Publications

2024
Telecom Language Models: Must They Be Large?
Nicola Piovesan, Antonio De Domenico, Fadhel Ayed
Submitted
Preprint

The increasing interest in Large Language Models (LLMs) within the telecommunications sector underscores their potential to revolutionize operational efficiency. However, the deployment of these sophisticated models is often hampered by their substantial size and computational demands, raising concerns about their viability in resource-constrained environments. Addressing this challenge, recent advancements have seen the emergence of small language models that surprisingly exhibit performance comparable to their larger counterparts in many tasks, such as coding and common-sense reasoning. Phi-2, a compact yet powerful model, exemplifies this new wave of efficient small language models. This paper conducts a comprehensive evaluation of Phi-2's intrinsic understanding of the telecommunications domain. Recognizing the scale-related limitations, we enhance Phi-2's capabilities through a Retrieval-Augmented Generation approach, meticulously integrating an extensive knowledge base specifically curated with telecom standard specifications. The enhanced Phi-2 model demonstrates a profound improvement in accuracy, answering questions about telecom standards with a precision that closely rivals the more resource-intensive GPT-3.5. The paper further explores the refined capabilities of Phi-2 in addressing problem-solving scenarios within the telecom sector, highlighting its potentials and limitations.

Linguistic Intelligence in Large Language Models for Telecommunications
Tasnim Ahmed, Nicola Piovesan, Antonio De Domenico, Salimur Choudhury
Submitted
Preprint

Large Language Models (LLMs) have emerged as a significant advancement in the field of Natural Language Processing (NLP), demonstrating remarkable capabilities in language generation and other language-centric tasks. Despite their evaluation across a multitude of analytical and reasoning tasks in various scientific domains, a comprehensive exploration of their knowledge and understanding within the realm of natural language tasks in the telecommunications domain is still needed. This study, therefore, seeks to evaluate the knowledge and understanding capabilities of LLMs within this domain. To achieve this, we conduct an exhaustive zero-shot evaluation of four prominent LLMs—Llama-2, Falcon, Mistral, and Zephyr. These models require fewer resources than ChatGPT, making them suitable for resource-constrained environments. Their performance is compared with state-of-the-art, fine-tuned models. To the best of our knowledge, this is the first work to extensively evaluate and compare the understanding of LLMs across multiple language-centric tasks in this domain. Our evaluation reveals that zero-shot LLMs can achieve performance levels comparable to the current state-of-the-art fine-tuned models. This indicates that pretraining on extensive text corpora equips LLMs with a degree of specialization, even within the telecommunications domain. We also observe that no single LLM consistently outperforms others, and the performance of different LLMs can fluctuate. Although their performance lags behind fine-tuned models, our findings underscore the potential of LLMs as a valuable resource for understanding various aspects of this field that lack large annotated data.

Data-driven Energy Efficiency Modelling in Large-scale Networks: An Expert Knowledge and ML-based Approach
David López-Pérez, Antonio De Domenico, Nicola Piovesan, Merouane Debbah
Submitted
Preprint

The energy consumption of mobile networks poses a critical challenge. Mitigating this concern necessitates the deployment and optimization of network energy-saving solutions, such as carrier shutdown, to dynamically manage network resources. Traditional optimization approaches encounter complexity due to factors like the large number of cells, stochastic traffic, channel variations, and intricate trade-offs. This paper introduces the simulated reality of communication networks (SRCON) framework, a novel, data-driven modeling paradigm that harnesses live network data and employs a blend of machine learning (ML)- and expert-based models. These mix of models accurately characterizes the functioning of network components, and predicts network energy efficiency and user equipment (UE) quality of service for any energy carrier shutdown configuration in a specific network. Distinguishing itself from existing methods, SRCON eliminates the reliance on expensive expert knowledge, drive testing, or incomplete maps for predicting network performance. This paper details the pipeline employed by SRCON to decompose the large network energy efficiency modeling problem into MLand expert-based submodels. It demonstrates how, by embracing stochasticity, and carefully crafting the relationship between such submodels, the overall computational complexity can be reduced and prediction accuracy enhanced. Results derived from real network data underscore the paradigm shift introduced by SRCON, showcasing significant gains over a state-of-theart method used by a operator for network energy efficiency modeling. The reliability of this local, data-driven modeling of the network proves to be a key asset for network energy-saving optimization.


