Spectrum Sensing and Spectrum Sensing Techniques
Info: 5436 words (22 pages) Dissertation
Published: 12th Dec 2019
In recent years there has been an enormous growth in wireless communication devices and wireless users. The ever increasing demand for higher speed and reliability made researchers think about intelligent radios like Cognitive Radios (CR). But major amount of spectrum is available for licensed users. There are various communication bodies like the International Telecommunication Union (ITU), European Conference of Postal and Telecommunications Administrations (CEPT) also European Telecommunications Standards Institute (ETSI) who works on standards of communication that defines the use of spectrum for licensed and unlicensed users.
Spectrum is a valuable resource in communication. Over the past few years as the use of various wireless technology is increasing rapidly so we either need more spectrum or make efficient use of current spectrum to satisfy their needs. One way of making efficient use of spectrum is employing spectrum sharing technique. There are many spectrum sharing techniques available like energy detection, cyclostationary feature detector, and matched filter. Of the above technique matched filter and cyclostationary gives more accurate result but they are bit complex and computationally harder as compared to energy detection technique. Energy detection is the simplest of above three and computationally less complex.
Survey has shown that at any given time only portion of spectrum is utilized. According to a report published by FCC (Federal Communications Commission – America) in 2003 has set a set of rule for efficient use of spectrum for licensed and unlicensed users. Also OFCOM (Office of Communications – United Kingdom) has noticed the underutilization of spectrum. At any given time only portion of spectrum is utilized. Even if the system says there are no frequencies available, there is still some frequency available. These available frequencies are known as spectrum holes or white spaces. Some of the bands are completely occupied by users while some bands go unutilized. And that is inefficient use of spectrum. We must have noticed that in cases of emergencies like the train bomb blast in July 2006 in Mumbai, India cellular network actually failed to support huge amount of customer at the same time, this was also the case during 9/11 in USA. If we were having cognitive radios at that time peoples would have been able to talk to their families and inform about their safety. As we are moving from 3G (Third generation) to 4G (Fourth Generation) we need to make certain changes in our wireless technologies. Below shown is the measurement of 0 GHz to 6 GHz at Berkley Wireless Research Center (BWRC).
1.3 Aim and Objective
This thesis emphasises on understanding Cognitive radios, the importance of spectrum sensing for today’s world, the issues regarding the same. A simplified Matlab code is used to support our thesis. As the thesis follows you will find more about cognitive radios and spectrum sensing with a technique to generate white space at a specific frequency.
The main objective of the thesis is to do survey on spectrum sensing and spectrum sensing techniques. Then do plan a model for the same. A model can be supported by a Matlab code. And after all this we need to analyse the model we suggested and further improvements that can be done in that.
1.4 Thesis Organisation
Our thesis is organised as follows:
Chapter.2 Literature Review
This chapter begins with a brief history about cognitive radios .Which is followed by a detail explanation of Cognitive radios and spectrum sensing and some of the spectrum sensing techniques used. At the end of this chapter a business model for spectrum sensing and multi resolution of CR is given.
Chapter.3 System Description
This chapter basically deals with sampling and its importance to us. Also you will find technique to get your spectrum at specific center frequency under “generation of White Spaces”.
Chapter.4 Simulation/Design Analysis
As the name suggests, this chapter deals with simulation model and detail explanation of the code used for spectrum sensing using energy detection. In this chapter we have shown the output of the simulation used.
Chapter.6 Future Work
This chapter suggests some of the future work to be done with this thesis which could be useful for further research in this field.
This gives the concluding part of the thesis.
Chapter 2 – Literature Review
In this chapter we are going to discuss about the cognitive radios, like what are they and why are they so important to us. The chapter begins with a short history of cognitive radios, which is followed by a general discussion on OFDM, some of the challenges faced by cognitive radios in real environment, and also some of the applications of cognitive radios. We have tried to explain spectrum sensing in brief and the techniques used for spectrum sensing in today’s world. Finally a business model for spectrum sensing is showed which is preceded by multi-resolution of Cognitive radios.
