CN105553574A - Support-vector-machine-based MAC protocol identification method in cognitive radio - Google Patents
Support-vector-machine-based MAC protocol identification method in cognitive radio Download PDFInfo
- Publication number
- CN105553574A CN105553574A CN201510243566.1A CN201510243566A CN105553574A CN 105553574 A CN105553574 A CN 105553574A CN 201510243566 A CN201510243566 A CN 201510243566A CN 105553574 A CN105553574 A CN 105553574A
- Authority
- CN
- China
- Prior art keywords
- mac protocol
- cognitive radio
- svms
- label
- agreement
- 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
Landscapes
- Mobile Radio Communication Systems (AREA)
Abstract
The invention discloses a support-vector-machine-based MAC protocol identification method in cognitive radio, thereby realizing identification on four kinds of MAC protocols. According to the invention, on the basis of an MAC protocol signal used for generating a main user network, signal sampling is carried out with a fixed frequency; a power feature and a time feature of the signal are extracted and a feature space is established; and an MAC protocol used by the main user network is identified by using a support vector machine. The practice shows that the MAC protocol identification method has characteristics of high practicability and reliability and the like.
Description
Technical field
What the present invention relates to is MAC protocol recognition methods based on SVMs in a kind of cognitive radio, belongs to cognitive radio technology field.
Background technology
Cognitive radio (CognitiveRadio, CR) be a kind of intelligent wireless communication system, it can sensing external environment, utilize artificial intelligence technology from environment learning, by changing some operating parameter in real time, its internal state is made to adapt to the statistically change of the wireless signal received.Cognitive radio mainly comprises the content of following three aspects: wireless environment analysis, channel status are estimated to control and Dynamic Spectrum Management with prediction modeling, transmitting power.Two the topmost targets realizing cognitive radio are: any time, the height reliable communication in any place and the effective utilization to frequency spectrum resource.
Cognitive radio is as a kind of smart frequency spectrum technology of sharing, by perception external radio environment dynamically to change radio parameter and consensus standard accesses primary user (PrimaryUser, PU) frequency spectrum cavity-pocket (SpectrumHole, thus reach the object making full use of valuable frequency spectrum resource SP).In order to make full use of frequency spectrum resource when not producing interference to primary user, secondary user (SecondUser, SU) needs physical layer and the MAC layer information of extracting primary user, comprises position, through-put power, MAC protocol and network traffics etc.Therefore, be one of its key technology to the perception of medium education (MediaAccessControl, MAC) agreement.
In cognitive radio, be divided into two classes to the research of MAC protocol, the first is a kind of special MAC protocol of wireless network; It two is perceive the MAC protocol that current network using accurately.In order to judge band limits and the time span thereof of frequency spectrum cavity-pocket accurately, to avoid producing interference to primary user, secondary user needs the MAC protocol considering its Web vector graphic that will access.Therefore, do not change procotol to allow time user can access current wireless network, the method for the MAC protocol proposing to use the method dynamic sensing current network of machine learning to use.
In conventional cognitive radio, MAC protocol identification can be divided into following several stages:
First stage: secondary user awareness also records the signal transmission of primary user's network, samples to signal with fixed frequency.
Second stage: according to sampled signal, extracts the power features of signal, comprises power average and power variance.
Phase III: utilize the signal characteristic construction feature space of extracting, and represent with the characteristic set with characteristic attribute, the data member of different agreement in characteristic set is identified, for training SVM (supportvectormachine) grader.
Fourth stage: utilize the grader trained, another stack features data set is classified, obtains the label of different agreement data, namely complete the identification to different MAC protocol.
Because traditional MAC protocol recognition methods only extracts power features as characteristic parameter, can only identify out based on the MAC protocol (as slottedALOHA) of competition with based on the MAC protocol (as TDMA) controlled, and cannot the MAC protocol in every class be distinguished further.In addition, traditional MAC protocol recognition methods is sampled and feature extraction to signal in desirable network environment, and reckon without the fading characteristic of wireless channel and the uncertainty of noise, therefore cannot be applied in actual cognitive radio networks.
The present invention is directed to the shortcoming of traditional MAC protocol recognition methods, carry out corresponding improvement, and propose a kind of MAC protocol recognition methods based on SVMs accordingly.
Summary of the invention
Goal of the invention: for the deficiency of feature extraction in traditional MAC protocol recognition methods, the present invention considers the fading characteristic of wireless channel and the uncertainty of noise, devises the MAC protocol recognition methods based on SVMs in a kind of cognitive radio.
