ISSN: 2366-1313

Design of AI Algorithm Using IOT By

Embedded System Technology
T Ravi Chandra, Research Scholar, Department of ECE ,
J.S University, Shikohabad.
Dr. Ramireddy Gangadhar Reddy , Associate Professor ,Supervisor,
Department of ECE, J.S University, Shikohabad.

Abstract - The expansion of artificial communication networks and the

intelligence has a substantial beneficial development of integrated circuit design,
influence on the lives of people, and the the study of embedded systems has
development of embedded technology has become one of the most active areas of
a direct impact on the expansion of the research within the information technology
Internet of Things. Both of these industry. Information technology has
developments are having an impact on the developed into a new industrial system in
growth of the Internet of Things. Research China as part of an effort to speed up the
is being done to determine how an country's process of industrialization. An
artificial intelligence algorithm may be embedded system is a kind of computer
included into an embedded device using system that integrates hardware and
this information. The embedded system software and is capable of doing certain
hardware configuration and the embedded specialized tasks all by itself. Embedded
system software running algorithm are systems may also communicate with one
both improved in order to boost the another. It is a comprehensive system that
research of embedded technology and raise incorporates cutting-edge computer
investments in people and material technology, semiconductor technology,
resources in the study of embedded electrical technology, and the specialized
systems. Gaining knowledge of western applications of a variety of different
sophisticated technology, as well as the industries. This innovation in knowledge
embedded technology that is employed in and technology is continual, and the
the Internet of Things, may help increase system relies heavily on both. This article
the performance of embedded systems. takes a look at the status quo of embedded
The results of the experiment show that the systems, provides a concise summary of
artificial intelligence algorithm is able to their component parts, and makes some
effectively improve the performance of the predictions on their potential future
embedded system. development. An embedded system is a
special form of computer system that is
Keywords: Artificial intelligence, constructed using software as well as
Embedded system, Hardware structure, physical computer components. It is
Embedded technology possible to modify software and hardware
in order to fulfill the specific needs of an
1. INTRODUCTION application system with regard to its
functions, dependability, cost, volume, and
As a result of the rapid expansion of power consumption. The combination of

Volume VII Issue I June 2022 www.zkginternational.com 202

 ISSN: 2366-1313

computer hardware and computer software achieve more effective software and
that constitutes an embedded system. In hardware co-design. This change was
most cases, an embedded operating system necessary in order to achieve more
and application software make up a effective software and hardware co-design.
device's software, while the embedded As a result of both the progression of
central processing unit (CPU) and external technology and the growing importance
peripherals make up the device's hardware. placed on system performance, an
The performance of the embedded system increasing number of embedded systems
might be improved by incorporating some are making use of specific hardware
of the fundamentals of artificial optimization and acceleration methods.
intelligence into the optimization process. This article combines a specific hardware
environment with an integrated platform in
2. EMBEDDED SYSTEM BASED order to improve the embedded
ON ARTIFICIAL environment, boost the performance of the
INTELLIGENCE artificial intelligence algorithm for
ALGORITHM recognition, and optimize the performance
of the embedded environment overall. The
 hardware configuration of bulk of embedded devices are composed
embedded system based on of embedded computer systems and the
artificial intelligence algorithm hardware that is connected with them. A
software layer, a hardware layer, a layer of
The continued growth of embedded system software, and an intermediate layer
systems correlates with the fast make up the majority of the embedded
development of embedded technologies. computer, which serves as the core of the
Embedded systems are becoming more system. Other layers include an
and more common. Embedded technology intermediate layer. The majority of the
is becoming more vital as its usage spreads time, when people talk about the software
layer, they are referring to the layer of
software that is used while developing
embedded systems. The hardware layer is
comprised of a variety of components,
including the memory, embedded CPU,
universal device interface, and others. It
serves as the principal module for the
embedded system. The RTOS, the
graphics system, the file system, and the
network protocol are the most important
across more and more applications. As parts of the software layer. The
technology has advanced and the intermediate layer, which is also referred
to as the driver layer software at times, sits
Fig.1 Basic structure of embedded between the system software and the
system underlying hardware. This separation is
what allows the system device driver to
function independently of the hardware.
requirements for system performance have As a direct consequence of this, the
continued to rise, embedded systems have hardware configuration of the embedded
shifted their focus from hardware design to system has been improved. The essential
software implementation in order to structure of the embedded system is shown