2023
FlexTrain: A Dynamic Training Framework for Heterogeneous Devices Environments
Mert Unsal, Ali Maatouk, Antonio De Domenico, Nicola Piovesan, Fadhel Ayed
NeurIPS 2023, 10-16 December 2023, New Orleans (USA)
Preprint

As deep learning models become increasingly large, they pose significant challenges in heterogeneous devices environments. The size of deep learning models makes it difficult to deploy them on low-power or resource-constrained devices, leading to long inference times and high energy consumption. To address these challenges, we propose FlexTrain, a framework that accommodates the diverse storage and computational resources available on different devices during the training phase. FlexTrain enables efficient deployment of deep learning models, while respecting device constraints, minimizing communication costs, and ensuring seamless integration with diverse devices. We demonstrate the effectiveness of FlexTrain on the CIFAR-100 dataset, where a single global model trained with FlexTrain can be easily deployed on heterogeneous devices, saving training time and energy consumption. We also extend FlexTrain to the federated learning setting, showing that our approach outperforms standard federated learning benchmarks on both CIFAR-10 and CIFAR-100 datasets.

TeleQnA: A Benchmark Dataset to Assess Large Language Models Telecommunications Knowledge
Ali Maatouk, Fadhel Ayed, Nicola Piovesan, Antonio De Domenico, Merouane Debbah, Zhi-Quan Luo
Submitted
Preprint Dataset

We introduce TeleQnA, the first benchmark dataset designed to evaluate the knowledge of Large Language Models (LLMs) in telecommunications. Comprising 10,000 questions and answers, this dataset draws from diverse sources, including standards and research articles. This paper outlines the automated question generation framework responsible for creating this dataset, along with how human input was integrated at various stages to ensure the quality of the questions. Afterwards, using the provided dataset, an evaluation is conducted to assess the capabilities of LLMs, including GPT-3.5 and GPT-4. The results highlight that these models struggle with complex standards related questions but exhibit proficiency in addressing general telecom-related inquiries. Additionally, our results showcase how incorporating telecom knowledge context significantly enhances their performance, thus shedding light on the need for a specialized telecom foundation model. Finally, the dataset is shared with active telecom professionals, whose performance is subsequently benchmarked against that of the LLMs. The findings illustrate that LLMs can rival the performance of active professionals in telecom knowledge, thanks to their capacity to process vast amounts of information, underscoring the potential of LLMs within this domain. The dataset has been made publicly accessible on GitHub.

Large Language Models for Telecom: Forthcoming Impact on the Industry
Ali Maatouk, Nicola Piovesan, Fadhel Ayed, Antonio De Domenico, Merouane Debbah
Submitted
Preprint

Large Language Models (LLMs) have emerged as a transformative force, revolutionizing numerous fields well beyond the conventional domain of Natural Language Processing (NLP) and garnering unprecedented attention. As LLM technology continues to progress, the telecom industry is facing the prospect of its potential impact on its landscape. To elucidate these implications, we delve into the inner workings of LLMs, providing insights into their current capabilities and limitations. We also examine the use cases that can be readily implemented in the telecom industry, streamlining numerous tasks that currently hinder operational efficiency and demand significant manpower and engineering expertise. Furthermore, we uncover essential research directions that deal with the distinctive challenges of utilizing the LLMs within the telecom domain. Addressing these challenges represents a significant stride towards fully harnessing the potential of LLMs and unlocking their capabilities to the fullest extent within the telecom domain.

High Altitude Platform Stations: the New Network Energy Efficiency Enabler in the 6G Era
Tailai Song, David López-Pérez, Michela Meo, Nicola Piovesan, Daniela Renga
Submitted
Preprint

The rapidly evolving communication landscape, with the advent of 6G technology, brings new challenges to the design and operation of wireless networks. One of the key concerns is the energy efficiency of the Radio Access Network (RAN), as the exponential growth in wireless traffic demands increasingly higher energy consumption. In this paper, we assess the potential of integrating a High Altitude Platform Station (HAPS) to improve the energy efficiency of a RAN, and quantify the potential energy conservation through meticulously designed simulations. We propose a quantitative framework based on real traffic patterns to estimate the energy consumption of the HAPS integrated RAN and compare it with the conventional terrestrial RAN. Our simulation results elucidate that HAPS can significantly reduce energy consumption by up to almost 30\% by exploiting the unique advantages of HAPS, such as its self-sustainability, high altitude, and wide coverage. We further analyze the impact of different system parameters on performance, and provide insights for the design and optimization of future 6G networks. Our work sheds light on the potential of HAPS integrated RAN to mitigate the energy challenges in the 6G era, and contributes to the sustainable development of wireless communications.

