Deep Neural Network Based Resource Allocation for V2X Communications

Abstract

This paper focuses on optimal transmit power allocation to maximize the overall system throughput in a vehicleto-everything (V2X) communication system. We propose two methods for solving the power allocation problem namely the weighted minimum mean square error (WMMSE) algorithm and the deep learning-based method. In the WMMSE algorithm, we solve the problem using block coordinate descent (BCD) method. Then we adopt supervised learning technique for the deep neural network (DNN) based approach considering the power allocation from the WMMSE algorithm as the target output. We exploit an efficient implementation of the mini-batch gradient descent algorithm for training the DNN. Extensive simulation results demonstrate that the DNN algorithm can provide very good approximation of the iterative WMMSE algorithm reducing the computational overhead significantly.Index Terms-Machine learning, deep learning, deep neural network, V2X, V2V, power control, resource allocation.

Role of Machine Learning in Resource Allocation Strategy over Vehicular Networks: A Survey

“…For example, the authors in [ 55 ] implemented logistic regression to determine the power fraction which can flexibly decrease BS/RSU transmit power according to the vehicle–RSU distance. Additionally, in [ 54 ], DNN was utilized to determine the optimal transmit power according to the channel realization and the channel gain of V2I and V2V links. Although training the algorithm requires a longer amount of time, the algorithm can provide a fast solution for a dynamic resource allocation decision.…”

“…By predicting resource availability, the system’s effectiveness is guaranteed by avoiding over-provisioning. J. Gao et al [ 54 ] implemented a DNN to approximate the weighted minimum square error (WMMSE) value by learning the mapping between the channel power gain as the input and the optimal power allocation as output in the V2V and V2I links. The results indicate that implementing DNN supervised learning improved the system performance compared to conventional supervised learning.…”

Low-Complexity Resource Allocation for Dense Cellular Vehicle-to-Everything (C-V2X) Communications

“…Thus, the spatial reuse gain of the cell was not utilized in [15]- [18]. Resource allocation for V2X communications using deep neural networks was addressed in [21], [22]. The authors of [21] considered a limited number of transmitting entities and only V2I links.…”

“…Resource allocation for V2X communications using deep neural networks was addressed in [21], [22]. The authors of [21] considered a limited number of transmitting entities and only V2I links. The authors of [22] focused on mobile edge computing in nondense environments.…”

Learning the Wireless V2I Channels Using Deep Neural Networks

“…In order to leverage the advantages of using a large group of data for communication performance improvement, several machine learning methods including supervised, unsupervised and reinforcement learning have been proposed based on the traditional approaches. The machine learning can be useful in analyzing communication environment variance, making decisions autonomously, transmission routing, network security, and system resource management [6], [8].…”