Technical scheme: based on the MAC protocol recognition methods of SVMs in the cognitive radio that the present invention proposes, mainly comprise following several stages.
First stage: secondary user awareness also records the signal transmission of primary user's network, sample, and sampling interval duration is much smaller than channel busy time and idle time of channel with fixed frequency to signal.Owing to considering the fading characteristic of wireless signal, at the signal power p that time i samples
iobeys index distribution, average power is p
m, then its probability density function is
When multiple primary user access channel transmit data simultaneously time, following joint probability density distribution function can be obtained:
So the power sampled at moment i is exactly all primary user's signal power sums, namely
Consider the noise problem of actual wireless channel, signal transmission adds Gaussian noise, n
ifor the instant noise power that moment i samples, the total real-time signal power so sampled at moment i is
P
ni=P
i+n
i(4)
Second stage: according to sampled signal extraction time characteristic sum power features.Record current channel condition due to secondary user awareness, if channel transfers channel busy condition to from channel clear, then record idle time of channel T
i(n), on the contrary then record the busy time T of channel
b(n).When collecting enough power sample, carry out the extraction of power features.N is as power sample number in definition, when collecting N number of power sample, can obtain power average
and power variance
as follows respectively:
Phase III: after settling signal feature extraction, power average, power variance, channel busy time and idle time of channel are formed four dimensional feature space, each sample namely in characteristic set has this four attributive character.In order to can train by the characteristic set pair SVM classifier collected, need to identify the data sample of often kind of agreement.In the present invention, main TDMA, slottedALOHA, pureALOHA and CSMA/CA agreement is as object to be identified, and wherein TDMA and CSMA/CA is based on control class, and slottedALOHA and pureALOHA is based on competition class.The data sample of mark TDMA protocol data sample to be the data sample of 1, slottedALOHA be 2, pureALOHA is 3, CSMA/CA be data sample is 4.
Fourth stage: utilize the grader trained, organizes to another characteristic data set do not identified and classifies, obtain the label of different agreement data, namely complete the identification to MAC protocol.
Beneficial effect: compared with traditional MAC protocol recognition methods, based on the MAC protocol recognition methods of SVMs in the cognitive radio that the present invention adopts, in joining day feature with while improving feature extraction dimension, consider again the fading characteristic of actual wireless channel and the uncertainty of noise.This not only can distinguish a greater variety of MAC protocol, also more realistic wireless communications environment.
Accompanying drawing explanation
Fig. 1 is method flow schematic diagram.
Fig. 2 is that (Fig. 2 a is slottedALOHA and TDMA power features figure to power features figure; Fig. 2 b is the power features figure of CSMA/CA and pureALOHA.)
Fig. 3 is temporal characteristics figure (the temporal characteristics figure of Fig. 3 a to be CSMA/CA and TDMA temporal characteristics figure, Fig. 3 b be slottedALOHA and pureALOHA.)
Fig. 4 is recognition result figure.
Embodiment
As shown in Figure 1, to the MAC protocol recognition methods based on SVMs in cognitive radio, idiographic flow is as follows:
(1) first generation has Rayleigh fading and contains Gaussian noise four kinds of MAC protocol signals, and with fixed frequency, signal is sampled, its sampling interval duration, much smaller than channel busy time and idle time of channel, collects power sample and the time samples of signal.If sampling obtains current channel condition and transfers channel busy condition to from channel clear, record idle time of channel T
i(n), on the contrary then record the busy time T of channel
b(n).
(2) when collecting enough samples, extract power features and temporal characteristics, wherein power features comprises power average and power variance, and temporal characteristics comprises channel busy time and idle time of channel.
(3), after completing feature extraction, the characteristic set with four-dimensional attributive character is set up.Then, the data sample of different agreement is identified, the label of TDMA agreement is 1, the label of slottedALOHA agreement is 2, the label of pureALOHA agreement is the label of 3, CSMA/CA agreement is 4, with this characteristic set with label, training SVM classifier, and according to test result Optimized model parameter.
(4) utilize the grader trained, the characteristic data set not carrying out label is classified, namely completes the identification to different MAC protocol.
Simulation result shows, MAC protocol recognition methods based on SVMs in the cognitive radio that the present invention proposes effectively can identify the MAC protocol that current network uses, compared with traditional MAC protocol recognition methods, a greater variety of agreement can be identified, and the increase of flow G along with network, the recognition accuracy of agreement is higher; When introducing noise, recognition effect fluctuation is comparatively large, so first need in actual applications to carry out denoising, and then identifies.