Energy efficient cell-free massive MIMO on 5G deployments: sleep modes strategies and user stream management
Felip Riera-Palou, Guillem Femenias, David López-Pérez, Nicola Piovesan, Antonio De Domenico
Submitted
Preprint

This paper proposes the utilization of cell-free massive MIMO (CF-M-MIMO) processing on top of the regular micro/macrocellular deployments typically found in current 5G networks. Towards this end, it contemplates the connection of all base stations to a central processing unit (CPU) through fronthaul links, thus enabling the joint processing of all serviced user equipment (UE), yet avoiding the expensive deployment and maintenance of dozens of randomly scattered access points (APs). Moreover, it allows the implementation of centralized strategies to exploit the sleep mode capabilities of current baseband/RF hardware to (de)activate selected Base Stations (BSs) in order to maximize the network energy efficiency and to react to changes in UE behaviour and/or operator requirements. In line with current cellular network deployments, it considers the use of multiple antennas at the UE side that unavoidably introduces the need to effectively manage the number of streams to be directed to each UE in order to balance multiplexing gains and increased pilot contamination.

Rate, Power, and Energy Efficiency trade-offs in Massive MIMO Systems with Carrier Aggregation
Alessio Zappone, David López-Pérez, Antonio De Domenico, Nicola Piovesan, Harvey Bao
IEEE Transactions on Green Communications and Networking
Paper

This work considers a multi-cell, multi-carrier massive MIMO network with carrier aggregation capabilities, and tackles both the rate versus power consumption and the rate versus energy efficiency (EE) trade-offs, by jointly optimizing the number of employed carriers, transmit antennas, base station density, and transmit power. Provably convergent algorithms for both trade-off problems are developed, together with closed-form results for the individual optimization of the considered resources, taking three main carrier aggregation techniques into account, namely inter-carrier, intra-carrier contiguous, and intra-carrier non-contiguous. Numerical results show how the use of carrier aggregation represents an effective way of increasing the network rate and EE, while keeping the power consumption at bay. By using carrier aggregation, it is possible to reduce the number of deployed antennas, without sacrificing the rate performance and increasing the system EE.

A Novel Metric for mMIMO Base Station Association in Aerial Highway Systems
Matteo Bernabè, David López-Pérez, Nicola Piovesan, Giovanni Geraci, David Gesbert
IEEE International Conference on Communications (ICC), 28 May-1 June 2023, Rome (Italy)
Preprint

In this article, we introduce a new metric for driving the serving cell selection process of a swarm of cellular connected unmanned aerial vehicles (CCUAVs) located on aerial highways when served by a massive multiple input multiple output (mMIMO) terrestrial network. Selecting the optimal serving cell from several suitable candidates is not straightforward. By solely relying on the traditional cell selection metric, based on reference signal received power (RSRP), it is possible to result in a scenario in which the serving cell can not multiplex an appropriate number of CCUAVs due to the high correlation in the line of sight (LoS) channels. To overcome such issue, in this work, we introduce a new cell selection metric to capture not only signal strength, but also spatial multiplexing capabilities. The proposed metric highly depends on the relative position between the aerial highways and the antennas of the base station. The numerical analysis indicates that the integration of the proposed new metric allows to have a better signal to interference plus noise ratio (SINR) performance on the aerial highways, resulting in a more reliable cellular connection for CCUAVs.

Sleep mode strategies for energy efficient cell-free massive MIMO in 5G deployments
Felip Riera-Palou, Guillem Femenias, David López-Pérez, Nicola Piovesan, Antonio De Domenico
IEEE International Conference on Communications (ICC), 28 May-1 June 2023, Rome (Italy)

Power Consumption Modeling of 5G Multi-Carrier Base Stations: A Machine Learning Approach
Nicola Piovesan, David López-Pérez, Antonio De Domenico, Xinli Geng, Harvey Bao
IEEE International Conference on Communications (ICC), 28 May-1 June 2023, Rome (Italy)
Preprint

The fifth generation of the Radio Access Network (RAN) has brought new services, technologies, and paradigms with the corresponding societal benefits. However, the energy consumption of 5G networks is today a concern. In recent years, the design of new methods for decreasing the RAN power consumption has attracted interest from both the research community and standardization bodies, and many energy savings solutions have been proposed. However, there is still a need to understand the power consumption behavior of state-ofthe-art base station architectures, such as multi-carrier active antenna units (AAUs), as well as the impact of different network parameters. In this paper, we present a power consumption model for 5G AAUs based on artificial neural networks. We demonstrate that this model achieves good estimation performance, and it is able to capture the benefits of energy saving when dealing with the complexity of multi-carrier base stations architectures. Importantly, multiple experiments are carried out to show the advantage of designing a general model able to capture the power consumption behaviors of different types of AAUs. Finally, we provide an analysis of the model scalability and the training data requirements.

