Available online at www.sciencedirect.

com
Available online at www.sciencedirect.com
ScienceDirect
ScienceDirect
Available online at www.sciencedirect.com
Procedia Computer Science 00 (2022) 000–000
Procedia Computer Science 00 (2022) 000–000 www.elsevier.com/locate/procedia
ScienceDirect www.elsevier.com/locate/procedia

Procedia Computer Science 207 (2022) 4123–4131

26th International Conference on Knowledge-Based and Intelligent Information & Engineering

26th International Conference on Knowledge-Based and Intelligent Information & Engineering
Systems (KES 2022)
Systems (KES 2022)
Data Architecture and Big Data Analytics in Smart Cities
Data Architecture and Big Data Analytics in Smart Cities
El Mehdi Ouafiq1*
1*
, Mourad Raif11, Abdellah Chehri22, Rachid Saadane11
El Mehdi Ouafiq , Mourad Raif , Abdellah Chehri , Rachid Saadane
1
SIRC-LaGeS, Hassania School of Public Works, Casablanca, Morocco
2
Department
1 of Applied Sciences,
SIRC-LaGeS, HassaniaUniversity
School of of Quebec,
Public Chicoutimi,
Works, QC,Morocco
Casablanca, Canada, G7H 2B1
[email protected];
2 [email protected];
Department of Applied Sciences, [email protected];
University of Quebec, Chicoutimi,[email protected]
QC, Canada, G7H 2B1
[email protected]; [email protected]; [email protected]; [email protected]

Abstract
Abstract
The smart city has become a persistent need and is no longer just a concept. The concept of smart cities heavily relies on collecting
enormous
The smart amounts of data.aThis
city has become paper proposes
persistent need and aisdata-management-based
no longer just a concept.solution for smart
The concept city,cities
of smart which is labeled
heavily Smart
relies Systems
on collecting
Oriented
enormousBig Data Architecture.
amounts of data. This Big
paperdata technologies
proposes have become essential
a data-management-based to the functioning
solution of cities.
for smart city, which The architecture
is labeled Smart includes
Systems
complex
Oriented components to be implemented
Big Data Architecture. Big databased on the architectural
technologies have becomerequirements. A data
essential to the migration
functioning of strategy was
cities. The proposed toincludes
architecture handle
the various
complex data sources
components to besuch as IoT devices,
implemented video
based on cameras, andrequirements.
the architectural drones. The A proposed approach
data migration also takes
strategy into account
was proposed data
to handle
the various and
processing datadata
sources suchThe
storage. as technical
IoT devices, video cameras,
constraints related toand drones.
data The proposed
processing approach
in a big-data also takes
environment areinto
alsoaccount
studied. data
We
also consider
processing anddata
datamodeling
storage. from a businessconstraints
The technical intelligence pointtoofdata
related viewprocessing
and a datainscience perspective.
a big-data environmentOur main goalstudied.
are also is to favor
We
the
alsofacilitation of the
consider data daily lifefrom
modeling practices in theintelligence
a business context of a point
smartof city by providing
view and a datathe city administrators
science perspective. Ourwithmain
a solution
goal isthat helps
to favor
them maintain of
the facilitation their
thecity smartly
daily and effectively.
life practices in the context of a smart city by providing the city administrators with a solution that helps
them maintain their city smartly and effectively.
© 2022 The Authors. Published by Elsevier B.V.
© 2022 The Authors. Published by ELSEVIER B.V.
This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0)
This
© is an
2022 open
The accessPublished
Authors. article underbyofthe CC BY-NC-ND
ELSEVIER license (https://creativecommons.org/licenses/by-nc-nd/4.0)
B.V.committee
Peer-review under responsibility the scientific of the 26th International Conference on Knowledge-Based and
This is an
Peer-review open access
under article
responsibilityunder
of the CC BY-NC-ND
scientific
Intelligent Information & Engineering Systems (KES 2022) license
committee (https://creativecommons.org/licenses/by-nc-nd/4.0)
of KES International
Keywords:
Peer-reviewSmart
undercities, big data of
responsibility analytics, Drone,committee
the scientific IoT, Data of
Management.
KES International
Keywords: Smart cities, big data analytics, Drone, IoT, Data Management.

