-
Optimal Resource Management for Hierarchical Federated Learning over HetNets with Wireless Energy Transfer
Authors:
Rami Hamdi,
Ahmed Ben Said,
Emna Baccour,
Aiman Erbad,
Amr Mohamed,
Mounir Hamdi,
Mohsen Guizani
Abstract:
Remote monitoring systems analyze the environment dynamics in different smart industrial applications, such as occupational health and safety, and environmental monitoring. Specifically, in industrial Internet of Things (IoT) systems, the huge number of devices and the expected performance put pressure on resources, such as computational, network, and device energy. Distributed training of Machine…
▽ More
Remote monitoring systems analyze the environment dynamics in different smart industrial applications, such as occupational health and safety, and environmental monitoring. Specifically, in industrial Internet of Things (IoT) systems, the huge number of devices and the expected performance put pressure on resources, such as computational, network, and device energy. Distributed training of Machine and Deep Learning (ML/DL) models for intelligent industrial IoT applications is very challenging for resource limited devices over heterogeneous wireless networks (HetNets). Hierarchical Federated Learning (HFL) performs training at multiple layers offloading the tasks to nearby Multi-Access Edge Computing (MEC) units. In this paper, we propose a novel energy-efficient HFL framework enabled by Wireless Energy Transfer (WET) and designed for heterogeneous networks with massive Multiple-Input Multiple-Output (MIMO) wireless backhaul. Our energy-efficiency approach is formulated as a Mixed-Integer Non-Linear Programming (MINLP) problem, where we optimize the HFL device association and manage the wireless transmitted energy. However due to its high complexity, we design a Heuristic Resource Management Algorithm, namely H2RMA, that respects energy, channel quality, and accuracy constraints, while presenting a low computational complexity. We also improve the energy consumption of the network using an efficient device scheduling scheme. Finally, we investigate device mobility and its impact on the HFL performance. Our extensive experiments confirm the high performance of the proposed resource management approach in HFL over HetNets, in terms of training loss and grid energy costs.
△ Less
Submitted 3 May, 2023;
originally announced May 2023.
-
LoRa-RL: Deep Reinforcement Learning for Resource Management in Hybrid Energy LoRa Wireless Networks
Authors:
Rami Hamdi,
Emna Baccour,
Aiman Erbad,
Marwa Qaraqe,
Mounir Hamdi
Abstract:
LoRa wireless networks are considered as a key enabling technology for next generation internet of things (IoT) systems. New IoT deployments (e.g., smart city scenarios) can have thousands of devices per square kilometer leading to huge amount of power consumption to provide connectivity. In this paper, we investigate green LoRa wireless networks powered by a hybrid of the grid and renewable energ…
▽ More
LoRa wireless networks are considered as a key enabling technology for next generation internet of things (IoT) systems. New IoT deployments (e.g., smart city scenarios) can have thousands of devices per square kilometer leading to huge amount of power consumption to provide connectivity. In this paper, we investigate green LoRa wireless networks powered by a hybrid of the grid and renewable energy sources, which can benefit from harvested energy while dealing with the intermittent supply. This paper proposes resource management schemes of the limited number of channels and spreading factors (SFs) with the objective of improving the LoRa gateway energy efficiency. First, the problem of grid power consumption minimization while satisfying the system's quality of service demands is formulated. Specifically, both scenarios the uncorrelated and time-correlated channels are investigated. The optimal resource management problem is solved by decoupling the formulated problem into two sub-problems: channel and SF assignment problem and energy management problem. Since the optimal solution is obtained with high complexity, online resource management heuristic algorithms that minimize the grid energy consumption are proposed. Finally, taking into account the channel and energy correlation, adaptable resource management schemes based on Reinforcement Learning (RL), are developed. Simulations results show that the proposed resource management schemes offer efficient use of renewable energy in LoRa wireless networks.
△ Less
Submitted 6 September, 2021;
originally announced September 2021.