2.1 History of cognitive radios
Dr. Joseph Mitola III was the first to introduce or propose the theory of Cognitive radios in 1999. According to Dr. Joseph Mitola Cognitive radios will be the radios that are smart and intelligent enough to find the available bandwidth in a spectrum. It will also have knowledge of right information that has to be passed to the user. And user does not have to take some extra effort for that. It is supposed to do this automatically. He has also mentioned in his PhD dissertation that CR is natural extensions of software defined radios. In 2002 the FCC published a report prepared by Spectrum Policy Task Force [SPTF] which says that majority of spectrum is underutilized. And there is actually is not shortage of spectrum but rather we need to make efficient use of the current spectrum. Also in same year 2002 Professor Cave from UK presented a report which speaks of the possibility of selling bandwidth to the user depending on their requirement. But it would not be fair to give unlicensed user allow to use licensed spectrum. So in December 2002 FCC issued a Notice of Enquiry (NOI) to see TV channel bands can be made available to unlicensed users. Then in 2003 FCC forms a set of rules and proposed interference temperature model for keep track of interference. Later in the same year Notice of Proposed Rulemaking (NPRM) tried to see into issues related to cognitive radio technology where it also pointed out Cognitive radios is a advanced technology which could help efficient use of spectrum by licensed users in own network and by sharing spectrum and with unlicensed users by negotiating when required. This encouraged many researchers in the field of cognitive radios. Major progress in cognitive radios took place in the year 2004, where FCC published NRPM which showed possibility of allowing secondary users to use licensed spectrum. FCC opened three bands for unlicensed users which are 6525 to 6700 MHz, 12.75 to 13.15 GHz
and 13.2125-13.25 GHz. This allowed cognitive devices to transmit six times more. Also IEEE standards are working parallel to the FCC’s. Spectrum pooling system by Professor Timo A Weiss from Karlsruhe University Germany), OFDM based Cognitive radios by professor Ian F Akyildiz et al from GIT (USA) are some of the promising works in Cognitive radios. there are many researches are done, many still going on in Europe, Asia, and America, exploring various aspects of cognitive radios. In Jan 2010 first call over a CR network was made in university of Oulu using CRAMNET (Cognitive Radio Assisted Mobile Ad Hoc Network).
2.2 Cognitive radio
With the development of wireless communication devices and technologies in WLAN and WAN spectrum is becoming scarcer. Low frequency bands which are near few GHz are very scarce and highly congested. In current wireless system we are using fixed spectrum allocation scheme. In fixed spectrum allocation scheme a part of spectrum is owned by an operator. Unlicensed users are not allowed to use that spectrum. This leads to the problem of spectrum scarcity. Survey has shown that more than 50% of the spectrum is underutilized. This is where Cognitive Radio (CR) is comes into picture. CR is introduced to solve the problem of spectrum sharing.
Cognitive Radios actually scans the spectrum and during scanning it looks for spectrum holes. The main objective of the cognitive radios is to look for opportunities or white spaces in spectrum band as quickly and as much as possible. And when we say opportunity, opportunity could be in time or frequency domain. Also when we locate this spectrum holes opportunistically we also need to vacant the occupied spectrum as soon as primary user comes back. Here primary user means the users whose spectrum we are using (licensed user) and secondary means the unlicensed users. Cognitive radio is a pattern for wireless communication technology in which either a network or a wireless node changes its transmission or reception parameters to communicate efficiently avoiding interference with licensed or unlicensed users. These altered parameters are associated with the active monitoring of several factors in the external and internal radio environment e.g. radio frequency spectrum, user behaviour.
Cognitive radio can be said as next generation of software defined radio (SDR). They are flexible in terms of their transmission characteristics in terms of frequency, bandwidth, ISP which makes smart decisions to configure the SDR at any point in time to achieve a particular goal. By combining these two technologies makes a radio intelligent and flexible and which helps to adapt it to the variations in the environment, user requirements as per the other radio users. Adaptation to changes and requirements should lead to highly reliable communication whenever and wherever required, while making efficient use of spectrum. Good cognitive radio uses analysis done for long period to know about the environment and also his own behaviour.
There are various parameters taken into account to decide transmission and reception changes, we can distinguish certain types of cognitive radio. The main two are as follow:
- Full Cognitive Radio : It is also known as “Mitola radio” in which every possible parameter which can be observed by a wireless node is taken into consideration to take decision
- Spectrum Sensing Cognitive Radio: It is the type in which only the radio frequency spectrum is considered.
And Depending on the parts of the spectrum available for cognitive radio, we can distinguish as:
- Licensed Band Cognitive Radio: It is the type in which cognitive radio is capable of using bands assigned to licensed users, apart from unlicensed bands, such as U-NII band or ISM band.