1. Introduction
1. Introduction
The concept of Smart Cities appeared in 1992. It is used to represent the cities of the future, whose objective is to
The concept
improve of Smart Cities
the functioning appeared
of these cities in 1992.
both It is used to of
economically, represent the cities
management, or of the future,
even whose objective
the ecological is toa
plan, from
improve the functioning of these cities both economically, of management, or even the ecological plan, from
sustainable development perspective. The purpose of Smart Cities is also to bequeath to future generations something a
sustainable development perspective. The purpose of Smart Cities is also to bequeath to future generations something

* Corresponding author.
*Email address: [email protected]
Corresponding author.
Email address: [email protected]
1877-0509 © 2022 The Authors. Published by ELSEVIER B.V.
This is an open
1877-0509 access
© 2022 Thearticle under
Authors. the CC BY-NC-ND
Published by ELSEVIER license
B.V.(https://creativecommons.org/licenses/by-nc-nd/4.0)
Peer-review under
This is an open responsibility
access of the CC
article under scientific committee
BY-NC-ND of KES
license International
(https://creativecommons.org/licenses/by-nc-nd/4.0)
Peer-review under responsibility of the scientific committee of KES International
1877-0509 © 2022 The Authors. Published by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0)
Peer-review under responsibility of the scientific committee of the 26th International Conference on Knowledge-Based and Intelligent
Information & Engineering Systems (KES 2022)
10.1016/j.procs.2022.09.475
4124 2 El MehdiE.Ouafiq et al. M.
M. Ouafiq, / Procedia Computer
Raif, A. Chehri, R. Science
Saadane207 (2022) 4123–4131
/ Procedia Computer Science 00 (2022) 000–000

sustainable to allow them to build their future on a good foundation. The intelligent technologies implemented in these
smart cities are aimed at residents, and the objective is to improve comfort, their standard of living, or even their safety.
As stated by [1] the issues addressed by sustainable urban development strategies are:
• The economic development of the city, including the creation of jobs.
• The creation of an environment conducive to the creation of new businesses.
• Improving the level of education in the city.
• Increasing the quality of life in the city with the improvement and maintenance of green spaces, the quality of
medical and social support services, and making life safer on the streets of the city.
• The modernization and expansion of the technical and IT infrastructure of the city.
• Increasing the leisure offer in the city and improving its tourist attractiveness.
As the foresight scenario above demonstrates, smart cities are brittle architectures. From technological, social and
governance points of view, they have multiple points of failure with cascading, systemic effects. Several definitions
of Smart Cities have been studied [2]-[3], to conclude that the axes around which Smart Cities are developing are (1)
industry, (2) academic, and (3) the governmental aspect.
The "Smart" aspect takes several forms in these areas. For the industry, it is a question of designing more innovative
products or services and integrated technology for technology. For the academic, it is about improving existing
technologies. Finally, for the governmental aspect, which may seem unclear, refers to the concept of smart, sustainable
growth to prevent urban sprawl, for example.
Six basic components for making a Smart City have been identified [1]:
• An intelligent governance system with transparent information exchange between residents, the city, municipal
services, and emergency services (police, firefighters).
• An intelligent economy, enabling an efficient flow of products, services, and knowledge at the city level and
between cities.
• Smart mobility is an interconnected, secure, and efficient system for the management of transport, logistics,
parking lots, and public transport.
• The intelligent environment is an innovative resource management system such as devices for energy storage,
reducing energy consumption, power supply management, smart lighting system, renewable energy development, or
waste management.
• Smart residents have access to education and training through modern telecommunication and information
technologies. In addition, the accompaniment of residents in terms of resource creativity of human potential
encourages the active participation of residents in the life of the city.
• An intelligent lifestyle, allowing to improve the quality of life, to develop better health services and infrastructures,
but also enlarge and diversify the culture and service offered.
In this research paper, we explore the different intelligent cities' challenges. Furthermore, we expose NiFi solution
for dataflow between various systems within a smart city. The NiFi was built to automate the flow of data and manage
the flow of information between systems. NiFi's fundamental design concepts closely relate to the main ideas of Flow-
Based Programming. These challenges can be resumed into two categories [1], as described in Table 1.