-
Federated Learning over Energy Harvesting Wireless Networks
Authors:
Rami Hamdi,
Mingzhe Chen,
Ahmed Ben Said,
Marwa Qaraqe,
H. Vincent Poor
Abstract:
In this paper, the deployment of federated learning (FL) is investigated in an energy harvesting wireless network in which the base station (BS) employs massive multiple-input multiple-output (MIMO) to serve a set of users powered by independent energy harvesting sources. Since a certain number of users may not be able to participate in FL due to the interference and energy constraints, a joint en…
▽ More
In this paper, the deployment of federated learning (FL) is investigated in an energy harvesting wireless network in which the base station (BS) employs massive multiple-input multiple-output (MIMO) to serve a set of users powered by independent energy harvesting sources. Since a certain number of users may not be able to participate in FL due to the interference and energy constraints, a joint energy management and user scheduling problem in FL over wireless systems is formulated. This problem is formulated as an optimization problem whose goal is to minimize the FL training loss via optimizing user scheduling. To find how the factors such as transmit power and number of scheduled users affect the training loss, the convergence rate of the FL algorithm is first analyzed. Given this analytical result, the user scheduling and energy management optimization problem can be decomposed, simplified, and solved. Further, the system model is extended by considering multiple BSs. Hence, a joint user association and scheduling problem in FL over wireless systems is studied. The optimal user association problem is solved using the branch-and-bound technique. Simulation results show that the proposed user scheduling and user association algorithm can reduce training loss compared to a standard FL algorithm.
△ Less
Submitted 16 June, 2021;
originally announced June 2021.
-
Testing for breaks in variance structures with smooth changes
Authors:
Ben Hajria Raja,
Khardani Salah,
Raïssi Hamdi
Abstract:
The problem of detecting variance breaks in the case of smooth time-varying variance structure is studied. It is highlighted that the tests based on (piecewise) constant specification of the variance are not able to distinguish between smooth non constant variance and the case where an abrupt change is present. Consequently, a new procedure for detecting variance breaks taking into account for smo…
▽ More
The problem of detecting variance breaks in the case of smooth time-varying variance structure is studied. It is highlighted that the tests based on (piecewise) constant specification of the variance are not able to distinguish between smooth non constant variance and the case where an abrupt change is present. Consequently, a new procedure for detecting variance breaks taking into account for smooth changes of the variance is proposed. The finite sample properties of the tests introduced in the paper are investigated by Monte Carlo experiments. The theoretical outputs are illustrated using U.S. macroeconomic data.
△ Less
Submitted 21 July, 2017;
originally announced July 2017.
-
Sum-rate Maximizing in Downlink Massive MIMO Systems with Circuit Power Consumption
Authors:
Rami Hamdi,
Wessam Ajib
Abstract:
The downlink of a single cell base station (BS) equipped with large-scale multiple-input multiple-output (MIMO) system is investigated in this paper. As the number of antennas at the base station becomes large, the power consumed at the RF chains cannot be anymore neglected. So, a circuit power consumption model is introduced in this work. It involves that the maximal sum-rate is not obtained when…
▽ More
The downlink of a single cell base station (BS) equipped with large-scale multiple-input multiple-output (MIMO) system is investigated in this paper. As the number of antennas at the base station becomes large, the power consumed at the RF chains cannot be anymore neglected. So, a circuit power consumption model is introduced in this work. It involves that the maximal sum-rate is not obtained when activating all the available RF chains. Hence, the aim of this work is to find the optimal number of activated RF chains that maximizes the sum-rate. Computing the optimal number of activated RF chains must be accompanied by an adequate antenna selection strategy. First, we derive analytically the optimal number of RF chains to be activated so that the average sum-rate is maximized under received equal power. Then, we propose an efficient greedy algorithm to select the sub-optimal set of RF chains to be activated with regards to the system sum-rate. It allows finding the balance between the power consumed at the RF chains and the transmitted power. The performance of the proposed algorithm is compared with the optimal performance given by brute force search (BFS) antenna selection. Simulations allow to compare the performance given by greedy, optimal and random antenna selection algorithms.
△ Less
Submitted 14 August, 2015;
originally announced August 2015.
