RESEARCH ACTIVITY

We carry out our research work in the LTI (Information Processing Laboratory) laboratories of the ESP of the UCAD, LIRT (laboratory of Computing, Networks and Telecommunications) of the ESP of the UCAD and the UNESCO Chair for Global Smart Disruptive Learning (GSDL) at TELUQ | University of Quebec in Montreal. This reflects the act of collaboration in our research activities.

We conduct our research in the field of Mobile Laboratories, Geodistributed Computing or Fog Computing and Wireless Infrastructures.

For Mobile Laboratories, our research is oriented around new challenges of the cohabitation of different connected objects, existing network equipment and heterogeneous, radically different wireless access networks, for the needs of the mobile user and the understanding of its environment. In mobility, the Mobile Laboratories are a set of mobile solutions or communicating mobile platforms / devices that can allow surveillance, monitoring, telemedicine, remote control, remote control, learning of one's environment,….

For Geodistributed Computing or Fog Computing and Wireless Infrastructures-Technologies, the challenges of communication and security in the Internet of Vehicles (IoV, Internet of Vehicle) and those of SDN network architectures (Software- Defined Network) to provide flexibility in the management of vehicle networks.

In addition, the advent of the Internet of Things (IoT) and mobility have precipitated a new mode of interaction redefining the role of human presence. This new role is no longer primarily to act and interact directly with the environment, but rather to supervise its automated unfolding.

However, the proposal of new models of Next Generation Mobile Laboratories is necessary. A New Generation Mobile Laboratory is then characterized by a massive use of intelligent objects, intelligent equipment and the omnipresence of heterogeneous wireless networks, thus defining new services and value-added services for transport, the city, industry, aviation, academia, education, research, military and governance of public / private institutions. 

The main objective of our research is to participate in the current scientific debate relating to IoT and IoV by proposing a model of Network of Intelligent Vehicle Sensors to facilitate the implementation of intelligence functions in modern vehicles and laboratories. new generation mobiles.

The specific research objectives are to propose Cloud infrastructure models, SDN-based network architecture models, lightweight security systems / schemes for the global support of new services and value-added services to users, to companies, universities, industries,…. Also, these new services will allow users to be entrepreneurial as independent digital service providers benefiting from the digital economy.

Motivation

         Digital and their permanent innovation arouse curiosity and arouse new services, value-added services among professionals, academics, military, individuals,…. Increasingly, mobility is more than a reality and we are witnessing a deployment of ubiquitous wireless technologies enabling delocalized collaboration.

However, a mobile device is often multi-interface and a modern vehicle is incorporated with several sensors, control / command systems and other interfaces…. With 5G technology and the Internet of Things (IoT), the Internet of Vehicles (IoV) is becoming a reality.

So, modern vehicles could be transformed into communicating mobile devices that can help with surveillance, monitoring, telemedicine, remote control, remote control, environmental learning,….

At the same time, public / private, industrial, university, government, family business structures, etc. are undergoing digital development. New types of wireless devices / equipment must necessarily be deployed in cohabitation with those already existing in the network to supervise and control ecosystems. To easily benefit from all the possibilities offered by digital platforms / systems, heterogeneous and ubiquitous wireless technologies, we have the right to believe that mobile solutions / devices should allow us to conduct and pursue activities of laboratory where there is no nearby service platform. In mobility, called new generation “Mobile Laboratories”, these “communicating” mobile devices must be designed and deployed to meet the needs of mobile users as closely as possible.

Results in the literature

To manage mobility and multihoming, most of the approaches proposed in the literature are local solutions, and only act on a part of the improvements required on the current architecture of the Internet (IP Protocol). For example, Mobile IP (MIP) mainly focuses on mobility. Other solutions also exist, such as: i3 (Internet Indirection Infrastructure), mSCTP and Shim6 among others. Despite all this, the problems of the architecture of the Internet still remain without an effective solution for mobility because the IP protocol is degraded by nature by its dual role of identifier and locator. This concern about finding an effective solution has been the subject of several debates within the IETF, thus the NSRG (Namespace Research Group) was activated between 1999 and 2003. A report was produced which clearly stipulates the use of another namespace other than 32-bit IPv4 addresses. With the gradual deployment of IPv6, interoperability between IPv4 and IPv6 remains another unresolved issue by the TCP / IP model.

