J.

Electrical Systems 20-10s (2024):2753-2756

1
Golubtsov Ivan Microprocessor Role in the Internet of
Sergeevich
Things (IoT) Domain: A Review

Abstract: - The microprocessor of any computing device should be right-provisioned as it acts as the brain of any embedded system and
the Internet of Things (IoT) is the new name of recent embedded systems. However, as the number of connected devices with the internet
is increased, fixed microarchitecture design is considered a drawback in the context of IoT. The aim of current paper is to illustrate the
previous research work related to the device microprocessor in the IoT domain. Additionally, the functions of the IoT applications are
classified and the IoT microarchitecture configurations are illustrated. Right-provisioned microprocessors ensure efficient IoT edge
computing and appropriate microarchitecture resources/configurations utilization.

Keywords: Manager; Motivation; Critical Thinking; Innovation Performance

1. INTRODUCTION:
An appropriate microarchitecture design is required in terms of area, speed and time constraints in order to
ensure optimal computational capabilities in the IoT domain. The 4th industrial revolution (industry 4.0) is
constructed by integrating the new digital technologies with the different industrial processes [1]. A
reconfigurable microarchitecture in terms of the pipelines stages is designed to ensure just-enough edge
computing at the IoT edge-level [2]. The microarchitecture characteristics of a wide range of state-of-the-art
embedded system microprocessor are analyzed and the microprocessor’s applicability to IoT computation is
evaluated using different evaluation metrics [10]. An increasing amount of research work concentrates on
understanding and detecting insights into various aspects of the IoT due to the expected growth of IoT
[3][4][11]. In the context of edge computing, the edge nodes’ processing capabilities and hardware components
must be considered [12]. Fog computing is proposed as a virtual platform that enables networking services,
compute and storage between edge nodes and cloud computing data centers [13]. By moving computations
closer to the edge nodes, both latency and bandwidth bottleneck are reduced using fog computing. If the edge
nodes are equipped with enough computing capabilities, then the data transmission will be minimized and thus
latency, power consumption and bandwidth bottleneck will be reduced. Edge mining has the potential to reduce
the transmitted data amount which will reduce power consumption and storage requirements [5]. Taxonomy for
a high level definition of IoT components with respect to hardware, middleware and presentation/data
visualization are presented [4]. Taxonomy to classify wireless sensor networks is presented; this classification is
according to network dynamics, data delivery models and different communication functions [6]. Based on the
characteristics of the data models, the algorithms are classified using high level visualization taxonomy [14].
The execution characteristics are determined by the application’s functions, so those functions reflect the
microprocessor requirements. The presented paper is organized as following: the IoT applications are classified
in section two, the IoT microarchitecture configurations are illustrated in section three, and, finally the paper is
concluded in section four.
2. FUNCTIONS CLASSIFICATIONS OF IOT APPLICATIONS:
Several application domains are offered computing potential by the IoT. The application domains include
healthcare, logistics and transportation, smart environments, social and personal domains [3]. An expansive
study of IoT use-cases is performed [10], and then by those use-cases, the application functions performed such
that the application classification provides a broad, tractable and high level description. Six application

1Robotics and Complex Automation, Bauman Moscow State Technical University, Moscow, Russia, 105005

Microprocessor, Internet of Things (IoT), Microarchitecture, Edge Computing, configuration.

[email protected]

Copyright © JES 2024 on-line : journal.esrgroups.org

2753
 J. Electrical Systems 20-10s (2024):2753-2756

functions form IoT application classification, they are: sensing, communication, image processing, compression,
security and fault tolerance [10]. Figure 1 illustrates the functions classifications of IoT applications and the
following subsections describe the application functions.

Figure 1: Functions Classifications of IoT Applications.