-
Joint Optimal Number of RF chains and Power Allocation for Downlink Massive MIMO Systems
Authors:
Rami Hamdi,
Wessam Ajib
Abstract:
This paper investigates the downlink of massive multiple-input multiple-output (MIMO) systems that include a single cell Base Station (BS) equipped with large number of antennas serving multiple users. As the number of RF chains is getting large, the system model considered in this paper assumes a non negligible circuit power consumption. Hence, the aim of this work is to find the optimal balance…
▽ More
This paper investigates the downlink of massive multiple-input multiple-output (MIMO) systems that include a single cell Base Station (BS) equipped with large number of antennas serving multiple users. As the number of RF chains is getting large, the system model considered in this paper assumes a non negligible circuit power consumption. Hence, the aim of this work is to find the optimal balance between the power consumed by the RF chains and the transmitted power. First, assuming an equal power allocation among users, the optimal number of RF chains to be activated is analytically found. Then, for a given number of RF chains we derive analytically the optimal power allocation among users. Based on these analysis, we propose an iterative algorithm that computes jointly the optimal number of RF chains and the optimal power allocation vector. Simulations validate the analytical results and show the high performance provided by the proposed algorithm.
△ Less
Submitted 20 July, 2015;
originally announced July 2015.
-
Stark widths of Ar III spectral lines in the atmospheres of subdwarf B stars
Authors:
Rafik Hamdi,
Nabil Ben Nessib,
Sylvie Sahal-Bréchot,
Milan S. Dimitrijević
Abstract:
Using semiclassical perturbation approach in impact approximation, we have calculated Stark widths for 32 spectral lines of doubly charged argon (Ar III). Oscillator strengths are calculated using Hartree-Fock method with relativistic correction (HFR) and an atomic model including 17 configurations. Energy levels are taken from NIST database. For perturbing levels for which the corresponding energ…
▽ More
Using semiclassical perturbation approach in impact approximation, we have calculated Stark widths for 32 spectral lines of doubly charged argon (Ar III). Oscillator strengths are calculated using Hartree-Fock method with relativistic correction (HFR) and an atomic model including 17 configurations. Energy levels are taken from NIST database. For perturbing levels for which the corresponding energy does not exist in NIST database, the calculated energies are used. Our widths are compared with the experimental results. The results presented here are of interest for modelling and investigation of stellar atmospheres since argon in different ionization stages is observed in many astrophysical objects. Finally, the importance of Stark broadening mechanism is studied in the atmospheric conditions of sdB stars. Electron impact Stark widths are compared to thermal Doppler widths as a function of temperature and optical depth of atmospheric layers.
△ Less
Submitted 6 November, 2013;
originally announced November 2013.
-
Atomic data and electron-impact broadening effect in DO white dwarf atmospheres: Si VI
Authors:
R. Hamdi,
N. Ben Nessib,
N. Milovanović,
L. Č. Popović,
M. S. Dimitrijević,
S. Sahal-Bréchot
Abstract:
Energy levels, electric dipole transition probabilities and oscillator strengths in five times ionized silicon have been calculated in intermediate coupling. The present calculations were carried out with the general purpose atomic-structure program SUPERSTRUCTURE. The relativistic corrections to the non-relativistic Hamiltonian are taken into account through the Breit-Pauli approximation. We ha…
▽ More
Energy levels, electric dipole transition probabilities and oscillator strengths in five times ionized silicon have been calculated in intermediate coupling. The present calculations were carried out with the general purpose atomic-structure program SUPERSTRUCTURE. The relativistic corrections to the non-relativistic Hamiltonian are taken into account through the Breit-Pauli approximation. We have also introduced a semi-empirical correction (TEC) for the calculation of the energy-levels. These atomic data are used to provide semiclassical electron-, proton- and ionized helium- impact line widths and shifts for 15 Si VI muliplet. Calculated results have been used to consider the influence of Stark broadening for DO white dwarf atmospheric conditions.
△ Less
Submitted 4 May, 2008;
originally announced May 2008.
