GDGhassan Dahman

Ghassan Dahman

*This is an algorithmically generated author record based on publicly available information.

Ghassan Dahman has a total of 30 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    New Approaches for Network Topology Optimization Using Deep Reinforcement Learning and Graph Neural Network

    by Mohammed Ali, Florent Duchesne, Ghassan Dahman, François Gagnon, Diala Naboulsi

    The exponential growth in Internet-connected devices has escalated the demand for optimized network topologies to ensure high performance. Traditional optimization methods often fall short in scalability and adaptability when it comes to network topology planning. In this paper, we address the challenge of transforming mesh topologies into tree topologies for wireless networks, with the objective of maximizing throughput. We propose two new methods: Path Selection with Rejection Strategy (PSRS), which leverages Message-Passing Neural Networks (MPNN), and Dual-Agent Tree Topology Exploration (DATTE), which employs Graph Attention Networks (GAT). These schemes integrate Deep Reinforcement Learning (DRL) and Graph Neural Networks (GNNs) to construct efficient tree topologies with the goal of maximizing the minimum throughput of the wireless network. Experimental results validate the scalability and performance gains of the proposed approaches, highlighting their potential for real-world applications.


  • Resilient Topology Design for Wireless Backhaul: A Deep Reinforcement Learning Approach

    by Ahmed Abdelmoaty, Diala Naboulsi, Ghassan Dahman, Gwenael Poitau, Francois Gagnon

    Ultra-dense 5G and beyond deployments are setting significant burden on cellular networks, especially for wireless backhauls. Today, a careful planning for wireless backhaul is more critical than ever. In this letter, we study the hierarchical wireless backhaul topology design problem. We introduce a Deep Reinforcement Learning (DRL) based algorithm that can solve the problem efficiently. We compare the quality of the solutions derived by our DRL approach to the optimal solution, derived according to the Integer Linear Program (ILP) formulation in our previous work. A simulation using practical channel propagation scenarios and different network densities proves that our DRL-based algorithm is providing a sub-optimal solution with different levels of resiliency. Our DRL algorithm is further shown to scale for larger instances of the problem.


  • Open Access

    On The Design of Resilient and Reliable Wireless Backhaul Networks

    by Ahmed Abdelmoaty, Diala Naboulsi, Ghassan Dahman, Gwenael Poitau, Francois Gagnon

    The exponential growth of traffic in mobile networks is leading to an increased pressure on the infrastructure of mobile networks and in particular on their backhaul networks. It is more critical than ever to carefully plan backhaul networks. In this paper, we formulate and solve the problem of hierarchical wireless backhaul network design. In our problem, we cover different requirements, namely: topology simplicity, network resiliency and link reliability. We formulate the problem as an Integer Linear Programming (ILP) problem, allowing us to solve the problem to optimality. Furthermore, we provide a graph theory based algorithm that allows to solve the problem overlarge scale. The proposed algorithm exploits the properties of the graph representing the network. The results of our evaluations in various network scenarios demonstrate the efficiency of our ILP formulation and the provided algorithm in keeping the backhaul network simple, resilient and reliable. Using a practical channel propagation model and different node densities that are representative of small-scale and large-scale urban environments, our results also show that even with high resiliency requirements, the network traffic can be backhauled with only5−10%of the nodes for the considered densities. Our results also demonstrate that our algorithm leads to near-optimal solutions in different scenarios.

  • Hyperparameter Free MEE-FP Based Localization

    by Rangeet Mitra, Georges Kaddoum, Ghassan Dahman, Gwenael Poitau

    In the context of outdoor localization over systems impaired by non-line of sight (NLoS), the minimum error entropy with fiducial points (MEE-FP) based methods have emerged as promising due to their excellent generalization and independence to statistics of NLoS. However, the performance of these approaches are well-known to depend on hyperparameters, such as, the spread parameter of the MEE-FP criterion. To enable MEE-FP based hyperparameter-free localization, we propose a modified Gauss-Newton based localization algorithm based on sampled kernel-widths. Next, analytical results are derived to demonstrate the asymptotic equivalence of the proposed kernel-width sampling based localization algorithm to its ideal fixed kernel width based counterpart. This equivalence is validated through computer simulations assuming typical non-Gaussian NLoS distributions, which motivates the hyperparameter-independence of the proposed localization algorithm and its generalization to different NLoS statistics.