- Unlicensed Band Cognitive Radio: This can only utilize unlicensed parts of radio frequency spectrum only or the bands which are free to use.
2.3 About OFDM
OFDM stands for Orthogonal Frequency Division Multiplexing. It is generally a type of Frequency Division Multiplexing (FDM) rather a special case in FDM. What makes it special is its orthogonal behaviour. Now the word orthogonal basically means mutually independence. When we say A is orthogonal to B, we mean that A has no vector in direction of B and vice versa or in other words A and B is mutually independent. Or the integral of two signals over one period is 0. In OFDM a single signal is first multiplexed and modulated independently to create orthogonal signals. This means the signal is first divided into number of smaller streams and then modulated further before transmission. For example imagine a slicing of cheese and grilling or cheese. Slicing is like FDM where whole data is sent in a bunch and grilling is like OFDM where the data to be sent is first divided into smaller data and then processed to transmit further like OFDM. Figure shown below is one more way to understand the concept of OFDM. On the left hand side is a big container which carry whole bunch of data at one time and take it to the destination. And on right hand side is four smaller containers where each carry smaller portion of data and take it to the destination. These smaller containers can be assumed as sub-carriers. And in case of OFDM they are orthogonal sub-carriers. The main advantage over here is that even if some of the cheese is lost during grilling, still we have not lost all the cheese.
In OFDM input data with high data rate is first passed thorough serial to parallel converter. This parallel divided data is then modulated on individually. And parallel to serial conversion is done before transmission. This parallel divided data is our sub-carriers. These sub-carriers must be orthogonal.
2.4 Cognition cycle
Above figure shows the rough model of Cognition Cycle. It reads the surrounding environment and makes decision accordingly. When we say decision we mean cognitive radio sense the requirement or urgency that may be in terms of changing a channel or churning from one technology to another depending upon the scenario. There are different stages where it observes the environment, learns it and then plans its action make required decision and then execute its plans. It is much like a radio with a power of thinking which was not been able before. The figure below shows the cognitive radio architecture suggested by Dr. Mitola.
2.4.1 Some important terms
188.8.131.52Wireless Environment/Outside world- It refers to a communication environment that includes any communication devices and the frequency bands they are working in.
184.108.40.206 Spectrum sensing- It is the technique used by cognitive radios to sense the spectrum. This action involves finding availability of white spaces or spectrum holes in the spectrum.
220.127.116.11 Spectrum Management- It involves catching the best spectrum available so as to satisfy user communication requirements. Cognitive radios should decide on the best spectrum band so that it can meet the Quality of service required for all available frequency bands, therefore these functions are necessary for Cognitive radios.
These management functions can be classified as:
- Spectrum analysis
- Spectrum decision
- Spectrum Mobility: It is defined as the process when a cognitive radio device exchanges its frequency of operation. Cognitive radio networks target to use the spectrum in a dynamic manner by allowing the radio terminals to operate in the best available frequency band, maintaining seamless communication requirements during the transition to better spectrum.
- Spectrum Sharing: providing the fair spectrum scheduling method. One of the major challenges in open spectrum usage is the spectrum sharing. It can be regarded to be similar to generic media access control MAC problems in existing systems
2.5Cognitive Radio Challenges
Three main problems experienced by CR are as follows:
- Interference (Mainly because of Hidden Nodes).
They are described as follows:
2.5.1 Interference and the Hidden Node Problem
Ideally while designing a CR we should consider that it does not have any impact on existing radio users, but in practically some impact is expected. If a particular user have non-cognitive radios, it is essential to study and make a note that how they would be affected by the interference of CR, mainly with respect to sharing resources such as spectrum, time, space etc. CR adaptive nature could be difficult to predict and thus making it hard to control the behaviour of a CR which will concern for user who faces CR interference issue. In communication industry the main concern about CR is the hidden node problem. This scenario arises when a CR is not capable to detect an interference with any of non-cognitive radios within its range, not only because of CR’s own spectrum sensing is ineffective but also due to some non-cognitive radios are hidden. For example, if a transmitting contemporary user is not in the range from the CR, its transmission power may not be strong enough at the CR’s location, it may be reduce than the noise floor which makes it more difficult to get detected. As the CR might not be able to detect a transmission by a contemporary user and similarly unaware of availability of the receiving by a contemporary user. Consequently, if it is confirmed as safe to use the contemporary user’s frequency and CR starts transmitting, at the contemporary receiving end it will create interference. The CR may have a limited view of spectrum provided from wide spectrum measurements which may cause interference with the receiving user. The localised spectrum view denotes that a CR should be potential to find transmitting user those are communicating below the noise level, since the strength of the received signal is very weak at the CR’s location. Similarly a situation can occur where the signal attenuate by distance, thus user transmission is blocked by obstacles such as buildings, towers, hills or mountains. For example consider a CR in a valley would have a limited picture of the surrounding radio environment, as compared to that if it were located on top of a hill.