Volume VII Issue I June 2022 www.zkginternational.com 203

 ISSN: 2366-1313

in Figure 1. storage of a large number of intelligent

programs, quick execution speed, flexible
addressing, high execution efficiency,
For embedded hardware systems, in fixed instruction length, and established
addition to the basic control components, it hardware conditions for embedded
is possible to take into account the storage, intelligent systems. Currently,
communication, debugging, and display technologies such as fuzzy logic, neural
functions, as well as any other auxiliary networks, artificial intelligence, and a
functions related to the embedded system. variety of others are being successfully
Memory, communication devices, and applied in a broad range of sectors, and the
displays are the three categories that may generally accepted theories are quite
be used to classify the widely used correct. The four most important
embedded peripherals. Memory is the components of the intelligent control
most common kind of embedded module are the knowledge base, the
peripheral. The multi-channel batch inference engine, the knowledge
operating system, the time-sharing acquisition program, the vast database, and
operating system, and the real-time the embedded MCU that stores the data. It
operating system are currently the three is not at all out of the question to develop
types of operating systems that have the the necessary hardware, software, and
greatest number of users. Linux is a group other elements for an embedded intelligent
of operating systems that are based on sensor. In light of this, the master control
UNIX and are used in these systems. The structure of the embedded system has been
task switching technology of the system given some optimization. Figure 2 depicts
enables the system to function as a the structure after it has been optimized.
multiuser environment that is also capable
of genuine multitasking. one of the Both the low-end embedded DSP
numerous advantages offered by the processor and the high-end embedded DSP
embedded operating system. The current processor are put to use in the field of
generation of embedded microprocessors computer communication. One of these
is characterized by the following two types of processors is based on an
embedded microprocessor, while the other
is based on a highly integrated embedded
chip system. The embedded
microprocessor with the intelligent control
module (fuzzy, neural, or artificial
intelligence technology) is comparable to
the intelligent control module, which
stands for the intelligent control module. It
is an intelligent control module that
integrates sensors into its system. It is able
to duplicate the manner in which human
Fig. 2 Main control structure of professionals solve problems and has
embedded system considerable expert knowledge in the
relevant disciplines. It is able to apply
characteristics: small size, low power sound reasoning and is skilled in some
consumption, low cost, high performance, information collection techniques, but in
online control, support for the thumb (16 general, its talents are far more advanced
bit) and arm (32 bit) dual instruction set, than those of specialists. In addition to
large capacity flash memory, low cost,

Volume VII Issue I June 2022 www.zkginternational.com 204

 ISSN: 2366-1313

being adaptive, transparent, and interactive, upon which the multitask real-time
it has a certain degree of intricacy as well processing platform of the embedded
as a level of difficulty. system is constructed. As illustrated in Fig.
3,
 Embedded system software
operation algorithm design based
on artificial intelligence
algorithm

Because of the continual development of

embedded systems, embedded technology
is now widely employed across a wide
variety of sectors and is essential to all
aspects of production as well as day-to-day
life. Embedded technology has become
more important in recent years. The word
"embedded system" may apply to a wide Fig. 3 Multitask real-time processing
range of software programs that are platform of embedded system
connected to computers. For specialized
computer systems that have high The embedded system is used most often
requirements for performance, reliability, in applications involving artificial
cost, volume, and power consumption, it is intelligence. The artificial intelligence data
desirable to use hardware and software for the core module of the whole system is
that can be customized. The embedded received by the embedded system, where it
system is very relevant and is inextricably is analyzed and processed before being
bound up with a number of different sent on to be processed further in line with
applications. It is possible to alter and the rules that have been defined. Before
personalize it to satisfy the specific needs sending the digital signals to the core
of each application and to concentrate key module for processing, the digital module
efforts on those applications that have the in this system converts the data from the
potential to advance the development of artificial intelligence into a signal format
embedded systems more effectively. The that the embedded system is able to
faultlessness of the function may be process. The optical module is the initial
directly attributed to the incorporation of component that finishes off the collecting
an MCU and software that is embedded. of data for artificial intelligence. After
The intelligence of embedded devices is being retrieved from the system
unaffected by the absence of standard positioning module by means of an
software for embedded systems. It's also artificial intelligence processing algorithm,
possible that it was written in a language the necessary artificial intelligence
that's quite popular, like VC++. characteristics are then subjected to
Application software, which serves as the processing and analysis before being
basis of embedded systems, is output and regulated in accordance with a
characterized by stringent criteria specified target value. When the
pertaining to reliability, quality, and acquisition time is set to A, the
storage capacity. The implementation of information collection may be accurately
multifunctional real-time presents defined by the solutions of linear equations;
challenges with more stringent criteria for this is possible because of the linearity of
real-time. This serves as the cornerstone the equations.

