A Survey on Machine Learning-Based Performance Improvement of Wireless Networks: PHY, MAC and Network Layer

Kulin, Merima; Kazaz, Tarik; Poorter, Eli De; Moerman, Ingrid

doi:10.3390/electronics10030318

Cited by 67 publications

(57 citation statements)

References 261 publications

(291 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition, hyperparameters also decide the efficiency and accuracy of model training via parameters such as learning rate (LR) of the gradient descent algorithm, activation function, regularization factors, or types of optimization algorithms [ 28 ]. Hyperparameters tuning and fining optimal configurations is challenging and time-consuming in a deep learning model [ 33 ]. Most widely used methods for hyperparameter selection are based on experience in training deep learning models; however, it is lack of logical reasoning and difficult to find an optimal set of hyperparameters for a DNN model with a given dataset.…”

Section: Proposed Modelmentioning

confidence: 99%

A Deep Neural Network-Based Multi-Frequency Path Loss Prediction Model from 0.8 GHz to 70 GHz

Nguyen

Cheema

2021

Sensors

View full text Add to dashboard Cite

Large-scale fading models play an important role in estimating radio coverage, optimizing base station deployments and characterizing the radio environment to quantify the performance of wireless networks. In recent times, multi-frequency path loss models are attracting much interest due to their expected support for both sub-6 GHz and higher frequency bands in future wireless networks. Traditionally, linear multi-frequency path loss models like the ABG model have been considered, however such models lack accuracy. The path loss model based on a deep learning approach is an alternative method to traditional linear path loss models to overcome the time-consuming path loss parameters predictions based on the large dataset at new frequencies and new scenarios. In this paper, we proposed a feed-forward deep neural network (DNN) model to predict path loss of 13 different frequencies from 0.8 GHz to 70 GHz simultaneously in an urban and suburban environment in a non-line-of-sight (NLOS) scenario. We investigated a broad range of possible values for hyperparameters to search for the best set of ones to obtain the optimal architecture of the proposed DNN model. The results show that the proposed DNN-based path loss model improved mean square error (MSE) by about 6 dB and achieved higher prediction accuracy R2 compared to the multi-frequency ABG path loss model. The paper applies the XGBoost algorithm to evaluate the importance of the features for the proposed model and the related impact on the path loss prediction. In addition, the effect of hyperparameters, including activation function, number of hidden neurons in each layer, optimization algorithm, regularization factor, batch size, learning rate, and momentum, on the performance of the proposed model in terms of prediction error and prediction accuracy are also investigated.

show abstract

Section: Proposed Modelmentioning

confidence: 99%

A Deep Neural Network-Based Multi-Frequency Path Loss Prediction Model from 0.8 GHz to 70 GHz

Nguyen

Cheema

2021

Sensors

View full text Add to dashboard Cite

show abstract

“…Some discussed techniques are divided according to the problems found in each layer of the protocol stack. A more detailed discussion and survey about these techniques, however, is found in [ 17 ]. As an example of such techniques, [ 18 ] proposed a fast machine learning algorithm by modeling the problem as a contextual multi-armed bandit one.…”

Section: Ai and Recent Ra Enhancements In Literaturementioning

confidence: 99%

Random Access Using Deep Reinforcement Learning in Dense Mobile Networks

Bekele

Choi

2021

Sensors

View full text Add to dashboard Cite

5G and Beyond 5G mobile networks use several high-frequency spectrum bands such as the millimeter-wave (mmWave) bands to alleviate the problem of bandwidth scarcity. However high-frequency bands do not cover larger distances. The coverage problem is addressed by using a heterogeneous network which comprises numerous small and macrocells, defined by transmission and reception points (TRxPs). For such a network, random access is considered a challenging function in which users attempt to select an efficient TRxP by random access within a given time. Ideally, an efficient TRxP is less congested, minimizing delays in users’ random access. However, owing to the nature of random access, it is not feasible to deploy a centralized controller estimating the congestion level of each cell and deliver this information back to users during random access. To solve this problem, we establish an optimization problem and employ a reinforcement-learning-based scheme. The proposed scheme estimates congestion of TRxPs in service and selects the optimal access point. Mathematically, this approach is beneficial in approximating and minimizing a random access delay function. Through simulation, we demonstrate that our proposed deep learning-based algorithm improves performance on random access. Notably, the average access delay is improved by 58.89% from the original 3GPP algorithm, and the probability of successful access also improved.

show abstract

“…With ML techniques, it is possible to autonomously learn the main features of each wireless channel from raw data as well as to develop models that are smarter and self-adaptive in the sense that they adjust to the wireless channel they are facing. In [7] [8] the authors provide a comprehensive survey on the latest research efforts in this field. The two most relevant conclusions are: 1) most works use simplistic simulation data that does not reflect the complexity of real world scenarios; 2) most practical implementation of ML models in wireless platforms are not realistic, because the existing computing facilities in network systems are not designed to support the training overhead, due to the need of collecting and transferring large amounts of data.…”

Section: Introductionmentioning

confidence: 99%

Wi-Fi Rate Adaptation using a Simple Deep Reinforcement Learning Approach

Ruben¹,

Almeida²,

Fontes³

et al. 2022

Preprint

View full text Add to dashboard Cite

The increasing complexity of recent Wi-Fi amendments is making optimal Rate Adaptation (RA) a challenge. The use of classic algorithms or heuristic models to address RA is becoming unfeasible due to the large combination of configuration parameters along with the variability of the wireless channel. Machine Learning-based solutions have been proposed in the state of art, to deal with this complexity. However, they typically use complex models and their implementation in real scenarios is difficult.We propose a simple Deep Reinforcement Learning approach for the automatic RA in Wi-Fi networks, named Data-driven Algorithm for Rate Adaptation (DARA). DARA is standardcompliant. It dynamically adjusts the Wi-Fi Modulation and Coding Scheme (MCS) solely based on the observation of the Signal-to-Noise Ratio (SNR) of the received frames at the transmitter. Our simulation results show that DARA achieves up to 15% higher throughput when compared with Minstrel High Throughput (HT) and equals the performance of the Ideal Wi-Fi RA algorithm.

show abstract

A Survey on Machine Learning-Based Performance Improvement of Wireless Networks: PHY, MAC and Network Layer

Cited by 67 publications

References 261 publications

A Deep Neural Network-Based Multi-Frequency Path Loss Prediction Model from 0.8 GHz to 70 GHz

A Deep Neural Network-Based Multi-Frequency Path Loss Prediction Model from 0.8 GHz to 70 GHz

Random Access Using Deep Reinforcement Learning in Dense Mobile Networks

Wi-Fi Rate Adaptation using a Simple Deep Reinforcement Learning Approach

Contact Info

Product

Resources

About