  • Performance Analysis of Information Theoretic Learning-Based Cooperative Localization

    by Rangeet Mitra, Georges Kaddoum, Ghassan Dahman, Gwenael Poitau

    In the context of localization over ad-hoc wireless networks, the effect of non-line-of-sight (NLoS) presents a severe performance bottleneck that causes outages in the signal-to-noise ratio (SNR) of the returns and significantly increases the location estimates' error-floor. The non-Gaussianity and the scenario-dependent distributions/statistics of the additive noise-process caused by the NLoS returns make the naive ML-based solutions sub-optimal. In this regard, information-potential (IP) based localization algorithms have recently been found viable, especially in the presence of arbitrary additive noise-processes. This letter presents analytical insights on the error-bounds of two IP based algorithms in a cooperative-localization context, namely, maximum correntropy (MCC) and the minimum error entropy with fiducial points (MEE-FP). The presented analytical results for the considered IP based approaches are validated using relevant computer-simulations assuming typical point-processes.


  • Error Analysis of Localization Based on Minimum-Error Entropy With Fiducial Points

    by Rangeet Mitra, Georges Kaddoum, Ghassan Dahman, Gwenael Poitau

    Localization has been a well-investigated problem in the past decade for several networks and is one of the most celebrated applications of adaptive signal processing. However, existing localization algorithms exhibit a significant degradation under non-line of sight (NLOS) conditions, especially in outdoor scenarios. Due to the inherent non-Gaussianity in the NLOS returns, generic mitigation of the degradations due to NLOS remains quite an open challenge. In this regard, information-theoretic learning (ITL) criteria are attractive due to their ability to adapt to arbitrary NLOS distributions and suppress NLOS-induced non-Gaussian processes. In this regard, this letter proposes the use of minimum error entropy with Fiducial points (MEE-FP) in the particular context of round-trip time of arrival (RTTOA) based localization. From the presented simulations, it is observed that the proposed MEE-FP based localization method delivers lower variance under severe NLOS conditions and is closer to the ideal maximum-likelihood solution than contemporary ITL based approaches. Lastly, analytical variance-expressions are derived for the proposed localization technique, which is validated by computer simulations.


  • Open Access

    Using Vertically Separated MIMO in Ship-to-Ship Communications

    by Ahmed Abdelmoaty, Ghassan Dahman, Amen Allah Bousselmi, Gwenael Poitau, Francois Gagnon

    It is well established that the performance of communication systems is improved when deploying multiple antennas at one or both of the transmit/receive sides; and the amount of the gained improvement depends on the level of multipath richness of the propagation channel. In ship-to-ship overwater channels, quantifying such an improvement is not an easy task due to the effect of evaporation duct on imposing complex range- and height-dependent patterns of the received signal level. In this study, based on the parabolic equations (PE) method and using realistic evaporation duct distributions, we conduct extensive simulations in order to quantify the link-level improvement achieved when using 2 × 2 vertically-spaced Multiple-Input Multiple-Output (MIMO) systems against 1 × 2 Single-Input Multiple- Output (SIMO) and Single-Input Single-Output (SISO) systems. Then, we analyze the implication of such link improvement on the performance of a system comprised of a network of randomly distributed ships. When evaluating the outage throughput at the 2nd percentile using realistic system parameters, it was found that 2×2 MIMO-MRC (maximum ratio combining) systems with 1 m antenna spacing are able to improve the outage throughput by three-fold compared to SISO systems. This improvement increases to one order of magnitude when the antenna spacing increases to 10 m. It was also found that, in all cases, assuming using the same vertical spacing, 1 × 2 SIMO-MRC systems capture about 60% of the improvement achieved by 2 × 2 MIMO-MRC systems. On the other hand, 1 × 2 SIMO-DIV (diversity combining) systems are very sensitive to antenna spacing, and when assuming using the same vertical spacing, they can capture from 20% and up to 55% of the improvement achieved by 2 × 2 MIMO-MRC systems if the antenna spacing increases from 1 m to 10 m, respectively.