Modelling User Transfer during Dynamic Carrier Shutdown in Green 5G Networks
Antonio De Domenico, David López-Pérez, Wenjie Li, Nicola Piovesan, Harvey Bao, Xinli Geng
IEEE Transactions on Wireless Communications
Paper

The energy consumption of the fifth generation (5G) of cellular technology is concerning for the mobile industry and the entire society. To minimize the environmental footprint and economic costs of 5G, it is necessary to adapt the transmission capabilities of networks to end-users’ quality of service requirements. In this paper, we focus on the carrier shutdown approach that enables a base station (BS) to autonomously switch off during low traffic periods, by transferring its load to neighbouring active BSs. More specifically, we propose a data-driven framework, constructed through real network measurements, which statistically characterizes the user equipment (UE) transfer across neighbouring BSs, when carrier shutdown operates. The implementation of this framework allows the 5G system to determine a poor load distribution due to energy saving mechanisms, prevent drastic reductions in UE performance, and ultimately estimate energy savings when activating carrier shutdown.


2022
Carrier Aggregation for Improved Rate versus Power trade-off in Massive MIMO Systems
Alessio Zappone, David López-Pérez, Antonio De Domenico, Nicola Piovesan, Harvey Bao
IEEE Global Communications Conference (GLOBECOM), 4-8 December 2022, Rio de Janeiro (Brazil)
Paper

This work considers a multi-cell, multi-carrier massive MIMO network with carrier aggregation capabilities, and tackles the rate versus power consumption trade-off, by jointly optimizing the number of employed carriers, transmit antennas, base station density, and transmit power. A provably convergent algorithm is developed together with closed-form results for the individual optimization of the considered resources, Numerical results show how the use of carrier aggregation represents an effective way of reducing the number of deployed antennas, reducing the power consumption without sacrificing the rate performance

Machine Learning and Analytical Power Consumption Models for 5G Base Stations
Nicola Piovesan, David López-Pérez, Antonio De Domenico, Xinli Geng, Harvey Bao, Mérouane Debbah
IEEE Communications Magazine
Paper Preprint

The energy consumption of the fifth generation (5G) of mobile networks is one of the major concerns of the telecom industry. However, there is not currently an accurate and tractable approach to evaluate 5G base stations (BSs) power consumption. In this article, we propose a novel model for a realistic characterisation of the power consumption of 5G multi-carrier BSs, which builds on a large data collection campaign. At first, we define a new machine learning architecture that allows modelling multiple 5G BS products. Then, we exploit the knowledge gathered by our machine learning framework to derive a realistic but also analytically tractable power consumption model, which can help driving both theoretical analyses as well as feature standardisation, development and optimisation frameworks. Notably, we demonstrate that such model has high precision, and it is able of capturing the benefits of energy saving mechanisms. We believe this analytical model represents a fundamental tool for understanding 5G BSs power consumption, and accurately optimising the network energy efficiency.

A Survey on 5G Radio Access Network Energy Efficiency: Massive MIMO, Lean Carrier Design, Sleep Modes, and Machine Learning
David López-Pérez, Antonio De Domenico, Nicola Piovesan, Harvey Bao, Xinli Geng, Song Qitao, Mérouane Debbah
IEEE Communications Surveys and Tutorials
Paper Preprint

Cellular networks have changed the world we are living in, and the fifth generation (5G) of radio technology is expected to further revolutionise our everyday lives, by enabling a high degree of automation, through its larger capacity, massive connectivity, and ultra-reliable low-latency communications. In addition, the third generation partnership project (3GPP) new radio (NR) specification also provides tools to significantly decrease the energy consumption and the green house emissions of next generations networks, thus contributing towards information and communication technology (ICT) sustainability targets. In this survey paper, we thoroughly review the state-of-the-art on current energy efficiency research. We first categorise and carefully analyse the different power consumption models and energy efficiency metrics, which have helped to make progress on the understanding of green networks. Then, as a main contribution, we survey in detail -- from a theoretical and a practical viewpoint -- the main energy efficiency enabling technologies that 3GPP NR provides, together with their main benefits and challenges. Special attention is paid to four key enabling technologies, i.e., massive multiple-input multiple-output (MIMO), lean carrier design, and advanced idle modes, together with the role of artificial intelligence capabilities. We dive into their implementation and operational details, and thoroughly discuss their optimal operation points and theoretical-trade-offs from an energy consumption perspective. This will help the reader to grasp the fundamentals of -- and the status on -- green networking. Finally, the areas of research where more effort is needed to make future networks greener are also discussed.