Volume VII Issue I June 2022 www.zkginternational.com 205

 ISSN: 2366-1313

operating system in order to cut down on

the amount of resources that are used by a
 variety of distinct systems. This operating
system ensures both the consistency of the
In the formula, the start time of the running time as well as the performance in
information collection is represented by real time of any sort of task that may be
the letters A1 and A2, while the start time required.
values are represented by the letters B1
and B2. When describing the current state  Optimization of embedded
of information collection, the software system operation process based
architecture of the mobile network on artificial intelligence
information intelligent acquisition system algorithm
based on Embedded MCU is characterized
by the data obtained from a linear equation. The use of intelligent sensor technology
This is performed by utilizing the data. and the usage of embedded technology
The output of the model for making are inextricably linked and cannot be
predictions, denoted by ym(k), is separated. Embedded technology
composed of two parts. The answer that provides the capacity for a standard
was given as input was 0, and one of the sensor to do calculations, connect with
parts was the free response to Yi (k). The other devices, and make choices based
ability to govern the moment before this on the collected data. The embedded
one has absolutely nothing to do with this software is able to interact with one
one, and neither does the ability to control another and can even establish a
the moment after this one or the moment connection to the Internet. Intelligent
after that one. The second kind of answer sensors equipped with intelligent
is called the forced response, and it is the algorithms and intelligent algorithms
model response that is introduced to the
system after the control function has been
played. The forced response is represented
by yf(k), which is the same as the zero
state response. The newly added control
function in predictive functional control
may be represented as a linear
combination of a number of existing
functions (fn) in certain implementations
of the technique. As a direct result of this,
the predictive functional control approach
is enhanced in the way that will be detailed
further below.Embedded application Fig.4 Information collection flow of
software is not the same as traditional embedded system
application software. It does this by
integrated into embedded systems will
allocating certain hardware resources in
be a part of the Internet of Things. As a
line with the user's expectations and by
direct consequence of this, embedded
selecting a particular hardware platform in
technology is an essential component of
accordance with the unique application
the Internet of Things. The goal of
circumstances. When developing
software and hardware collaborative
embedded application software, we often
synthesis is to maximize the potential of
make use of a trustworthy embedded
a target by using the model's function