Table 1. Smart City Challenges – Global View.

Perspective Requirement
Business The cities supposed to be more productive and considers the logistic
facilities at the first level.
Public The Smart-City should remedy to the Global-Problems difficulties, e.g.
• Pandemic.
• Crime and Social disorder.
• Cyber Security Risks.
• Traffic and Transportation.
 El Mehdi Ouafiq et al. / Procedia Computer Science 207 (2022) 4123–4131 4125
E. M. Ouafiq, M. Raif, A. Chehri, R. Saadane / Procedia Computer Science 00 (2022) 000–000 3

2. Connectivity, Data Analytics in Smart Cities

Technological innovations facilitate the implementation of smart city infrastructure. The lifeblood of a smart city,
data and connectivity power enables everything from digital twins, fiber networks, and the Internet of Things (IoT),
4G, 5G, robotics, cloud, edge computing, and artificial intelligence (AI), machine learning [4]. The primary purpose
of a Smart City is to collect data from drones, video cameras, and IoT sensors, which the relevant departments of the
city will analyze to make "quick or preventive" decisions. Also, data analytics-based integrated command and control
centers can be planned mainly for the missions of Smart Cities. Predictive analytics can be developed to track clusters
zones, suspected cases, Ambulances, disinfection services, and quarantined people to provide the latest information
based on real-time data monitoring. In every pandemic period, such as Covid-19, the city can be mapped based on a
geospatial information system (GIS) from which the city administration can monitor the areas affected and create
buffer zones.
Privacy concerns and the integrity of personal data are just part of the debate over smart cities and a crucial part of
the intersection between technological vulnerabilities and human-centered and societal dynamics [5].
As described in Table 2, the heart of smart city planning is data collection and data analytics which can help us
accelerate the digital transformation to get value from our Smart-Cities. As a matter of fact, a smart city project stands
on:
1. Strategy;
2. Data collection;
3. Data regulation and data governance to implement compliance measures to consider the rights and the
privacy laws which might restrict certain data collection and the use of it, especially when tracking
people's behavior;
4. Data analytics which is the source of truth of the smart-city.

Table 2. Smart City Stakes.

Smart City Stakes Details

Building blocks - Drones;
- Telecommunications networks (4G, 5G);
- Smartphones and IoT sensors;
- Big data centers;
- AI-Computation to analyze the data and find a solution;
Areas - Streets;
- Traffic signals;
- Buildings; Airport
Data Collection - Data migration strategy from the building blocks;
- Data architecture;
- Technical components.
Data Analytics - More energy efficient transport: e.g., traffic management application when
Benefits an accident is happening;
- Upgraded and smart water supply and waste-management facilities: e.g.,
sewage leak alarms;
- Heating and lighting networks;
- Secure space: e.g., crime reduction applications;
- Flood control;
- Responsive and interactive city administration: e.g., E-government.