2021
Mobile Traffic Forecasting for Green 5G Networks
Nicola Piovesan, Antonio De Domenico, David López-Pérez, Harvey Bao, Xinli Geng, Xie Wang, Mérouane Debbah
IEEE Global Communications Conference (GLOBECOM), 7-11 December 2021, Madrid (Spain)
Paper

The energy consumption and carbon footprint of the fifth-generation (5G) of mobile technology is a current concern to mobile network operators (MNOs). These are currently attempting to lower both their carbon emissions and electricity bills by investigating new schemes that allow adapting the network transmission capabilities to the end-users' quality of service (QoS) requirements. Many of such schemes rely on accurate traffic forecasting, and as a consequence, there is a large effort on investigating novel machine learning (ML) algorithms, which fed by network measurement data and empowered by the computing capabilities of dedicated hardware, can help modelling and predicting users' behaviours. Most of the works in the literature, however, focus on predicting the traffic when energy saving features, e.g. carrier shutdown, are not implemented or activated. However, the prediction task becomes much more challenging when energy saving features are adopted due to their impact to the actual measured traffic. In this paper, we consider a scenario in which part of the base stations implement energy saving schemes, which allow them to dynamically switch off part of their hardware to reduce their power consumption. Then, we present a ML framework based on graph convolutional networks (GCNs) for traffic forecasting in such dynamic scenarios, and compare its performance with other statistical and ML prediction algorithms. The proposed GCN framework provides significant accuracy gains. Moreover, we provide an analysis of the impact of spatial correlation ---captured by the GCN model--- on the achieved performance.

Energy Efficiency of Multi-Carrier Massive MIMO Networks: Massive MIMO Meets Carrier Aggregation
David López-Pérez, Antonio De Domenico, Nicola Piovesan, Xinli Geng, Harvey Bao, Mérouane Debbah
IEEE Global Communications Conference (GLOBECOM), 7-11 December 2021, Madrid (Spain)
Paper

The energy consumption of cellular networks, despite the high energy efficiency of the fifth generation (5G) of mobile technology, is still a challenge. The fundamental problem arises due to the complexity of optimising the operation of the available rich set of energy efficiency features in large-scale deployments. To assist such optimisation, a large body of research --with the resulting understanding and algorithms-- exists, particularly on the energy efficiency of single-cell massive multiple-input multiple-output systems. However, other fundamental cellular features, such as those relating to multi-carrier systems, remain largely unexplored. In this paper, we show how multi-carrier features, such as carrier aggregation, can play a significant role in energy savings, and question the need for hundreds of antennas and transceiver chains at the base stations as an urgent solution to increase the energy efficiency of next generation networks.

Forecasting Mobile Traffic to Achieve Greener 5G Networks: When Machine Learning is Key
Nicola Piovesan, Antonio De Domenico, Matteo Bernabè, David López-Pérez, Harvey Bao, Xinli Geng, Xie Wang, Mérouane Debbah
IEEE International Workshop on Signal Processing Advances in Wireless Communications (SPAWC), 27-30 September 2021, Lucca (Italy)
Paper

To minimise the energy consumption of the fifth generation (5G) of mobile technology, it is necessary to adapt the transmission capabilities of 5G networks to the end-users' quality of service (QoS) requirements. In this line, researchers are investigating novel machine learning (ML) algorithms, which fed by network measurement data and empowered by the computing capabilities of dedicated hardware, can help modelling and predicting user behaviour. In this paper, we focus on the task of forecasting mobile traffic, which is a key step to dynamically adapt network parameters to space-time traffic variations by optimizing energy efficiency features, such as carrier shutdown. Specifically, we present a state-of-the-art ML framework based on Graph convolutional Networks, and compare its performance with simpler schemes, characterized by a lower complexity. Our analysis highlights advantages and drawbacks of these solutions, paying special attention to the importance of the error metric selection, feature pre-processing and historical data availability.



2020
Deep Reinforcement Learning for Autonomous Mobile Networks in Micro-Grids
Marco Miozzo, Nicola Piovesan, Dagnachew Temesgene, Paolo Dini
Deep Learning for Unmanned Systems
Chapter

In this chapter, we describe the design of controlling schemes for energy self-sustainable mobile networks through Deep Learning. The goal is to enable an intelligent energy management that allows the base stations to mostly operate off-grid by using renewable energies. To achieve this goal, we formulate an on-line grid energy and network throughput optimization problem considering both centralized and distributed Deep Reinforcement Learning implementations. We provide an exhaustive discussion on the reference scenario, the techniques adopted, the achieved performance, the complexity and the feasibility of the proposed models, together with the energy and cost savings attained. Results demonstrate that Deep Q-Learning based algorithms represent a viable and economically convenient solution for enabling energy self-sustainability of mobile networks grouped in micro-grids.