Claims (4)
1. in a cognitive radio based on the MAC protocol recognition methods of SVMs, it is characterized in that: the secondary user in cognition network is by monitoring and recording the signal transmission of primary user's network, with fixed frequency to signal sampling, extract power features and the temporal characteristics of signal, set up four dimensional feature space, and utilize SVMs to identify the MAC protocol used in primary user's network.
2. in cognitive radio as claimed in claim 1 based on the MAC protocol recognition methods of SVMs, it is characterized in that: consider the fading characteristic of wireless channel in real network and add noise jamming, extract timely power and the Channel holding time of signal, calculate power average and variance as power features, selective channel busy time and idle time of channel are as temporal characteristics, thus set up four dimensional feature space, and represent with the characteristic set with four characteristic attributes.
3. in cognitive radio as claimed in claim 1 based on the MAC protocol recognition methods of SVMs, it is characterized in that: after completing the feature extraction of different protocol signals, produce the characteristic set that two array dimensions are identical.Select the data of its a pair different agreement to identify, wherein the label of the label of TDMA agreement to be the label of 1, slottedAloha agreement be 2, pureAloha agreement is the label of 3, CSMA/CA is 4, it can be used as the training dataset of the grader of SVM.
4. in cognitive radio as claimed in claim 1 based on the MAC protocol recognition methods of SVMs, it is characterized in that: utilize the grader trained, organize to another characteristic data set do not identified to classify, obtain the label of different agreement data, namely complete the identification to different MAC protocol.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510243566.1A CN105553574A (en) | 2015-05-13 | 2015-05-13 | Support-vector-machine-based MAC protocol identification method in cognitive radio |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510243566.1A CN105553574A (en) | 2015-05-13 | 2015-05-13 | Support-vector-machine-based MAC protocol identification method in cognitive radio |
Publications (1)
Publication Number | Publication Date |
---|---|
CN105553574A true CN105553574A (en) | 2016-05-04 |
Family
ID=55832529
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510243566.1A Pending CN105553574A (en) | 2015-05-13 | 2015-05-13 | Support-vector-machine-based MAC protocol identification method in cognitive radio |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105553574A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107231427A (en) * | 2017-06-19 | 2017-10-03 | 中国人民解放军理工大学 | MAC protocol recognition methods based on SVMs |
CN109818811A (en) * | 2019-03-25 | 2019-05-28 | 电子科技大学 | A method of identification distribution class and mixing class MAC protocol |
CN111988102A (en) * | 2020-08-31 | 2020-11-24 | 华侨大学 | GRU network-based MAC information identification method, device, equipment and storage medium |
US10972997B2 (en) | 2017-10-26 | 2021-04-06 | Benchmark Electronics, Inc. | Hybrid time division multiple access (TDMA) and carrier sense multiple access (CSMA) for interference avoidance method therefor |
CN115514690A (en) * | 2022-09-01 | 2022-12-23 | 西北工业大学 | Non-cooperative underwater acoustic network MAC protocol identification method |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TW200904105A (en) * | 2007-07-12 | 2009-01-16 | Univ Nat Chiao Tung | A concurrent transmission MAC method for throughput enhancement and reduction of spectrum sensing in cognitive radio and device thereof |
CN101594617A (en) * | 2009-06-19 | 2009-12-02 | 电子科技大学 | A kind of control channel method for building up of hierarchical spectrum sharing network |
CN101867960A (en) * | 2010-06-08 | 2010-10-20 | 江苏大学 | Comprehensive evaluation method for wireless sensor network performance |
CN102104875A (en) * | 2010-12-30 | 2011-06-22 | 青海西部矿业科技有限公司 | Novel cognitive radio multi-channel MAC protocol |
US8699430B2 (en) * | 2008-10-10 | 2014-04-15 | The Trustees Of The Stevens Institute Of Technology | Method and apparatus for dynamic spectrum access |
-
2015
- 2015-05-13 CN CN201510243566.1A patent/CN105553574A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TW200904105A (en) * | 2007-07-12 | 2009-01-16 | Univ Nat Chiao Tung | A concurrent transmission MAC method for throughput enhancement and reduction of spectrum sensing in cognitive radio and device thereof |
US8699430B2 (en) * | 2008-10-10 | 2014-04-15 | The Trustees Of The Stevens Institute Of Technology | Method and apparatus for dynamic spectrum access |