Smart Street Lights can provide multiple functions like surveillance cameras, weather data collection for further
visualization, management of traffic (also sending signals to facilitate driverless cars), providing chargers for e-
vehicles, converting 4G and 5G signals of telecommunication using Wi-Fi hotspots [6]. Figure I describe how many
Data Analytics based Smart Street lights can conduct Smart Systems.
4126 4 El MehdiE.Ouafiq et al.M.
M. Ouafiq, / Procedia Computer
Raif, A. Chehri, Science 207
R. Saadane (2022) 4123–4131
/ Procedia Computer Science 00 (2022) 000–000

Drones will play an essential role in urban environments, offering multiple advantages in various sectors such as
crisis management, medical transport, security, safety, and law enforcement [7]-[8]. For example, cities like Singapore
and Dubai have already implemented Intelligent Police Systems based on massive networks of video cameras and AI-
based solutions from a security perspective. As shown in Figure 1, in our research, from a data analytics perspective,
we considered that drones could complement these video-camera networks by providing mobile aerial vision and
increasing the responsiveness and efficiency of law enforcement agencies.

Figure 1. Smart Systems based on Data Analytics

From a Data Analytics perspective, smart cities contain three types of data-nature:
1. DGM data or data generated by machines is mainly collected from Drones, IoT, sensors, smartphones, and other
intelligent devices;
2. DHO or data from human origins is primarily generated from social media, smartphones, and other sources;
3. DMP or data mediated by the process mostly in administrations.
As described in Table III, the nature of data should be handled differently in terms of data acquisition, storage, and
technical components.

Table 3. Different nature of Smart City’s data.

Data Data Source Data Ingestion Data Storage Constraints Technical Components
Source Constraints
Nature
DGM Drones Handling micro- Satellite Images are stored as separate NiFi; Kafka
IoT Devices batches and Streams. files. Flume
Sensors Handling real-time Handling huge amounts of Small-Files Spark Streaming
data processing in Hadoop, which is made for storing
Video only massive files. AWS Kinesis Data
Cameras Streams
Many small files will make the
NameNode run out of memory. AWS Kinesis Data
Firehose
AWS Simple Queue
Service
DMP Relational Parallelizing batch A logical and physical Data Model . Sqoop; Spark
database data acquisition jobs should be defined to store the massive . Oozie;
management data in a way that can be processed
system efficiently. Data-Factory
(RDBMS) . AWS Database
Migration Service
DHO Social Media Handling data From security perspective this data . Spark; Python
 El Mehdi Ouafiq et al. / Procedia Computer Science 207 (2022) 4123–4131 4127
E. M. Ouafiq, M. Raif, A. Chehri, R. Saadane / Procedia Computer Science 00 (2022) 000–000 5

and quality should land on a permanent storage area

Manual Files before pushing into the Data Lake

3. Data Architecture

Under the MapReduce framework of the Hadoop Ecosystem, we can store and process the massive data coming
from different sources, guaranteeing more scalability and high availability [9]. Furthermore, as shown in Figure 2,
with the Hadoop architecture, the data can be distributed across multiple nodes on a single or multiple on-premise,
hybrid, and cloud clusters, based on horizontal scalability.
Data will not land permanently on a single server but across multiple ones, which leads to more flexibility in
managing the required resources by adding other nodes to the cluster instead of powering one existing node, which is
the case of vertical scalability.
In our research, we adapted the best of the existing techniques for Smart Cities and Smart Farming Analytics. As
described in Table IV, the proposed architecture, labeled IISOBA, has two vertical layers:
1. Batch-Layer: This is responsible for processing, transforming, and storing the smart city’s batch and micro-batch
data using Spark, Python, Kyling, HDFS, Hive, and Impala. And then building data workflows using Oozie and
airflow in for of actions built as a directed acyclic-graph;
2. Real-Time-Layer: Data pipelines will be built to consume, transform, and store streaming data that will be joined
directly with the Batch-Data on the Enrichment-Layer.

Table 4. IISFOBA Architecture

Vertical Horizontal Specificity Issue and Solution

Layers Layers
Batch Shared Area It is built on top of a network file system in the form
Layer of a group of folders structured based on the business
logic.
Landing-Area for the Ephemeral-Data (DHOs) and
the data from less-secure systems (especially DGMs
and DHOs) is also considered a gateway of the
Hadoop platform.

