-08-20 | 2020-11-17 | 重庆大学 | Bus travel time prediction method considering truck mixing rate based on RFID data |
Non-Patent Citations (3)
| Title |
|---|
| CHAO YANG: ""Unsupervised Drowsy Driving Detection With RFID"", 《IEEE》 * |
| RUIQIN BAI: ""RNN-Based Demand Awareness in Smart Library Using CRFID"", 《CHINA COMMUNICATIONS》 * |
| 陈静云: ""车间物联网数据采集关键技术研究"", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Tesfamariam et al. | Earthquake induced damage classification for reinforced concrete buildings | |
| Chen et al. | Fuzzy sensor fusion based on evidence theory and its application | |
| US20210341646A1 (en) | Weather parameter prediction model training method, weather parameter prediction method, electronic device and storage medium | |
| CN112131261B (en) | Community query method and device based on community network and computer equipment | |
| CN107295537A (en) | A kind of method and system for wireless sensor network reliability of testing and assessing | |
| CN112560981A (en) | Training method, apparatus, device, program and storage medium for generating countermeasure model | |
| CN105912773A (en) | Novel intelligent stamping process design method based on data mining technology | |
| Cheriguene et al. | A new hybrid classifier selection model based on mRMR method and diversity measures | |
| CN115734274A (en) | Cellular network fault diagnosis method based on deep learning and knowledge graph | |
| Yuan et al. | Human resource planning and configuration based on machine learning | |
| CN117407781B (en) | Equipment fault diagnosis method and device based on federal learning | |
| De Vita et al. | A Semi-Supervised Bayesian Anomaly Detection Technique for Diagnosing Faults in Industrial IoT Systems | |
| Shrivastava et al. | Selection of efficient and accurate prediction algorithm for employing real time 5G data load prediction | |
| CN118035866A (en) | Fault diagnosis method for aluminum extrusion equipment based on expert network | |
| CN116665482B (en) | Parking space recommending method and device based on intelligent parking | |
| CN108829846A (en) | A kind of business recommended platform data cluster optimization system and method based on user characteristics | |
| US20230297895A1 (en) | Method and apparatus for selective ensemble prediction based on dynamic model combination | |
| Poczeta et al. | Application of fuzzy cognitive maps with evolutionary learning algorithm to model decision support systems based on real-life and historical data | |
| CN113067785A (en) | Channel selection method suitable for backscattering communication network | |
| CN115619099A (en) | Transformer substation safety protection evaluation method and device, computer equipment and storage medium | |
| CN109978138A (en) | The structural reliability methods of sampling based on deeply study | |
| CN113452473B (en) | Method, apparatus and storage medium for predicting optimal channel in backscatter | |
| CN115130620A (en) | Power consumption mode identification model generation method and device for power equipment | |
| Yi | Logistics scheduling optimisation and allocation of intercultural communication trade under internet of things and edge computing | |
| Li | Sports policy and training decision support method based on wireless sensor network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210702 |