  • Open Access

    Average Vector-Symbol Error Rate Closed-Form Expression for ML Group Detection Receivers in Large MU-MIMO Channels With Transmit Correlation

    by Byron P. Maza, Ghassan Dahman, Georges Kaddoum, FranC’ois Gagnon

    In this paper, we derive a closed-form expression for the evaluation of the average vector-symbol error rate (VER) of group detection followed by maximum-likelihood (GD-ML) receivers in large multi-user multiple-input multiple-output (MU-MIMO) systems with transmit side correlated Rayleigh channels. We assume M antennas at the base station (BS), N closely-located, single-antenna user equipment (UEs) with load factor λ=N/M, and N ≫1 consequently, we evaluate the performance of GD-ML receivers as the load factor grows to unity. The derived expression requires a negligible correlation at the receive side of the communication channel. Hence, from a practical point-of-view, when considering scenarios with a large number of UEs, the derived analytical expression is generally more applicable for systems with a distributed massive number of BS antennas. Numerical results are provided to validate our derived expression. We observe that the GD-ML with Nu group size achieves a diversity order proportional to M - N + Nu. Moreover, we show that for small group sizes, the analytical and simulation results remain close, and at moderate to high signal-to-noise ratio (SNR), the derived expression very closely matches the simulations, whereas this match becomes perfect as the users’ side correlation increases. We also demonstrate that GD-ML outperforms the zero-forcing (ZF) and minimum mean-squared error (MMSE) receivers, in terms of VER; where for high λ, GD-ML exploits the maximum spatial multiplexing gain. Moreover, in terms of floating-point operations (FLOPs), we show that GD-ML receivers have almost the same complexity as ZF and MMSE where the ML detection stage adds a negligible complexity compared to the channel matrix inversion operation.


  • Extending the ITU-R P.530 Deep-Fading Outage Probability Results to SIMO-MRC and MIMO-MRC Line-of-Sight Systems

    by Francois Gagnon, Ghassan Dahman, Gwenael Poitau

    In this letter, we propose a method extending the deep-fading outage probability results of the ITU-R P.530 to the single-input multiple-output (SIMO)-maximum ratio-combining (MRC) and MIMO-MRC cases. By deducing the similarity between the analytical expressions of correlated Rayleigh fading and the ITU-R P.530 formulations, we utilize the ITU-R P.530 correlation information in order to estimate the outage probability of SIMO-MRC and MIMO-MRC systems. Finally, we demonstrate the significant reduction (of several orders of magnitude) in the worst month deep-fading outage probability that can be achieved using SIMO-MRC and MIMO-MRC for different fading depths and antenna spacings.


  • Open Access

    Spatial Separation of Closely-Located Users in Measured Massive MIMO Channels

    by Jose Flordelis, Fredrik Rusek, Xiang Gao, Ghassan Dahman, Ove Edfors, Fredrik Tufvesson

    We investigate the ability of Massive multiple-input multiple-output (MIMO) systems to spatially separate up to eighteen users located close to one another in line-of-sight (LOS) propagation conditions, in both indoor and outdoor environments. For that, we use fully-synchronous measured channels at 2.6 GHz of single-antenna users moving within a small area and concurrently communicating with a base station (BS) equipped with a compact 128-port array. To quantify the degree of spatial user separability, we use three scalar metrics, namely the achievable sum-rates, the condition number of the channel matrix, and the angle to interference factor. Our results show that Massive MIMO with zero-forcing (ZF) or regularized ZF (RZF) can spatially separate nine, even eighteen, concurrent users at practical SNR values even in the challenging case of dominant LOS propagation. In particular, signal-to-noise ratio losses relative to ideal (non-interfering and equally strong) channels can be reduced dramatically compared to standard multiuser MIMO systems, which typically have the same number of users as BS antennas. Our findings suggest that with RZF or ZF the ratio of BS antennas to number of served users should be at least three to four, to harvest most of the available spatial gains that the environment can offer. Although orthogonality and array gains complement each other, for the suggested ratios of antennas to users, the main contribution to improving system performance, measured in sum-rates, comes from the orthogonality gain.