Joint Load Control and Energy Sharing for Renewable Powered Small Base Stations: a Machine Learning Approach
Nicola Piovesan, David López-Pérez, Marco Miozzo, Paolo Dini
IEEE Transactions on Green Communication and Networking
Paper

The deployment of dense networks of small base stations represents one of the most promising solutions for future mobile networks to meet the foreseen increasing traffic demands. However, such an infrastructure consumes a considerable amount of energy, which, in turn, may represent an issue for the environment and the operational expenses of the mobile operators. The use of renewable energy to supply the small base stations has been recently considered as a mean to reduce the energy footprint of the mobile networks. In this paper, we consider a hierarchical structure in which part of the base stations are powered exclusively by solar panels and batteries. Base stations are grouped in clusters and connected in a micro-grid. A central controller enables base station sleep mode and energy sharing among the base stations based on the available energy budget and the traffic demands. We propose three different implementations of the controller through Machine Learning models, namely Imitation Learning, Q-Learning and Deep Q-Learning, capable of learning optimal sleep mode and energy sharing policies. We provide an exhaustive discussion on the achieved performance, complexity and feasibility of the proposed models together with the energy and cost savings attained.

Network Resource Allocation Policies with Energy Transfer Capabilities
Nicola Piovesan
Ph.D. dissertation
Thesis

During the last decades, mobile network operators have witnessed an exponential increase in the traffic demand, mainly due to the high request of services from a huge amount of users. The trend is of a further increase in both the traffic demand and the number of connected devices over the next years. The traffic load is expected to have an annual growth rate of 53% for the mobile network alone, and the upcoming industrial era, which will connect different types of devices to the mobile infrastructure including human and machine type communications, will definitely exacerbate such an increasing trend. The current directions anticipate that future mobile networks will be composed of ultra dense deployments of heterogeneous Base Stations (BSs), where BSs using different transmission powers coexist. Accordingly, the traditional Macro BSs layer will be complemented or replaced with multiple overlapping tiers of small BSs (SBSs), which will allow extending the system capacity. However, the massive use of Information and Communication Technology (ICT) and the dense deployment of network elements is going to increase the level of energy consumed by the telecommunication infrastructure and its carbon footprint on the environment. Current estimations indicates that 10% of the worldwide electricity generation is due to the ICT industry and this value is forecasted to reach 51% by 2030, which imply that 23% of the carbon footprint by human activity will be due to ICT. Environmental sustainability is thus a key requirement for designing next generation mobile networks. Recently, the use of Renewable Energy Sources (RESs) for supplying network elements has attracted the attention of the research community, where the interest is driven by the increased efficiency and the reduced costs of energy harvesters and storage devices, specially when installed to supply SBSs. Such a solution has been demonstrated to be environmentally and economically sustainable in both rural and urban areas. However, RESs will entail a higher management complexity. In fact, environmental energy is inherently erratic and intermittent, which may cause a fluctuating energy inflow and produce service outage. A proper control of how the energy is drained and balanced across network elements is therefore necessary for a self-sustainable network design. In this dissertation, we focus on energy harvested through solar panels that is deemed the most appropriate due to the good efficiency of commercial photovoltaic panels as well as the wide availability of the solar source for typical installations. The characteristics of this energy source are analyzed in the first technical part of the dissertation, by considering an approach based on the extraction of features from collected data of solar energy radiation. In the second technical part of the thesis we introduce our proposed scenario. A federation of BSs together with the distributed harvesters and storage devices at the SBS sites form a micro-grid, whose operations are managed by an energy management system in charge of controlling the intermittent and erratic energy budget from the RESs. We consider load control (i.e., enabling sleep mode in the SBSs) as a method to properly manage energy inflow and spending, based on the traffic demand. Moreover, in the third technical part, we introduce the possibility of improving the network energy efficiency by sharing the exceeding energy that may be available at some BS sites within the micro-grid. Finally, centralized controller implementations based on supervised and reinforcement learning are proposed in the last technical part of the dissertation. The controller is in charge of opportunistically operating the network to achieve efficient utilization of the harvested energy and prevent SBSs blackout.

Modeling the Environment in Deep Reinforcement Learning: the case of Energy Harvesting Base Stations
Nicola Piovesan, Marco Miozzo, Paolo Dini
IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), 4-8 May 2020, Barcelona (Spain)
Paper

In this paper, we focus on the design of energy self-sustainable mobile networks by enabling intelligent energy management that allows the base stations to mostly operate off-grid by using renewable energy. We propose a centralized control algorithm based on Deep Reinforcement Learning. The single agent is able to learn how to efficiently balance the energy inflow and spending among base stations observing the environment and interacting with it. In particular, we provide a study on the performance achieved by this approach when considering different representations of the environment. Numerical results demonstrate that using a good level of abstraction in the choice of the representation variables may enable a proper mapping of the environment into actions to take, so as to maximize the numerical reward.