Volume VII Issue I June 2022 www.zkginternational.com 206

 ISSN: 2366-1313

application and constraint conditions as system's computing and communication

inputs. After that, it decides the system resources. You have the option of selecting
calculation, communication resources these resources from the permanent
and job allocation, map drawing, platforms that are already in place, or you
processing, voltage adjustment, and so may include them into the design of the
on, and ultimately arrives at the system, which is comprised of a number of
software and hardware structure of the components such as a processor, memory,
system that is optimal. It is essential to communication unit, and specialized
do research on the approach for a hardware accelerator. This is the typical
variety of computer architectures, such design that the optimization of the
as single processor, dual processor, operating process for an embedded system
multi-core, and multi-core systems. The looks like.
number of applications that make use of
embedded systems is now developing at In order to successfully execute the
a rapid rate, and concurrently, there is automated design of embedded systems, a
an increasing need for the creation of powerful optimization algorithm has to be
new embedded systems. The vast created in line with the specific stages of
majority of today's embedded devices system synthesis and the upper limit of
are not up to the task of meeting the design parameters. Assignment and
needs of actual applications. mapping are two types of subproblems that
are notoriously difficult to solve when they
The embedded platform provides the basis occur together. If jobs are considered to be
for embedded artificial intelligence independent constants, then the intervals at
systems. This platform is comprised of which they occur and the life cycles that
graphic gathering equipment, artificial they follow will compete with one another.
intelligence processing equipment, The approach that is used to calculate task
artificial intelligence display equipment, priority is as follows. When designing
and artificial intelligence processing automated embedded systems, certain
equipment. To ensure that embedded synthesis techniques and design parameter
systems perform their functions as limits need to be taken into consideration.
efficiently as possible, the method of Additionally, an efficient approach for
information collection used by these optimization has to be developed. In
systems has been enhanced by using the general, combinatorial optimization issues
concepts of artificial intelligence. Because are instances of subproblems that have a
the system level synthesis of embedded large range of potential solutions and
systems needs the co-design of both numerous problems to solve. Some
software and hardware, this research refers examples of these problems are
to it as hardware/Software. The concepts assignment, mapping, and processing.
of co-synthesis and system integration are Many different approaches to optimization
synonymous with one another. In the have been investigated and put into
beginning, the function, capabilities, and practice. The cycle value for the job is a,
constraints of the embedded system are the number of tasks that need to be
discussed. Following that, the system is completed is n, and the priority of the
specified via the use of task diagrams, work is r. The task is then processed via a
formal language, and natural language. mix of processing management tools and
The process of selecting or allocating the artificial intelligence based on this
architecture of the system is concurrent information. If tasks are independent
with the selection or allocation of the constants, then they will not interact with

Volume VII Issue I June 2022 www.zkginternational.com 207

 ISSN: 2366-1313

one another over the whole of their

respective lifecycles and time periods.
After some time, the task prioritization
algorithm was eventually solved.

Using the approach described before, the

idle time di(t) of task margin is then
subjected to further normalization. If Ei (t)
indicates the amount of time the CPU was
used to complete the work and I represents
the amount of time left before the task is
due to the system, then the technique for
sorting tasks based on time point may be
achieved as follows.

The idea of artificial intelligence is used to

standardize the operational parameters of
embedded systems. This is done in order
to establish whether or not the system's
given responsibilities can be managed
centrally. It is required to make certain that
the scheduling limit of tasks is
proportional to the rate of resource usage
in order to accomplish the monotonous
processing of the system operating ratio.
This is necessary in order to ensure that
the artificial intelligence dynamic task
priority change algorithm is followed.
According to the research that came before
this one, one of the advantages of using
artificial intelligence algorithms is that
they allow for a high system operation
utilization rate, which in turn permits the
implementation of a monotonous work
processing ratio. The operational
consequences of the step system

Fig. 5 Optimization of embedded system

operation process

3. ANALYSIS OF
EXPERIMENTAL RESULTS

The following conditions constitute the

test environment's parameters: The

Volume VII Issue I June 2022 www.zkginternational.com 208

 ISSN: 2366-1313

processing delay performance was Fig. 6 comparison test results

measured after the real-time operating It can be seen from the figure that the
system terminal PC had the virtual system can simulate the running
machine ar ware and the kuntu2.0 environment of embedded application
operating platform added to it. The software and realize dynamic
processing delay performance was identification of software and hardware by
determined by setting 30 processing tasks combining with fault injection technology.
at random. The processing algorithm It has better performance and has more
determines the delay allocation function by advantages than traditional test behavior
using the embedded notion as a starting and function. This design can simulate the
point. The next step is to install a web-core running environment of embedded system
2.6 CPU operating at 1.6 GHz together and promote the operation of embedded
with 21 GB of system memory. The application software. Compared with the
Windows 8 development platform and traditional test behavior and function, this
Windows Server 2007 are the two options design has higher software test
that users will choose. The hardware of the performance and better guarantees the
sensor is kept simple by using a single- running effect of the system.
chip microcontroller with six bits of
memory and 256 bits of programmable 4. CONCLUSION
memory. The use of 10 KB of RAM in This picture demonstrates how, when
conjunction with a 12 bit digital signal paired with fault injection technology, the
converter results in an increase in the rate system has the potential to perform
at which information is sent over mobile dynamic identification of software and
networks. When the mcs1210 is utilized as hardware by simulating the operational
the principal chip for the embedded MCU, environment of embedded application
the number of external receiving software. It performs better than the
components that can be employed to conventional test behavior and function,
construct a multi-node sensing network is and it gives a greater number of
restricted to a very small number. The advantages. This architecture may be able
figure demonstrates the correctness of the to simulate the operating environment of
embedded system in contrast to both the the embedded system and provide support
traditional system and the system by for the capabilities of embedded
showing the results of regression testing. application software. In comparison to
more traditional approaches to testing
behavior and function, the running impact
guarantees and overall software test
performance offered by this architecture
are much better.