Raw Zone Where raw data (from DMPs and DGMs) will be The bzip2 compression algorithms can be used
stored directly from the sources, and DHOs data will to save more storage space which saved the
be pushed from the Shared Area into external tables storage space 70% times better than a raw text
based on an HDFS directory with metadata assigned file, with almost a compatible performance
to it in Hive Metastore to handle the schema on-read. capacity.

Structured Data will be cleaned, have a specific structure and

Layer data types, and be stored in the form of partitions in
Avro format. In this layer, Historical-Data will be
4128 El Mehdi Ouafiq et al. / Procedia Computer Science 207 (2022) 4123–4131
6 E. M. Ouafiq, M. Raif, A. Chehri, R. Saadane / Procedia Computer Science 00 (2022) 000–000

stored as per the data vault data model, which proved

capable of handling historical data

Trusted Zone The data will be stored based on the business logic
defined in a Logical-Data-Model in the form of a
Snowflake schema. This layer will be considered the
source of truth of the smart-city analytics and the
centralized Hadoop Data Warehouse of the Smart
City. Since only the needed columns for calculation
will be used on the analysis queries, the data files
should be stored in Parquet format

Real- Enrichment It is based on transforming data and enriching it with The New Hadoop Archive, which uses a hash
Time & Layer calculations of the Dashboard KPIs. In this layer, the table containing the information of indexes and
Batch- data will be stored in data marts and accessible via splits it across multiple index files, is highly
Layer data visualization tools. To enhance the NameNode recommended for this perspective because it
performance and the memory capacity, after outperformed the HAR by 85.47%
successful execution of the Enrichment-Layer
workflow, another workflow is supposed to be
executed to delete the raw zone files so that we can
lessen the number of metadata handled by the
NameNode.

Figure 2. Actionable Smart City based information and services.

There are multiple options to ingest Drones’ data based on the location, model, requirements, and networking. The
Message Queuing Telemetry Transport (MQTT) can be a better option when the Drone and Data Processing platform
exists on the same network / WiFi. The recommended data platform, in this case, should be built on top of a Cloud
MQTT server on which UAVs data. Since we include the Data Management aspects in our research, data security
 El Mehdi Ouafiq et al. / Procedia Computer Science 207 (2022) 4123–4131 4129
E. M. Ouafiq, M. Raif, A. Chehri, R. Saadane / Procedia Computer Science 00 (2022) 000–000 7

should be considered when sending the data to the protected internal Hadoop Cluster firewall. Thus, on the network
file system, which is the kernel of the shared area of the IISFOBA. A cron job can be used by a remote node with an
attached sense hat to send the messages of the MQTT to a broker of the cloud-hosted MQTT. As a matter of fact, they
can be accessed by offline download, UDP, TCP, and REST.
The primary differences are the scope of complexity, the rate of change necessary to adapt, and that, at scale, the
edge case becomes a common occurrence. NiFi is built to help tackle these modern dataflow challenges [10]-[12].
Then, as we suggested in Table 3, NiFi can be used to subscribe to the queue and collect the messages
asynchronously. Given the fact that the most flexible packaging file format should be a Json format. Table 5 represents
the Data Flow using NiFi [13].
Table 5. Data Flow using NiFi strategy.

