“…to measure the training performance of the secondary ANN, where m is an M × 1 vector produced by brute-force search algorithms by (4) or (5), in which the entries corresponding to the selected relays are ones, while the entries corresponding to the other relays are zeros.…”
Section: B Implementation Methodsmentioning
confidence: 99%
“…Because most communication nodes are expected to be equipped with channel estimation functionality in next generation networks, the CSI represented by {G(m, k)} is assumed to be perfectly known. Also, by the relay selection criteria given in (4) and 5, we can also employ brute-force search algorithms to determine the optimal sets of relays M (•) , i.e. labels, corresponding to given {G(m, k)}.…”
Section: Supervised Machine Learning Via Cascaded Artificial Neuramentioning
confidence: 99%
“…For simplicity, we normalize the hyperparameters by α pr bulk = α se bulk = α pr ps = α se ps = 1 and β pr bulk = β se bulk = β pr ps = β se ps = 1. We generate 10 7 training data sets by {G 1 (m, k)}, {G 2 (m, k)} ∼ exprnd(1) via the embedded exponentially distributed random number function on MATLAB as well as their labels by the exhaustive search according to (4) or (5). To mitigate the randomness of channel realizations and reveal the statistical nature of the proposed CANN based relay selection enabling technique, we average all simulation results over all previous trials under the identical simulation configurations for illustration.…”
Section: A Simulation Configurationsmentioning
confidence: 99%
“…To fulfill the increasingly high requirements of data rate, energy efficiency, coverage and reliability, cooperative relaying and relay-assisted communications have attracted many researchers' attention [2]. In order to further enhance the spectral efficiency of cooperative relaying and utilize the existing infrastructures in fourth generation (4G) wireless communication networks, multi-carrier transmission has been incorporated in cooperative relaying, which results in multi-carrier cooperative systems [3], [4]. This conception is believed to play a key role in even sixth generation (6G) networks [5].…”
Cooperative relaying has been adopted as one of the most important techniques to enhance the energy efficiency and coverage. Multi-carrier relay selection is an efficient method to allocate spatial/spectral resources in cooperative relay networks and provides diversity gain. However, the implementation of multicarrier relay selection is not straightforward, and could render the high system complexity (for centralized implementation schemes) or long processing delay (for distributed implementation schemes). These issues hinder the promotion and implementation of multicarrier relay selection for intelligent vehicular communications. To mitigate aforementioned issues, we propose an enabling technique of multi-carrier relay selection based on sensing fusion (SF) and cascaded artificial neural networks (CANNs) for intelligent vehicular communications. We employ two well-known multicarrier relay selection schemes, i.e. bulk and per-subcarrier relay selection, to verify the effectiveness of the CANN based enabling technique. With the powerful processing ability with intelligent vehicles, the numerical results illustrate a promising vision of applying CANNs to enable multi-carrier relay selection for fast deployment in intelligent vehicular communication networks.
“…to measure the training performance of the secondary ANN, where m is an M × 1 vector produced by brute-force search algorithms by (4) or (5), in which the entries corresponding to the selected relays are ones, while the entries corresponding to the other relays are zeros.…”
Section: B Implementation Methodsmentioning
confidence: 99%
“…Because most communication nodes are expected to be equipped with channel estimation functionality in next generation networks, the CSI represented by {G(m, k)} is assumed to be perfectly known. Also, by the relay selection criteria given in (4) and 5, we can also employ brute-force search algorithms to determine the optimal sets of relays M (•) , i.e. labels, corresponding to given {G(m, k)}.…”
Section: Supervised Machine Learning Via Cascaded Artificial Neuramentioning
confidence: 99%
“…For simplicity, we normalize the hyperparameters by α pr bulk = α se bulk = α pr ps = α se ps = 1 and β pr bulk = β se bulk = β pr ps = β se ps = 1. We generate 10 7 training data sets by {G 1 (m, k)}, {G 2 (m, k)} ∼ exprnd(1) via the embedded exponentially distributed random number function on MATLAB as well as their labels by the exhaustive search according to (4) or (5). To mitigate the randomness of channel realizations and reveal the statistical nature of the proposed CANN based relay selection enabling technique, we average all simulation results over all previous trials under the identical simulation configurations for illustration.…”
Section: A Simulation Configurationsmentioning
confidence: 99%
“…To fulfill the increasingly high requirements of data rate, energy efficiency, coverage and reliability, cooperative relaying and relay-assisted communications have attracted many researchers' attention [2]. In order to further enhance the spectral efficiency of cooperative relaying and utilize the existing infrastructures in fourth generation (4G) wireless communication networks, multi-carrier transmission has been incorporated in cooperative relaying, which results in multi-carrier cooperative systems [3], [4]. This conception is believed to play a key role in even sixth generation (6G) networks [5].…”
Cooperative relaying has been adopted as one of the most important techniques to enhance the energy efficiency and coverage. Multi-carrier relay selection is an efficient method to allocate spatial/spectral resources in cooperative relay networks and provides diversity gain. However, the implementation of multicarrier relay selection is not straightforward, and could render the high system complexity (for centralized implementation schemes) or long processing delay (for distributed implementation schemes). These issues hinder the promotion and implementation of multicarrier relay selection for intelligent vehicular communications. To mitigate aforementioned issues, we propose an enabling technique of multi-carrier relay selection based on sensing fusion (SF) and cascaded artificial neural networks (CANNs) for intelligent vehicular communications. We employ two well-known multicarrier relay selection schemes, i.e. bulk and per-subcarrier relay selection, to verify the effectiveness of the CANN based enabling technique. With the powerful processing ability with intelligent vehicles, the numerical results illustrate a promising vision of applying CANNs to enable multi-carrier relay selection for fast deployment in intelligent vehicular communication networks.
“…Due to the high EE and anti-interference property, IM technique has been widely used in cognitive radio (CR) networks [14], cooperative networks [15], [16], and multiple access (MA) schemes [12], [17], [19]. In [14], the inactive subcarriers of orthogonal frequency division multiplexing with index modulation (OFDM-IM) are utilized to transmit the secondary user's signals in CR networks. In [15], the mapping scenario between information bits and subcarrier activation patterns (SAPs) is adaptively chosen for OFDM-IM relay networks.…”
Index modulation multiple access (IM-MA) is recently proposed to exploit the IM concept to the uplink multiple access system, where multiple users transmit their own signals via the selected time slots. However, the computational complexity of the optimal maximum-likelihood (ML) detection in IM-MA is tremendously high when the number of users or time slots is large. In this letter, we propose a low-complexity detection method for IM-MA, which is inspired by the log likelihood ratio (LLR) algorithm. In addition, because of the heavy search burden for all LLR values, we further propose a suboptimal method to determine the permutation set, which records the number of users allocated to each time slot. Simulation results and the complexity analysis verify that the proposed detection performs closely to the optimal ML detection with reduced computational complexity.
In this article, we propose to use shuffled frog leaping algorithm (SFLA) in order to provide the quality of service (QoS) requested by the secondary user (SU) in orthogonal frequency division multiplexing-based cognitive radio (CR) systems. The performance of the SFLA was evaluated through three transmission modes (emergency, multimedia, and low battery). SFLA provided the best possible configuration in a multicarrier context and thus proved its effectiveness to satisfy the needed QoS for the SU in CR systems. The obtained simulation results showed that SFLA offers better performance compared with genetic algorithms.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.