REFERENCES:

1) Xu Z, Yujie W, Chang L and

Zonghai C 2018 A novel approach
of battery pack state of health
estimation using artificial
intelligence optimization algorithm
J. Power Sources 376 191
2) KuoChing Y, LungWen T, KoHan

Volume VII Issue I June 2022 www.zkginternational.com 209

 ISSN: 2366-1313

L, YiWei C, ShouChieh H, 9) Shihong Y, Zhigao Z, Tao W and

YuSheng L and RayJade C 2020 Qingfeng G 2018 An efficient joint
The development an artificial compression and sparsity
intelligence algorithm for early estimation matching pursuit
sepsis diagnosis in the intensive algorithm for artificial intelligence
care unit Int. J. Med. Inform 141 application Futur. Gener.Comp.
104176 Syst.86 603
3) Wang Z and Fang B 2019 10) Zheqi Y,Amir M A, Adnan Z, Hadi
Correction to: Application of H, Qammer H A 2020 An
combined kernel function artificial overview of neuromorphic
intelligence algorithm in mobile computing for artificial intelligence
communication network security enabled hardware-based hopfield
authentication mechanismJ. neural network IEEE Access 8
Supercomputing 75 5965 67085
4) Pandarakone S E, Mizuno Y and 11) Billah M , Rashid M and Bairagi A
Nakamura H 2019 A comparative K 2020 Embedded System Based
study between machine learning on Obstacle Detector Sensor to
algorithm and artificial intelligence Prevent Road Accident by Lane
neural network in detecting minor Detection and Controlling Int. J.
bearing fault of induction motors Intell. Transp.Syst. Res.18 331
Energies 12 2105 12) R. F. R. Suélia de Siqueira, I. João
5) Xuejing Z,Chen W and Jinxia Yoshiyuki and G R Oscar
S,Jianzhou W 2019 Research and Fernando 2019 Embedded system
application based on the swarm for lighting control of LED light
intelligence algorithm and artificial source applied to therapy of
intelligence for wind farm decision radiofrequency ablation in
system Renew.Energy 134 681 hepatocellular carcinoma IEEE
6) Zhenjie F, Qiang H, Yonglei Z, Latin Am. Trans.17 1671
Wei R, JunYi G, Tao L, Cheng C, 13) Jiangchao L,Liqian C,Xavier R
Wencong L, Shixun C, Jincang Z 2018 Automatic verification of
and Tongyi Z 2019 Method of embedded system code
artificial intelligence algorithm to manipulating dynamic structures
improve the automation level of stored in contiguous regions IEEE
Rietveld refinement Comput. Trans. Comput-Aided Des. Integr.
Mater. Sci.156 310 Circuits Syst. 37 2311
7) M. Mourad M, Hussein A. K and 14) HsuChih H and ShaoKang L 2019
Aziza I. H 2019 Artificial A Hybrid Metaheuristic Embedded
intelligence based cooperative System for Intelligent Vehicles
spectrum sensing algorithm for Using Hypermutated Firefly
cognitive radio networks Proc. Algorithm Optimized Radial Basis
Comput. Sci.163 19 Function Neural NetworkJ. IEEE
8) Bhamidipati S, Kim Kyeong J, Sun Trans. Ind. Inform.15 1062
H, Orlik P 2019 Artificial- 15) ChangHo C, DongHyeon K,
intelligence-based distributed Beomguk P, Yeoseon C, Wonyong
belief propagation and recurrent C and DongPyo K 2018 Air-water
neural network algorithm for wide- interfacial fluidic sonolysis in
area monitoring systems IEEE superhydrophobic silicon-
Netw.34 64 nanowire-embedded system for fast

Volume VII Issue I June 2022 www.zkginternational.com 210

 ISSN: 2366-1313

water treatment J. Chem. Eng 358

1594

Volume VII Issue I June 2022 www.zkginternational.com 211