2.5.2 Security Concerns
CR may be vulnerable to malicious effect, resulting into unexpected or problematic behaviour of individual CRs or complete networks. This problem springs up from the potential to re-program CRs in an unauthorized way. Hacking or placing a vulnerable code, virus on a network might enable criminals to steal valuable information from a CR through electronically, fool a network operator into charging others for services or achieve potentially widespread denial-of-service. A considerable amount of regulatory work will require to be done to clarify who would be responsible for the various security areas of CR, software developers, manufacturers, network operators and CR users themselves may all have a role to play. The CR security issue is closely related to that of SDR, which already discussed and hence not repeated here. Instead, a brief summary of the issues is given. Downloading software updates over an air interface poses some specific problems for security. Several digital signatures will be required for each piece of downloaded software in order to meet likely regulatory requirements. Exactly who is necessary to authorize software downloads must be standardized before any large scale deployment of over-the-air updates can be realized.
2.5.3 Burden of Control and Regulatory challenges
A CR in reality will have some effect on different spectrum users the compliance of these new radios is likely to focus on a “Policy module” defined within a CR, which will determine the boundaries of CR behaviour. It is important while studying CR; to consider how users would not be affected by interference from CR devices and the exact operation and nature of a proposed CR policy system must be understood carefully. It is likely to include a detail case study of the specifications and characteristics of all the contemporary users for a specific CR or complete network of CR’s may share the resources. The effort in controlling CR devices, it is necessary to ensure their behaviour is properly or not, even in the case of faulty or tampered devices, that measures are quickly implemented to intense problems. This will involve policies and standards which are created in a ‘universal’ digitally interpretable policy, so that all CRs can understand the same. Monitoring techniques and powerful algorithms are required to enable detection and identification of ‘bad’ CRs and in this way it provides traceability to find or determine who is responsible for the issue. In addition to these challenges, spectrum regulators and spectrum managers will require providing access to licensed spectrum in such a way that is traceable, transparent and highly dynamic. If CRs are allowed to cover international territory additional effort will be required, due to the necessity to provide and collaborate cooperation with other countries. Assuming that acceptable control of CR policy behaviour is technically possible and feasible, it may turn out to be such a great burden that it will be simply not economically viable and the benefits of CR are outweighed by this burden.
2.6 Important Applications for Cognitive Radio are
- Downloading of audio and video files on mobile handsets. This application requires moderate data rates and near-ubiquitous coverage.
- Emergency services communications: It requires interoperability and a moderate data rate with local coverage.
- Broadband wireless networking: Very high data rate required but CR users have option to accept limited coverage, e.g. hot spots.
- Multimedia wireless and sensor networking: Broad range of data rates may be required.
2.7 Spectrum Sensing
Spectrum sensing is the process performing measurements on the part of spectrum and on the basis of measured data making a decision related spectrum usage. As the requirement and quantity of users is getting increased day by day, it is necessary for ISPs to have large amount of spectrum in order to achieve the QOS (Quality of Service). This leads the interest in unlicensed spectrum access and spectrum sensing is vital concept of this. In a situation where there are licensed user and any unlicensed exists, licensed user (primary user) is to be protected and no unlicensed user can interfere any licensed user’s operation and such cases Spectrum sensing is also useful to detect the existence or non existence of a primary user. Spectrum sensing is an important concept for exploring spectrum opportunities for the secondary spectrum usage in real-time. It detects the unused spectrum and shares it without any noticeable interference with other users. It is an important requirement of the Cognitive Radio in order to sense spectrum holes. Detecting primary users is the most efficient way to detect spectrum holes.