               In addition, with Digital, several concepts related to vehicular communication have been developed and research in this field remains very active. The first applications of vehicle networks are mainly related to safety and security. The main objective being to reduce road accidents through a system for exchanging and disseminating warning messages when a dangerous event occurs. Architectures, protocols and their implementations are traditionally based on VANET network architectures and are carried out through projects supported mainly by governments and international institutions (United States, Japan, European Union, etc.), research organizations and consortia (ETSI, IEEE, ISO, SAE, ASTM…).

However, several characteristics related to the vehicle, i.e. its mobility / communication model, its perception methods and the types of services / applications used make the concept of the Internet of Vehicles (IoV) different from that of the Internet of Things. For some time now, new applications have started to settle in the application landscape of vehicular networks. These applications, grouped around the Internet of Vehicles, have above all a commercial interest and arouse great economic attractiveness among private operators.

In the literature, several Cloud models for IoV are proposed. Depending on these models, several services can be deployed. The ubiquitous view of IoV means that a communication model with local reach may not be suitable for all services / applications. In addition, a communication model with a global reach is sometimes unnecessary, expensive, not optimal due to the local interest / relevance of a service. A compromise would then be to use a model where the local or global scope of each piece of information is determined before its processing and dissemination. Communication in the IoV takes place in peer-to-peer or client-server mode relying on a complex and heterogeneous conglomerate of wireless entities.

Designated under different names, three (03) Cloud architectures are mainly used including Cloud-based Architecture / Vehicles using Clouds / Vehicles to Clouds, Vehicle-based architecture / Vehicular Clouds / Vehicles as Clouds and Hybrid architecture / Hybrid Vehicular Clouds / Vehicles with Cloud .

Following the vision of the IoT, we realize that the sensors will have to be placed everywhere. Indeed without sensors, there is no IoT. The "S2aaS" (Sensing as a Service) model is presented today as a solution for transforming raw data from sensors into "information" and "knowledge" which cannot be acquired by traditional methods. S2aaS takes advantage of vehicle mobility to follow the 4A (Anywhere, Anytime, by Anyone and Anything) vision. The integration of S2aaS into the IoV is therefore of unprecedented economic interest.

In the literature, most of the work relating to S2aaS attacks the challenge of the taxonomy of existing solutions. An S2aaS model is proposed to use the various sensors incorporated in Smartphones providing a collection service managed by several Cloud servers. The S2aaS model authors emphasize three points: the ability to support multiple mobile platforms, the ability to encourage individuals to share their data, and energy saving. The existing S2aaS models follow three categories: the “Search” service provider, the “GPS-less Sensing / Scheduling” service and the “Recruitment” service provider. Another S2aaS model is split into four layers: sensors and sensors owners; sensors publishers (SPs); extended service providers (ESPs); and sensors data consumers. Some examples of the use of these models are presented including garbage management, smart agriculture and environmental management. The "CASSARAM" model takes care of the research, selection and classification of sensors. Context sensitivity is also used by combining quantitative reasoning techniques and semantic research to improve the performance of the associated system. A virtual sensor management framework based on the e-SOA (event-driven Service Oriented Architecture) paradigm is also proposed in the literature. This Framework makes it easier to query connected objects and associated services; likewise a semantic access control mechanism is used. For a network of virtual sensors, two (02) algorithms are proposed, in particular RADV and RADE. The RADV algorithm makes it possible to virtualize sensors from the Cloud of Objects. RADE is a distributed algorithm for estimating a set of unknown parameters.

