Citation: Akallouch, M.; Akallouch,
O.; Fardousse, K.; Bouhoute, A.;
Berrada, I. Prediction and Privacy
Scheme for Traffic Flow Estimation
on the Highway Road Network.
Information 2022, 13, 381. https://
doi.org/10.3390/info13080381
Academic Editors: Antonio Comi,
Jianbo Li and Junjie Pang
Received: 6 June 2022
Accepted: 27 July 2022
Published: 9 August 2022
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2022 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
information
Article
Prediction and Privacy Scheme for Traffic Flow Estimation on
the Highway Road Network
Mohammed Akallouch
1,
* , Oussama Akallouch
1
, Khalid Fardousse
1
, Afaf Bouhoute
1
and Ismail Berrada
2
1
Faculty of Sciences Dhar El Mahraz, Sidi Mohammed Ben Abdellah University, Fez 30050, Morocco
2
School of Computer Sciences, Mohammed VI Polytechnic University, Benguerir 43150, Morocco
* Correspondence: mohammed.akallouch@usmba.ac.ma
Abstract: Accurate and timely traffic information is a vital element in intelligent transportation
systems and urban management, which is vitally important for road users and government agencies.
However, existing traffic prediction approaches are primarily based on standard machine learning
which requires sharing direct raw information to the global server for model training. Further, user
information may contain sensitive personal information, and sharing of direct raw data may lead to
leakage of user private data and risks of exposure. In the face of the above challenges, in this work,
we introduce a new hybrid framework that leverages Federated Learning with Local Differential
Privacy to share model updates rather than directly sharing raw data among users. Our FL-LDP
approach is designed to coordinate users to train the model collaboratively without compromising
data privacy. We evaluate our scheme using a real-world public dataset and we implement different
deep neural networks. We perform a comprehensive evaluation of our approach with state-of-the-art
models. The prediction results of the experiment confirm that the proposed scheme is capable of
building performance accurate traffic predictions, improving privacy preservation, and preventing
data recovery attacks.
Keywords: traffic flow forecasting; federated learning; privacy-preserving
1. Introduction
Traffic flow forecasting has long been considered a core and crucial element of Intelli-
gent Transportation Systems. By providing an accurate and timely estimation of the traffic
states, traffic predictive systems bring increased flexibility and efficiency to transportation
systems. They strongly contribute to improving traffic information [1] and efficiency, lead-
ing consequently to smarter mobility. Accurate and timely traffic information is vitally
important for both personal travelers and government companies, as it has the potential
to provide real-time traffic reports to predict future states, which can help government
authorities to control traffic, to make better-informed decisions, and estimate the traffic
congestion system by predicting the upcoming traffic flow in a smart city environment [2].
Empowered by technological advances and the wide availability of traffic data, re-
search on Traffic Flow Forecasting has been continually advancing. In fact, TFP has always
been a challenging task due to the stochasticity of traffic and dynamic conditions such
as weather conditions, calendar (i.e., time of day, day of week), accidents, events, etc.
Considerable efforts have been made by researchers from different fields to tackle the TFP
problem, trying to provide accurate and more reliable predictions. From mathematical
and statistical modeling to more recent data-driven methods, different approaches have
been proposed. The data-driven approach formulates TFP as a time series forecasting
problem, which aims to predict the traffic states based on historical data, collected by using
multiple sensors (e.g., radars, cameras, mobile devices, etc.). More recent TFP approaches
rely on deep learning models (e.g., Recurrent Neural Networks, Convolutional Neural
Networks) to automatically learn the deep features of traffic data automatically and solve
the prediction problem.
Information 2022, 13, 381. https://doi.org/10.3390/info13080381 https://www.mdpi.com/journal/information