Data Flow Phase Data Ingestion Steps

Drone Image’s Data Directory of Drone images that should be parsed
Data Flow 1. Bring data from the image directory using GetFile;
2. HDFS storage of raw images of Flow-file using PutHDFS;
3. Passing an API Key by adding an attribute using Update
Attribute;
4. Json conversion of the body using ReplaceText;
5. Calling the HTTP endpoint using InvokeHTTP;
6. Changing the name of the file ${filename:replace('jpg','json')}
using UpdateAttribute;
7. Json file storage in Hadoop using PutHDFS;
8. Collect the values of the attribute using ExtractImageMetadata;
9. Json metadata storage in Hadoop using PutHDFS;
10. Replacing images by Json = attributes conversion into a Json file
using AttributesToJson;
11. Changing the name of the file ${filename:replace('jpg','json')}
using UpdateAttribute;
12. Json file storage in Hadoop using PutHDFS;
13. Attributes collection from image by Apache Tika using
ExtraMediaMetadata;
14. Json file storage in Hadoop using PutHDFS;
15. Resizing the original image to a specific pixel image using
ResizeImage;
16. Resized image storage in HDFS using PutHDFS.
Hive DDL Metadata should be assigned in Hive Metastore to the data that has been
put into in HDFS:
Create External Table MONITORING(
[Columns] [Data Types], …
)
ROW FORMAT SERDE 'org.apache.hive.hcatalog.data.JsonSerDe'
LOCATION ‘[The directory of data in HDFS]’
Apache Phoenix Table for Create Table DRONES_INPUT (
Drone Data DATE_ID VARCHAR NOT NULL PRIMARY KEY,
[Columns] [Data Types], …);

UPSERT Query UPSERT INTO DRONES_INPUT (datekey, [VALUES]) values

('${'COLUMN_VALUE'}',...)
Displaying drone data Data can be displayed by a class Java:
package com.dataflowdeveloper;
public class DRONE_OUTPUT implements Serializable {}
Image Data Analytics Machine learning algorithms can then be used to some Data Analytics
4130 El Mehdi Ouafiq et al. / Procedia Computer Science 207 (2022) 4123–4131
8 E. M. Ouafiq, M. Raif, A. Chehri, R. Saadane / Procedia Computer Science 00 (2022) 000–000

e.g. classifying images.

4. IISFOBA Abstraction Layer

An abstraction layer can be built on top of the IISFOBA architecture to handle the schema-on-read problems on
the hive data warehouse and data quality simultaneously since we deal with a different source of data. Most of it comes
with other data structures.
As described in Figure 3, the kernel of the abstraction layer is a configuration table named ABSCT with parquet
file format since it is columnar based. Each column will be solicited for specific usage.
The ABSCT will be responsible for creating DDL (deep data locality) for each file in HDFS (Hadoop Distributed
File System) at the RAW-Zone’s level, based on the inferred schema from the file and then affect the Metadata to it
in Hive Metastore. In addition, the table will be connected to data quality configuration tables to launch technical
controls and business rules verifications before ingestion of data from the Raw-Zone to the structured layer.

Figure 3. The scalability in Hadoop Architecture

5. Conclusion

Smart cities can be defined as those that effectively integrate physical, digital, and human systems in urban
environments to deliver sustainable, prosperous, and inclusive outcomes for their citizens.
In this paper, we have designed what a smart city as a platform can look like, which combines drones, networks,
and data analytics performances to deliver digital and intelligent services to make life better in the smart city by
engaging stakeholders and citizens straightly.
At the same time, we defined the data migration strategy and the data architecture to process and ingest Smart
City’s data coming from different sources, especially from drones and IoT devices. The proposed architecture handles
the real-time process and should be combined with Batch processing workflows to build the pipelines where data
flows from the source to the smart city dashboards and intelligent machine algorithms. A NiFi was built to automate
the flow of data between systems. A core philosophy of NiFi has been that even at a very large scale, guaranteed
delivery is a must. It is also necessary that the architecture should be a built-in layer to handle data quality, data
compression, data security, and data modeling on top of Hadoop’s Data Lake. Therefore, from this perspective and in
our future research will be continued to optimize the setup of the abstraction layer of the proposed architucture.
E. M. Ouafiq, M. Raif, A. Chehri, R. Saadane / Procedia Computer Science 00 (2022) 000–000 9