2.1. Sensing Function:
Sensing includes data collection about case or event. The data can be temperature, motion or pressure, etc.
sensing is common in several domains of the applications. Data/activities/information of interest is collected for
over and above processing and decision performing. For example: in sensor fusion [7], the sensed data using
different sensors are merged to produce a more robust and accurate quantitatively/qualitatively data. The
algorithms that used for sensor fusion can occupy different levels of memory/compute intensity.
2.2. Communications Function:
Considering the fact of the IoT’s intrinsic connected structure, communications is the most common IoT
application functions. Many nodes are connected together and the data travels through them. Wi-Fi and
Bluetooth are examples of the communication technologies. Transfer control protocol (TCP) and the emerging
6lowpan (IPv6) are examples of the communication protocols. Software defined ratio (SDR) is a communication
system where filters, modems and other physical layer functions that are presented in hardware are implemented
in software [15]. With SDR, both instruction memory and data footprints are small. However, its applications
are compute intensive.
2.3. Image processing Function:
Signal processing is the major function in image processing. Face recognition, traffic sign recognition and
automatic license plate recognition are IoT applications which use image processing. Several image processing
applications and use-cases are expected to get more significant in the coming future [8]. More computation
capabilities are required with image processing as it includes matrix multiplications. In addition to that, large
amount of data is to be stored either as input, intermediate or output, so this raise the memory size requirements.
2.4. Compression Function:
Compression ensures that the data is transmitted fastly, analyzed or/and retrieved by reducing the
communication requirements. The resource-constrained concept of the IoT devices can be achieved using
compression by making the required memory amount size small at the edge-level. Lossless and lossy are the
main two types of compression. With first type, i.e. lossless compression, in order to represent the data
concisely, the redundant data are deleted in a statistical way. With second type, i.e. lossy compression, the
perceptibility of data in question is exploited, but the unnecessary data is removed. Lossless compression type
needs more memory due to its computational requirements.

2754
 J. Electrical Systems 20-10s (2024):2753-2756

2.5. Security Function:

The device and the data integrity are maintained with the help of security applications which keep the sensitive
data save from unauthorized access. Malicious attacks affect the IoT devices which functioning in open
environments. Data encryption ensures data confidentiality [9]. The encrypted data is created using an
encryption algorithm which creates the need of decryption process. Encryption speed is related to the memory
access latency for storage and data retrieval, so this is again forms both memory and compute intensively.
2.6. Fault tolerance Function:
When the IoT devices are utilized in unattended environments, the system should be capable to function
properly if a failure of some their component accrues. This is referred to as fault tolerance and it is important for
quality of service (QoS) insurance. There are two types of fault tolerance and they are hardware-based and
software-based. First type i.e., hardware-based fault tolerance are a storage devices and use redundancy. Second
type includes algorithms and applications that make operations like error detection and correction, cyclic-
redundancy checks and memory scrubbing.
3. CONFIGURATIONS OF IOT MICROARCHITECTURE:
A wide and an extensive study is performed on the state-of-the-art in commercial-off-the-shelf (COTS)
architectures of the embedded systems microprocessor [10]. This study is from several manufactures and range
from low-end microcontrollers to high performance/high-end embedded systems microprocessors. The
microprocessors are categorized based on several characteristics. Those characteristics include execution type,
off-chip memory support, on-chip memory, number of pipeline stages, cores number and frequency. Table 1
summarizes the presented configurations. It is clear from table 1 that: there are four microarchitecture
configurations. First configuration represents low-performance and low-power microcontrollers units (MCUs)
like ARM cortex M4 [16] presented in many IoT-targeted MCUs from many developers. Those developers are
STMicroelectronics [17], Freescale Semiconductors [18] and Atmel [19]. Second configuration represents
recently-developed IoT-targeted CPUs like the Intel Quark Technology [20]. Third configuration represents
mid-range CPUs like the ARM cortex A7 found in several general purpose embedded systems. Finally, fourth
configuration represents high-performance/high-end embedded systems CPUs like the ARM cortex A15. The
main difference between the two execution types is that: the out-of-order execution type allows instructions to
execute as soon as the instruction becomes available, whereas with in-order execution type the instructions must
execute in program order.
Table 1: Configurations of the IoT Microarchitecture
First Conf. Second Conf. Third Conf. Fourth Conf.
Cortex M4 from Cortex A7 from Cortex A15 from
CPU family Intel Quark
ARM ARM ARM
Out-of-order
Execution Type In-order Execution In-order Execution In-order Execution
Execution
Off-chip memory 512 KB flash 2 GB RAM 2 GB RAM 1 TB RAM
32 KB 32 KB
Instructions/Data Instructions/Data
On-chip memory None None
Level 1, 1 MB Level 1, 2 MB
Level 2 Level 2
Number of
three five eight fifteen
pipeline stages
Cores number one one four four
CPU frequency 48MHZ 400MHZ 1GHZ 1.9GHZ