2.8 Spectrum Sensing Techniques Available
Spectrum sensing plays a vital role in cognitive radios. And the type of spectrum sensing techniques to choose more or less depends upon the spectrum sensing technique. A method such as energy detection proves to be one of the simplest of all, but it doesn’t works well at low SNR, varying noise levels, fading. On the other hand technique such as Matched filters shows better performance, but they comes complex receiver design. We are going to discuss some of these techniques as we proceed further.
2.8.1 Matched Filter Technique
This is the technique which takes minimum amount of sensing time. In this method of spectrum detection, receiver receives a pilot signal along with the data that is sent by the transmitter. A pilot signal is a single frequency that is used for synchronisation. All the secondary’s those are struggling for spectrum should have knowledge of this pilot signal. There should` be tight timing synchronisation between primary and secondary. They are also required to know about the kind of modulation being used, pulse shaping. Also secondary’s must have another receiver for every primary. This kind of techniques also fails when there is frequency offset. Examples of this technique are TV signals, CDMA with pilot, also used in OFDM.
2.8.2 Energy Detection technique
This is the most simples of all techniques. In this the receiver has no knowledge of the transmitted signal. The receivers need not to have knowledge about the modulation type or any kind of pilot signal. Earlier energy detection was done with the help of a LPF (Low Pass Filter), Digital to Analog converter (D/A), and square law device that used to calculate the energy of the signal. Later it is done by making use of fast Fourier transform (FFT). This is known as periodogram method in energy detection.
2.8.2 Cyclostationary Feature Detection
Signals are modulated with sine waves or cyclic prefix as in OFDM. And they are periodic. This periodic property of a signal helps it to be cyclostationary. This technique basically uses this principle of spectral correlation to detect the spectrum. Even if signals have similar PSD (Power Spectral Density) but they have do not have similar spectral correlation.
2.9 Multi-resolution for Cognitive Radio Sensing
The concept of multi-resolution for Cognitive Radio can be applied with different methods but, the basic idea is the same. The whole spectrum is first sensed by using a coarse resolution. After this first step fine resolution sensing is done on a part of interested bands. In this way CR avoids itself from sensing the spectrum at one time and thus saving time and power. In this way, the sensing time is reduced and the power also been saved from unrequited computations. Also the multiple antenna architecture helps parallel processing and enables to reduce the sensing time. But, it increases the chip area and consumption of power which is not desirable. Also for coarse resolution sensing the mixer has to produce many frequencies and
Also it should switch to one frequency for beginning the fine resolution sensing. If the signal is low pass signal then we can use fine resolution to scan the whole spectrum. Because low pass signal has low center frequencies and its sampling is doable. But for pass band signals it is not feasible to scan the whole spectrum. Because, for example say if we have some signal with center frequency of 850 MHz, it is not practically possible to do sample that signal. As according to Nyquist theorem sampling rate should be at least twice that of center frequency. Therefore it is practically not possible to sample a signal at 1600 Mega Hertz.
2.10 Business Model for Spectrum Sensing
So far we have discussed about Cognitive radios and spectrum sensing in details. Since this is telecom, and telecom involves huge capital investment. One of the most costly things in telecom is getting the license itself and then comes the infrastructure and installations etc. Currently most of countries work on static spectrum allocation basis. For spectrum sensing to work we need some kind of regulation or set of rules that all will be ready to work with. A team from Brussels University has suggested a model for the same. The same model is discussed in brief below. This model is divided into four main categories.
Ownership simply means the ownership of license. One who has license is authorized to use particular band of spectrum. And if another licensee wants to share a spectrum then it will depends on parameters discussed below. If the operator is unlicensed then there is now issue of ownership.
Exclusivity means whether or not a particular operator is exclusively assigned a band of spectrum. That will be issue of regulator to decide to exclusively assign a spectrum to a specific user. If it is assigned exclusively then nobody can access that band of frequency, and if not then those bands of frequencies will be available for sharing.
Tradability means whether or not a terminal is allowed to switch between frequencies from different operators. If tradability is permitted frequency band or bands can be auctioned for sale or given on lease.
It is possible that some of frequency bands can be accessed by number of RAT’s (radio Access Technology) or may be limited to a particular RAT. If frequency bands are not available to number of RAT’s then that band need to address more issues, such as setting technical conditions to access the band and coordinating the cooperation between multiple technologies.
- Unlicensed: Unlicensed deals with the band of frequencies which are free to use, like ISM band. Common example for this is Wi-Fi which operates in 2.4GHz. This band of frequency is available to all and there is special condition to access this band.