SCIENTIFIC PRODUCTION

Our research resulted in fifteen (15) publications. These publications are published in IEEE Xplore, IGI Global and iCEER (INEER), and are indexed in the international databases Elsevier SCOPUS, DBLP, INSPEC-journalsList, IET-Journals and Springer.

I have participated in several international conferences and am Reviewer in several journals and program committees.

The scientific output is listed below:

1. A new vehicle-based vehicle sensor communication model seen as an intelligent digital platform.

This new model of an "Intelligent Vehicle Sensor Network" or IVSN is based on both spatial sensitivity, data sensitivity, group sensitivity and context sensitivity. In this model, the vehicle has a “self-awareness” because it is equipped with processing and reasoning capacities on the data coming from its incorporated sensors. It is then sensitive to the contexts of its contextual neighborhood.

Then, from this IVSN model, a new model of Cloud services for IoV, named S2aaS (Sensing as a Service), is proposed.

This new S2aaS model is composed of lower layers constituting the Local Cloud (sensors embedded in vehicles, on-board computer (OBU), roadside infrastructure (RSU, Road Side Unit)) and upper layers constituting the Conventional Cloud ( Middleware, Application servers). Middleware is the heart of S2aaS and is responsible for transforming information into knowledge and then into understanding for the user.

2. A new VANET (Vehicular Ad Hoc Network) network architecture based on SDN (Software-Defined Network) technologies

This new architecture is proposed for the communication of intelligent vehicles. In validation, this SDN architecture for vehicular communication is implemented as a prototype consisting of SDN Switches, SDN WiFi access points and SDN controllers. The SDN Backbone ensures the transport of control planes relating to mobility and routing schemes to dedicated SDN controllers. In this new SDN architecture proposed for vehicular communication, network intelligence is centralized in the mobility controller called SDNVANET which exclusively manages the VANET control plane (V2V and V2I communication, Vehicle-to-Infrastructure). This controller manages the status and topology of vehicles for packet routing.

3. A new AEAD (Authenticated Encryption with Associated Data) scheme based on the MQTT (Message Queue Telemetry Transport-Sensor Network) protocol.

In the SDN architecture proposed for vehicular communication, the network access level consists of WiFi-SDN access points. We are therefore looking for a lightweight messaging protocol to communicate the sensors incorporated into vehicles in an environment where disconnections / reconnections are frequent. In the IoT, sensor disconnections are not uncommon, but it is still necessary to be able to retransmit the information when reconnecting. Thus, based on the MQTT-SN protocol dedicated to constrained objects, the AEAD scheme uses the ChaCha20 to encrypt the data in MQTT messages, and the Poly1305 to authenticate the sensors incorporated in the vehicles. The proposed security scheme provides end-to-end encryption. Chacha20 and Poly1305 are lightweight cryptographic algorithms whose software implementations perform particularly well compared to schemes based on the AES algorithm.

4. Thesis:

In our unique thesis work, our major contribution is named “Wireless gateways / interfaces hypervisor”. The latter maintains communications of mobile laboratory activities and embedded components during travel (crossing different heterogeneous wireless networks). The hypervisor selects the most favorable wireless access network based on contextual criteria (reliability, cost, latency, etc.) and best adapts the flow or distribution of component communications between the active wireless gateways / interfaces .

In the very near future, the proposed Next Generation Mobile Laboratory Model is expected to become very common, for example, to have sensor networks in cars, buses, trains, airplanes to allow access to services when the vehicle is moving. This hypervisor then acts as a mobile router, i.e. equipment that interconnects the mobile laboratory with access networks via multiple integrated wireless gateways / interfaces.

One of the results of the thesis is the partial implementation of a new micro-mobility, macro-mobility and multihoming scheme based on the HIP protocol (Host identity Protocol). Implementing HIP therefore means modifying an OS (Linux, Windows, etc.) of terminal equipment without touching, for example, its intermediate equipment, in particular routers (Cisco or D-Link equipment).

We have more than nine (09) years of experience as a teacher-researcher in the field of research in networks and telecommunications.