2755
 J. Electrical Systems 20-10s (2024):2753-2756

4. CONCLUSION:
Taking the important role of the device microprocessor in the IoT domain, the current paper reviews the
research work related to it in the mentioned domain. In addition to that, the IoT applications functions are
classified into sensing, communications, image processing, compression, security and fault tolerance,
respectively. The IoT microarchitecture configurations are summarised also. The microprocessor optimization
and computing paradigms for the IoT are to be served as an extended work of current paper. Table 1 shows all
comparisons made in this work between the studied IoT microprocessors. It is clear that the differences in
performances among all the different IoT microprocessors are related to the microarchitecture of each of the
microprocessors.
REFERENCES:
[1] W. Ghder Soliman, D. V. R. K. Reddy, Traditional versus fuzzy optimal coupling in multi-axis distributed motion
control: a pre- iot- integration procedure, Advanced Control for Applications 2 (4) (2020) e57.
arXiv:https://onlinelibrary.wiley.com/doi/pdf/10.1002/adc2.57, doi:https://doi.org/10.1002/adc2.57. URL
https://onlinelibrary.wiley.com/doi/abs/10.1002/adc2.57
[2] W. G. Soliman, B. K. Priya, D. A. Reddy, P. V. S. Anusha, D. V. R. K. Reddy, Reconfigurable microarchitecture-
based pmdc prototypedevelopment for iot edge computing utilization, IEEE Sensors Journal 21 (2) (2021) 2334–2345.
doi:10.1109/JSEN.2020.3020362.
[3] Atzori, L., Iera, A., and Morabito, G. “The internet of things: A survey”. Computer networks, 54(15). 2010, pp. 2787-
2805.
[4] Gubbi, J., Buyya, R., Marusic, S. and Palaniswami, M. “Internet of Things (IoT): A vision, architectural elements, and
future directions”. Future generation computer systems, 29(7). 2013, pp.1645-1660.
[5] Gaura, E.I., et all. “Edge mining the internet of things”. IEEE Sensors Journal, 13(10). October 2013, pp.3816-3825.
[6] Tilak, S., Abu-Ghazaleh, N.B. and Heinzelman, W. “A taxonomy of wireless micro-sensor network models”. ACM
Sig. Mob. Comput. & Comm. Review, 6(2). 2002, pp. 28-36.
[7] Nakamura, E.F., Loureiro, A.A. and Frery, A.C. “Information fusion for wireless sensor networks: Methods, models,
and classifications”. ACM Computing Surveys (CSUR), 39(3). 2007, pp.9-es.
[8] Akyildiz, I.F., Melodia, T. and Chowdhury, K.R. “A survey on wireless multimedia sensor networks”. Computer
networks, 51(4). December 2007, pp.921-960.
[9] Singh, S.P. and Maini, R. “Comparison of data encryption algorithms”. International Journal of Computer Science and
Communication, 2(1). January 2011, pp.125-127.
[10] Adegbija, T., Rogacs, A., Patel, C., and Gordon-Ross, A. “Enabling right-provisioned microprocessor architectures for
the internet of things”. In ASME Int. Mech. Eng. Cong. & Expo.. American Society of Mechanical Engineers. (57571).
November 2015, p. V014T06A001.
[11] Giusto, D., Iera, A., Morabito, G., and Atzori, L. (Eds.). “The internet of things”. 20th Tyrrhenian workshop on digital
communications. Springer Science & Business Media. 2010.
[12] Sánchez-Sinencio, E. “Smart nodes of internet of things (IoT) a hardware perspective view & implementation”. In 24th
edit. Proce. of the great lakes sympo. on VLSI. May 2014, pp. 137-138.
[13] Bonomi, F., Milito, R., Zhu, J. and Addepalli, S. “Fog computing and its role in the internet of things”. In the 1st edit.
Proce. of the MCC workshop on Mob. cloud comput. August 2012, pp. 13-16.
[14] Tory, M. and Moller, T. “Rethinking visualization: A high-level taxonomy”. In IEEE Symposium on information
visualization. IEEE. October 2004, pp. 151-158.
[15] Lee, H., Lin, Y., Harel, Y., Woh, M., Mahlke, S., Mudge, T. and Flautner, K. “Software defined radio–a high
performance embedded challenge”. In Int. Con. on High-Perf. Emb. Archit. & Compil. Springer, Berlin, Heidelberg.
November 2005, pp. 6-26.
[16] Cortex-M4 Processor – ARM. http://www.arm.com/products/processors/cortex-m/cortex-m4-processor.php.
[17] STMicroelectronics. www.st.com
[18] Freescale Semiconductors. www.freescale.com
[19] Atmel Corporation. www.atmel.com
[20] Intel Quark Technology. http://www.intel.com/content/www/us/en/processors/quark/intel-quark-technologies.html.

2756