- Single RAT Pool: This pool is related to a group of licensee’s which are not exclusively assigned any band of frequency and using same RAT.
- Multi RAT Pool: It is similar to Single RAT Pool except that it has multiple RATs.
- Single RAT Market: In this each operator is assigned with a separate frequency, but that can be accessed by the conditionally secondary’s.
- Multi RAT Market: In this the operator is a licensee and is exclusively assigned with a band of frequency, and also with tradability.
- Flexible Operator and Static Spectrum: If a particular band of spectrum is exclusively assigned to an operator and without tradability. And if there is only one RAT, then it is known as Static Spectrum else Flexible Operator.
2.11 Survey outcome
As per the literature review we concludes to use Energy detection technique for spectrum sensing as it is the fastest spectrum detection technique available and also it is simpler as compared to other techniques. There are many researches done in this field and many are still doing. Because of the survey we get knowledge about various techniques available in market. This led us to do a simulation model with one of the spectrum sensing technique. Now as far as selection of spectrum sensing technique is concerned, we selected Energy detection technique.
The same could be seen in next part of the thesis which includes its design and implementation in simulation model.
Chapter 3 – Simulation Design
3.1 Sampling and Its Importance
Sampling is a process of converting the continuous analog signal to a discreet analog signal and the samples signal is the discreet time representation of the original signal. If the message is coming from a digital source, then it is in the form to be processed by digital communication systems. But in real life not every signal is digital, message signal can be analog. In situation like these, we have to first convert the analog signal into discreet time signal, this is sampling. For this process to work well, sampling rate should be selected carefully or in other words it should satisfy Nyquist criterion. And Nyquist Criterion says that the sampling frequency should be at least twice the maximum frequency in signal.
Fs ≥ 2fmax or T=1/ fmax
Where “Fs” is sampling frequency, fmax is maximum frequency in the signal. T=Sampling period.
Sampling is like reading a signal in analog form and taking its value at that instant of time. So more the samples we take better the resolution of the signal and, signal can be recovered more accurately. But if we take less number of samples, then resolution of the signal decreases. If we go on reducing the sampling rate, then times comes when it is difficult to recover the original signal from the sampled signal or in other words original information in the signal is lost. This is also known as aliasing. Aliasing is the effect which takes place if the signal is sampled less than twice the maximum frequency.
3.2 Types of Signal
3.2.1 – Time limited signal- It is a signal which exists for only certain duration of time. Out of this duration, signal does not exist. A rectangular pulse of duration “T” seconds can be considered as time limited signal.
x(t)=A … for 0 3.2.2 – Band Limited signals- It is a signal which has a frequency spectrum which exists only over a certain range of frequency. The value of signal outside this range of frequency is zero. Mod(X(f))=A ….-B Low pass sampling theorem states that- a) A band limited signal of finite energy, which has no frequency components higher than W Hertz, is c
3.3 Sampling Of Low pass and Band Pass signal
3.2.2 – Band Limited signals- It is a signal which has a frequency spectrum which exists only over a certain range of frequency. The value of signal outside this range of frequency is zero.
Mod(X(f))=A ….-B Low pass sampling theorem states that- a) A band limited signal of finite energy, which has no frequency components higher than W Hertz, is c
Low pass sampling theorem states that-
a) A band limited signal of finite energy, which has no frequency components higher than W Hertz, is c
Cite This Work
To export a reference to this article please select a referencing stye below:
Related ServicesView all
Related ContentAll Tags
Content relating to: "Information Technology"
Information Technology refers to the use or study of computers to receive, store, and send data. Information Technology is a term that is usually used in a business context, with members of the IT team providing effective solutions that contribute to the success of the business.
Design, build and User-test a Curriculum Vitae Mobile Application for Software Developers
Table of Contents 1. Introduction 2. Review of Literature / Background 3. Methodology/Sources of data 4. Analysis 5. Discussion 6. Reflection 7. Conclusion 8. References 9. Appendices Projec...
Machine Learning Techniques to Predict the Failure of Air Pressure Systems in Trucks using Sensor Data
ABSTRACT The aim of the research is to analyze the sensor reading data using machine learning techniques to predict the failure of air pressure systems (APS) in trucks and to reduce the maintenance...
DMCA / Removal Request
If you are the original writer of this dissertation and no longer wish to have your work published on the UKDiss.com website then please: