Enforcing the forgetting (or unlearning) of specific portions of a neural network's (NN) memory is necessary for both security and ethical considerations. Moreover, targeted forgetting can enhance the network's performance by eliminating the learning derived from compromised data. Conventional approaches, such as retraining the model from scratch, are associated with significant time and energy costs, while alternative methods can degrade the neural network's performance in terms of complexity and predictive accuracy. To address these challenges, this work introduces two novel approaches that mitigate the drawbacks associated with traditional unlearning methods. Both approaches leverage the Long Short-Term Memory (LSTM) network, chosen for its superior learning quality and efficient processing time—crucial factors in such algorithms. The first approach involves modifying the LSTM architecture by coarse tune the weight matrix of the forget gate to facilitate controlled forgetting. The second approach follows the idea of intruders by inducing poluuted data to the system, and the idea is based on learning the model through corrupted/bad data, which causes the LSTM memory to be changed at specific time intervals that correspond to the data that should be forgotten. We present a performance evaluation of the second approach, with a future study planned for the performance assessment of the first approach. The evaluation demonstrates the superiority of the proposed solutions in terms of quality of service and implementation simplicity.

Balancing sustainable energy with cellular deployment can be challenging. However, multi-user MIMO antennas for dynamic positioning can benefit both objectives. This study presents an optimization algorithm leveraging dynamic drone deployment to alleviate cellular communication congestion in densely populated areas. Two deployment strategies are examined: prioritizing energy efficiency by minimizing antenna footprint, and optimizing service quality by placing serving offloading drones. A novel Voronoi planning tool models both approaches, aiding in estimating energy consumption across various scenarios, particularly in the context of modern multi-user massive MIMO and beamforming techniques. A comparative energy analysis for different techniques is conducted, along with establishing a theoretical minimal coverage antenna footprint.

The real implementation of a recurrent neural network (RNN) in a low complexity IoT device is evaluated in order to predict the time series of power consumption in tertiary buildings. The RNN type long short-term memory (LSTM) algorithm is adapted for a 32-bit microcontroller unit (MCU) and the backpropagation (BP) algorithm is implemented in-house. We therefore demonstrate that Intelligent IoT (IIoT) devices, such as the Espressif ESP32 MCU, not only implement neural networks (NNs), but also learn on their own. The resulting IIoT architecture has been proven to operate efficiently and compared to the traditional computer-based learning platform. The selected results confirm that stand-alone IoT devices are a truly efficient solution that adds flexibility to the architecture, reduces storage and computation costs, and is more energy-friendly. As a conclusion, it is practically more efficient to exploit low-power and processing-time IIoT for our prediction use case rather than relying on server based distributed systems.

Two models are proposed to optimize unmanned aerial vehicles (UAVs) traffic offloading in cellular networks. The first model optimizes the realtime traffic service, while delaying non-realtime traffic in the cell buffers (BS and the currently serving drone). Delayed traffic is then transmitted later when free resources are available and has a maximum service delay limit. The proposed model provides a heuristic solution to minimize the losses in non-realtime traffic based on the maximum delay and the size of the cell buffers. The second model completes the work by providing an optimal Integer Linear Programming solution to minimize the number of needed UAV assuming no data loss and buffering availability. It is also parameterized with the same delay limits as the first model. The performance of the proposed models is studied using the call detail record (CDR) dataset of the city of Milan cellular network provided by Telecom Italia. The two models showed great QoS performance based on the maximum delay of non-realtime traffic and cell buffer size compared to models that do not include the buffering and the delay limits.

Base station (BS) capacity for cellular networks is fixed and therefore cannot be easily adapted to support temporary high demands due to unusual traffic (demonstration, sales, competitions, etc.). Unmanned Ariel Vehicle (UAV)/drone can be used as an additional temporary bandwidth (BW) provider capable of covering multiple cells.Two placement approaches are investigated to optimize UAV deployment for this purpose. The first adopts a rule based approach without cooperation: we place UAVs in the cells where the exceeding demand is the highest, allowing each UAV to only serve a single cell at a time. In the second algorithm, relying on the Knapsack problem, we design a solution where UAVs cooperate by sharing a portion of their available BW with neighbouring cells in order to cover all the demand with the lowest possible amount of drones. Their performance is compared using two months of real dataset from the Milan Cellular Network provided by Telecom Italia.

Cellular deployment of new generations faces a coverage challenge due to the non-line-of-sight (NLOS) between clients' devices and base station (BS). Therefore, relaying on using the emerging technology; intelligent reflective surface (IRS) to reconfigure wireless signal propagation is considered the best solution that can address the mentioned challenge. Additionally, choosing the position of the IRS is not an easy task as the clients are mobile. Hence, there is a need for an efficient model to elect the best positions of the IRSs for a better network performance. In this work, two fold model is proposed to provide an automated solution to optimize IRS positions. The first one is the mixed integer linear programming (MILP) that solves the IRS positions problem in a classical way. Whereas the second one is based on the reinforcement learning optimization (RLO) with complex encoder and decoder network architecture to provide fast learning of the MILP results with a low mean square error. The proposed RLO model's validity is studied using 10 days of mobile dataset and actual cellular BSs' positions in the city of Rome (Italy). This study is based on the use of long short term memory (LSTM) and gated recurrent unit (GRU). The results show a significant performance of the proposed model based on LSTM compared to GRU.

In multimedia delivery, especially for mobile users, the Multi-Path Transmission Control Protocol (MPTCP) can ameliorate the sustainability of multimedia sessions continuity. MPTCP establishes a number of sub-flows equals to the number of the available wireless active connections so, the chance of service interruptions decreases. This paper investigates the performance evaluation of using MPTCP in heterogeneous wireless networks (HetNets). Through this evaluation, we propose two dimensional Markov model for MPTCP seamless handover between two wireless connections. Then, we implement the handover algorithm using MPTCP open source Linux Kernel. Moreover, an implementation for Intelligent Transportation System (ITS) platform is conducted in order to valorise the MPTCP aspect in such real time platforms. The overall results show that MPTCP outperforms classical methodologies of handover.

Multi-modal traffic management in Intelligent Transportation Systems (ITS) aims to provide a more efficient traffic regulation to passengers and reduce congestion and obstruction in the roads. In spite of the outstanding progress made in this research filed; traffic management still a very challenging problem regarding the multiple factors that have to be taken into account in any proposed solution. To tackle this problem, this paper introduces a collaborative model based context awareness multi-modal traffic management aiming at providing an efficient way to manage the traffic inside a transportation station. In this model (Multi-Layers Stations: the stations that have different intersections for different means of transport), the traffic management is based on a Q-learning technique that takes into account the context awareness parameters to provide more potent decisions. The learning technique offers the opportunity to the system (transportation station) to adapt dynamically its decision (choice of the best transportation mean) based on feedbacks provided by the passengers traveling from that specified station and thus optimize their journey through the transportation network. The efficacy of our proposed technique is validated through extensive simulations for different layers of transport means like metros, trains, and buses. Our proposal holds for any ITS system decisions provided the availability of real-time traces about the passengers passing by any station.

This paper aims at using the novel concept of LTE-based Device-to-Device communications (D2D) as a new alternative for Intelligent Transportation Systems (ITS) vehicular communications in advanced 4G networks and beyond. We propose the Cellular Vehicular Network (CVN) solution as a reliable and scalable operator-assisted opportunistic architecture that supports hyper-local ITS services as 3GPP Proximity Services (ProSe). A distributed D2D architecture is proposed based on a hybrid clustering approach to organize vehicles into dynamic clusters. The proposed solution includes a network setup phase based on enhanced LTE authorization and authentication procedures for ITS nodes, and an LTE direct discovery phase. A BCMP queuing network is used to model the CVN core network and to evaluate the impact of core network entities load on the network setup delay. The results are validated with Matlab and compared to the network setup delays of an existing solution.

LTE-based Device-to-Device (D2D) communications have been envisioned as a new key feature for short range wireless communications in advanced and beyond 4G networks. We propose in this work to exploit this novel concept of D2D as a new alternative for Intelligent Transportation Systems (ITS) Vehicle-to-Vehicle/Infrastructure (V2X) communications in next generation cellular networks. A 3GPP standard architecture has been recently defined to support Proximity Services (ProSe) in the LTE core network. Taking into account the limitations of this latter and the requirements of ITS services and V2X communications, we propose the CVN solution as an enhancement to the ProSe architecture in order to support hyper-local ITS services. CVN provides a reliable and scalable LTE-assisted opportunistic model for V2X communications through a distributed ProSe architecture. Using a hybrid clustering approach, vehicles are organized into dynamic clusters that are formed and managed by ProSe Cluster Heads which are elected centrally by the CVN core network. ITS services are deemed as Proximity Services and benefit from the basic ProSe discovery, authorization and authentication mechanisms. The CVN solution enhances V2V communication delays and overhead by reducing the need for multi-hop geo-routing. Preliminary simulation results show that the CVN solution provides short setup times and improves ITS communication delays.

Sensor networks suffer from limited capabilities such as bandwidth, low processing power, and memory size. There is therefore a need for protocols that deliver sensor data in an energy-efficient way to the sink. One of those techniques, it gathers sensors' data in a small size packet suitable for transmission. In this paper, we propose a new Effective Data Aggregation Protocol (DAP) to reduce the energy consumption in Wireless Sensor Networks (WSNs), which prolongs the network lifetime. This work uses in-network aggregation approach to distribute the processing all over the aggregation path to avoid unbalanced power consumption on specific nodes until they run out. Simulation results prove that DAP, compared to other protocols, achieves more data aggregation percentage and less power consumption for a one data harvesting round.

Sensor Networks are considered the heart of a wide range of applications. e-Health is one of those applications. This paper proposes a new network and Context Awareness service architecture for e-Health services. The proposal integrates the Ubiquitous Sensor Network (USN) with the IP Multimedia Subsystem (IMS)-based Next Generation Network (NGN). It provides a detailed signaling model for different healthcare scenarios based on SIP. The proposal involves modifications made to SIP to match the new e-Health features provided.