El Mehdi Ouafiq et al. / Procedia Computer Science 207 (2022) 4123–4131 4131

References

[1] Barbara Kos. (2019). "Intelligent Transport Systems (ITS) in Smart City," Springer Proceedings in Business and Economics, in: Michał
Suchanek (ed.), Challenges of Urban Mobility, Transport Companies and Systems, pages 115-126, Springer.
[2] Mosannenzadeh, Farnaz & Vettorato, Daniele. (2014). "Defining Smart City, A Conceptual Framework Based on Keyword Analysis". TeMA
- Journal of Land Use, Mobility and Environment. 683-694. 10.6092/1970-9870/2523.
[3] Chehri, A., Sharma, T., Debaque, B., Duclos, N., Fortier, P. (2022). Transport Systems for Smarter Cities, a Practical Case Applied to Traffic
Management in the City of Montreal. In: Littlewood, J.R., Howlett, R.J., Jain, L.C. (eds) Sustainability in Energy and Buildings 2021 . Smart
Innovation, Systems and Technologies, vol 263. Springer, Singapore. https://doi.org/10.1007/978-981-16-6269-0_22
[4] Ahmed, Imran & Jeon, Gwanggil & Chehri, Abdellah & Hassan, Mohammad. (2021). Adapting Gaussian YOLOv3 with transfer learning for
overhead view human detection in smart cities and societies. Sustainable Cities and Society. 70. 102908. 10.1016/j.scs.2021.102908.
[5] Joe Burton, & Simona Soare. (2020). Smart Cities, Cyber Warfare and Social Disorder. Zenodo. https://doi.org/10.5281/zenodo.6143265
[6] A. Chehri and H. T. Mouftah, "New MMSE Downlink Channel Estimation for Sub-6 GHz Non-Line-of-Sight Backhaul," 2018 IEEE Globecom
Workshops (GC Wkshps), 2018, pp. 1-7, doi: 10.1109/GLOCOMW.2018.8644436.
[7] A. Chehri, G. Jeon, I. Fofana, A. Imran and R. Saadane, "Accelerating Power Grid Monitoring with Flying Robots and Artificial Intelligence,"
in IEEE Communications Standards Magazine, vol. 5, no. 4, pp. 48-54, December 2021, doi: 10.1109/MCOMSTD.0001.2000080.
[8] S. H. Alsamhi, O. Ma, M. S. Ansari and F. A. Almalki, "Survey on Collaborative Smart Drones and Internet of Things for Improving Smartness
of Smart Cities," in IEEE Access, vol. 7, pp. 128125-128152, 2019, doi: 10.1109/ACCESS.2019.2934998.
[9] Diaconita, V., Bologa, A. R., & Bologa, R. (2018). Hadoop Oriented Smart Cities Architecture. Sensors (Basel, Switzerland), 18(4), 1181.
https://doi.org/10.3390/s18041181
[10] H. Isah and F. Zulkernine, "A Scalable and Robust Framework for Data Stream Ingestion," 2018 IEEE International Conference on Big Data
(Big Data), 2018, pp. 2900-2905, doi: 10.1109/BigData.2018.8622360.
[11] Ouafiq, E.M.; Saadane, R.; Chehri, A. Data Management and Integration of Low Power Consumption Embedded Devices IoT for Transforming
Smart Agriculture into Actionable Knowledge. Agriculture 2022, 12, 329. https://doi.org/10.3390/agriculture12030329
[12] El Mehdi Ouafiq, Rachid Saadane, Abdellah Chehri, Seunggil Jeon, AI-based modeling and data-driven evaluation for smart farming-oriented
big data architecture using IoT with energy harvesting capabilities, Sustainable Energy Technologies and Assessments, Volume 52, Part A,
2022, 102093, ISSN 2213-1388, https://doi.org/10.1016/j.seta.2022.102093.
[13] A. Pandya et al., "Privacy preserving sentiment analysis on multiple edge data streams with Apache NiFi," 2019 European Intelligence and
Security Informatics Conference (EISIC), 2019, pp. 130-133, doi: 10.1109/EISIC49498.2019.9108851.