2019
Coordinated Load Control of Renewable Powered Small Base Stations through Layered Learning
Marco Miozzo, Nicola Piovesan, Paolo Dini
IEEE Transactions on Green Communications and Networking
Paper

The massive deployment of Small Base Stations (SBSs) represents one of the most promising solutions adopted by 5G cellular networks to meet the foreseen huge traffic demand. The usage of renewable energies for powering the SBSs attracted particular attention for reducing the energy footprint and, thus, mitigating the environmental impact of mobile networks and enabling cost saving for the operators. The complexity of the system and the variability of the harvesting process suggest the adoption of learning methods. Here, we investigate techniques based on the Layered Learning paradigm to control dense networks of SBSs powered solely by solar energy. In the first layer, SBSs locally select switch ON/OFF policies according to their energy income and traffic demand based on a Heuristically Accelerated Reinforcement Learning method. The second layer relies on an Artificial Neural Network that estimates the network load conditions to implement a centralized controller enforcing local agent decisions. Simulation results prove that the control of the proposed framework mimics the behavior of the upper bound obtained offline with Dynamic Programming. Moreover, the proposed layered framework outperforms both a greedy and a distributed Reinforcement Learning solution in terms of throughput and energy efficiency under different traffic conditions.

Joint Load Control and Energy Sharing for Autonomous Operation of 5G Mobile Networks in Micro-Grids
Nicola Piovesan, Dagnachew Temesgene, Marco Miozzo, Paolo Dini
IEEE Access
Paper

In this paper, we focus on the design of energy self-sustainable mobile networks, by enabling intelligent energy management that allows the base stations to mostly operate off-grid by using renewable energies. Many papers are available in the literature on this problem, however, we are approaching this issue from a different angle. In fact, we advocate for future mobile networks with a hierarchical cell structure and powered by energy harvesting hardware. Base stations within the same geographical area are grouped in a micro-grid and operate almost autonomously from the power grid. To achieve this goal, we target the design of optimal traffic and computational load control method with energy sharing within the micro-grid. We solve the optimization problem using a graph-based method and we demonstrate, via software simulations, that a combination of load control plus energy sharing represents a viable and economically convenient solution for enabling energy self-sustainability of mobile networks grouped in micro-grids.



2018
Unsupervised Learning of Representations from Solar Energy Data
Nicola Piovesan, Paolo Dini
IEEE International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC), 9–12 September 2018, Bologna (Italy)
Paper Preprint

In this paper, we propose an unsupervised method to learn hidden features of the solar energy generation from a PV system that may give a more accurate characterization of the process. In a first step, solar radiation data is converted into instantaneous solar power through a detailed source model. Then, two different approaches, namely PCA and autoencoder, are used to extract meaningful features from the traces of the solar energy generation. We interpret the latent variables characterizing the solar energy generation process by analyzing the similarities of 67 cities in Europe, North-Africa and Middle-East through an agglomerative hierarchical clustering algorithm. This analysis provides also a comparison between the feature extraction capabilities of the PCA and the autoencoder.

Optimal Placement of Baseband Functions for Energy Harvesting Virtual Small Cells
Dagnachew Temesgene, Nicola Piovesan, Marco Miozzo, Paolo Dini
IEEE Vehicular Technology Conference: VTC2018-Fall, 27–30 August 2018, Chicago (USA)
Preprint

Flexible functional split in Cloud Radio Access Network (CRAN) greatly overcomes fronthaul capacity and latency challenges. In such architecture, part of the baseband processing is done locally and the remaining is done remotely in the central cloud. On the other hand, Energy Harvesting (EH) technologies are increasingly adopted due to sustainability and economic advantages. Power consumption due to baseband processing has a huge share in the total power consumption breakdown of smaller base stations. Given that such base stations are powered by EH, in addition to QoS constraints, energy availability also conditions the decision on where to place each baseband function in the system. This work focuses on determining the performance bounds of an optimal placement of baseband functional split option in virtualized small cells that are solely powered by EH. The work applies Dynamic Programming (DP), in particular, Shortest Path search is used to determine the optimal functional split option considering traffic requirements and available energy budget.

Optimal Direct Load Control of Renewable Powered Small Cells: Performance Evaluation and Bounds
Nicola Piovesan, Marco Miozzo, Paolo Dini
IEEE Wireless Communications and Networking Conference (WCNC), 15-18 April 2018, Barcelona (Spain)
Paper Preprint

In this paper, we propose an optimal direct load control of renewable powered small base stations based on Dynamic Programming. The optimization is represented using Graph Theory and the problem is stated as a Shortest Path problem. The proposed optimal algorithm is able to adapt to the varying conditions of renewable energy sources and traffic demands. We analyze the optimal ON/OFF policies considering different energy and traffic scenarios. Then, we evaluate network performance in terms of system drop rate and grid energy consumption. The obtained results are compared with a greedy approach. This study allows to elaborate on the behavior and performance bounds of the system and gives a guidance for approximated policy search methods.