CN101594617A (en) * | 2009-06-19 | 2009-12-02 | 电子科技大学 | A kind of control channel method for building up of hierarchical spectrum sharing network |
CN101867960A (en) * | 2010-06-08 | 2010-10-20 | 江苏大学 | Comprehensive evaluation method for wireless sensor network performance |
CN102104875A (en) * | 2010-12-30 | 2011-06-22 | 青海西部矿业科技有限公司 | Novel cognitive radio multi-channel MAC protocol |
Non-Patent Citations (2)
Title |
---|
SANQING HU; YU-DONG YAO; ZHUO YANG: "MAC protocol identification using support vector machines for cognitive radio networks", 《IEEE WIRELESS COMMUNICATIONS》 * |
ZHUO YANG: "MAC protocol classification in a cognitive radio network", 《THE 19TH ANNUAL WIRELESS AND OPTICAL COMMUNICATIONS CONFERENCE (WOCC 2010)》 * |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107231427A (en) * | 2017-06-19 | 2017-10-03 | 中国人民解放军理工大学 | MAC protocol recognition methods based on SVMs |
CN107231427B (en) * | 2017-06-19 | 2020-04-07 | 中国人民解放军理工大学 | MAC protocol identification method based on support vector machine |
US10972997B2 (en) | 2017-10-26 | 2021-04-06 | Benchmark Electronics, Inc. | Hybrid time division multiple access (TDMA) and carrier sense multiple access (CSMA) for interference avoidance method therefor |
CN109818811A (en) * | 2019-03-25 | 2019-05-28 | 电子科技大学 | A method of identification distribution class and mixing class MAC protocol |
CN111988102A (en) * | 2020-08-31 | 2020-11-24 | 华侨大学 | GRU network-based MAC information identification method, device, equipment and storage medium |
CN111988102B (en) * | 2020-08-31 | 2022-03-04 | 华侨大学 | GRU network-based MAC information identification method, device, equipment and storage medium |
CN115514690A (en) * | 2022-09-01 | 2022-12-23 | 西北工业大学 | Non-cooperative underwater acoustic network MAC protocol identification method |
CN115514690B (en) * | 2022-09-01 | 2023-08-22 | 西北工业大学 | Non-cooperative underwater acoustic network MAC protocol identification method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107124381B (en) | Automatic identification method for digital communication signal modulation mode | |
CN105553574A (en) | Support-vector-machine-based MAC protocol identification method in cognitive radio | |
CN110086737B (en) | Communication signal modulation mode identification method based on graph neural network | |
Hu et al. | MAC protocol identification using support vector machines for cognitive radio networks | |
Sarala et al. | Spectrum energy detection in cognitive radio networks based on a novel adaptive threshold energy detection method | |
CN102571230A (en) | Distributed collaborative signal identification method based on blind estimation of higher order statistics and signal to noise ratio | |
CN104767692A (en) | Network traffic classification method | |
CN111832417A (en) | Signal modulation pattern recognition method based on CNN-LSTM model and transfer learning | |
CN113779247B (en) | Network fault diagnosis method and system based on intention driving | |
Zinonos et al. | Grape leaf diseases identification system using convolutional neural networks and Lora technology | |
Yang et al. | MAC protocol classification in a cognitive radio network | |
CN101895494B (en) | Stochastic resonance preprocessing-based digital modulation mode automatic identification method | |
CN110225001A (en) | A kind of dynamic self refresh net flow assorted method based on topic model | |
CN105656826A (en) | Modulation recognizing method and system based on order statistics and machine learning | |
Liu et al. | Dynamic traffic classification algorithm and simulation of energy Internet of things based on machine learning | |
Dakic et al. | LoRa signal demodulation using deep learning, a time-domain approach | |
CN105099587B (en) | Cognitive radio frequency spectrum sensory perceptual system detection performance parameter preparation method under mobile context | |
CN112836994A (en) | Big data visualization method for online remote education and online education server | |
CN107995278A (en) | A kind of scene intelligent analysis system and method based on metropolitan area level Internet of Things perception data | |
CN106936527A (en) | A kind of multi-user's cooperation frequency spectrum sensing method based on double threshold | |
Zhang et al. | A spectrum sensing algorithm for OFDM signal based on deep learning and covariance matrix graph | |
CN102158254B (en) | Method and device for receiving signal | |
CN101808327B (en) | Method and system based on visual cognition in radio communication | |
Lackner et al. | User tracking based on behavioral fingerprints | |
CN110390253A (en) | The modulation mode of communication signal recognition methods extracted based on a variety of points of shape spectrum signatures |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20160504 |