Energy sustainable paradigms and methods for future mobile networks: A survey
Nicola Piovesan, Angel F. Gambin, Marco Miozzo, Michele Rossi, Paolo Dini
Elsevier Computer Communications, 2018.
Paper Preprint

In this survey, we discuss the role of energy in the design of future mobile networks and, in particular, we advocate and elaborate on the use of energy harvesting (EH) hardware as a means to decrease the environmental footprint of 5G technology. To take full advantage of the harvested (renewable) energy, while still meeting the quality of service required by dense 5G deployments, suitable management techniques are here reviewed, highlighting the open issues that are still to be solved to provide eco-friendly and cost-effective mobile architectures. Several solutions have recently been proposed to tackle capacity, coverage and efficiency problems, including: C-RAN, Software Defined Networking (SDN) and fog computing, among others. However, these are not explicitly tailored to increase the energy efficiency of networks featuring renewable energy sources, and have the following limitations: (i) their energy savings are in many cases still insufficient and (ii) they do not consider network elements possessing energy harvesting capabilities. In this paper, we systematically review existing energy sustainable paradigms and methods to address points (i) and (ii), discussing how these can be exploited to obtain highly efficient, energy self-sufficient and high capacity networks. Several open issues have emerged from our review, ranging from the need for accurate energy, transmission and consumption models, to the lack of accurate data traffic profiles, to the use of power transfer, energy cooperation and energy trading techniques. These challenges are here discussed along with some research directions to follow for achieving sustainable 5G systems.



2017
Optimal Direct Load Control of Renewable Powered Small Cells: A Shortest Path Approach
Nicola Piovesan, Paolo Dini
Internet Technology Letters, Wiley, 2017.
Paper

In this letter, we propose an optimal direct load control of renewable powered small base stations (SBSs) in a two-tier mobile network based on dynamic programming (DP). We represent the DP optimization using Graph Theory and state the problem as a Shortest Path search. We use the Label Correcting Method to explore the graph and find the optimal ON/OFF policy for the SBSs. Simulation results demonstrate that the proposed algorithm is able to adapt to the varying conditions of the environment, namely renewable energy arrivals and traffic demands. The key benefit of our study is that it allows to elaborate on the behavior and performance bounds of the system.



2016
Cooperative Localization in WSNs: a Hybrid Convex/non-Convex Solution
Nicola Piovesan, Tomaso Erseghe
IEEE Transactions on Signal and Information Processing over Networks, 2016.
Paper Preprint

Data Analytics for Smart Parking Applications
Nicola Piovesan, Leo Turi, Enrico Toigo, Borja Martinez, Michele Rossi
MDPI Sensors, 2016.
Paper

We consider real-life smart parking systems where parking lot occupancy data are collected from field sensor devices and sent to backend servers for further processing and usage for applications. Our objective is to make these data useful to end users, such as parking managers, and, ultimately, to citizens. To this end, we concoct and validate an automated classification algorithm having two objectives: (1) outlier detection: to detect sensors with anomalous behavioral patterns, i.e., outliers; and (2) clustering: to group the parking sensors exhibiting similar patterns into distinct clusters. We first analyze the statistics of real parking data, obtaining suitable simulation models for parking traces. We then consider a simple classification algorithm based on the empirical complementary distribution function of occupancy times and show its limitations. Hence, we design a more sophisticated algorithm exploiting unsupervised learning techniques (self-organizing maps). These are tuned following a supervised approach using our trace generator and are compared against other clustering schemes, namely expectation maximization, k-means clustering and DBSCAN, considering six months of data from a real sensor deployment. Our approach is found to be superior in terms of classification accuracy, while also being capable of identifying all of the outliers in the dataset.

Distributed algorithms for localization in Wireless Sensor Networks
Nicola Piovesan
M.Sc. thesis, 2016.
Thesis  

We introduce the concept of localization in Wireless Sensor Networks, starting from ranging measurements available at sensor nodes. We explore different solutions available in the literature and then we introduce the algorithms proposed by two recent publications. Finally, we propose a new hybrid ADMM algorithm that, considering both the original non-convex problem and a convexification of the problem, allows to obtain better performances.



2013
Support Vector Machines for classification and statistical regression
Nicola Piovesan
B.Sc. thesis, 2013.
Thesis  

Il lavoro studia inizialmente l’utilizzo di Support Vector Machine per la classificazione. Vengono introdotte SVM lineari per dati separabili e non separabili e SVM non lineari con Kernel. In seguito, quanto visto in via teorica viene applicato a due casi di classificazione (Iris di Fisher e classificazione dei dati di una analisi SPECT). Per finire vengono introdotte le SVM per la regressione statistica.