Optimizing Cognitive Networks: Reinforcement Learning Meets Energy Harvesting Over Cascaded Channels

This paper presents a reinforcement learning (RL) based approach to improve the physical layer security (PLS) of an underlay cognitive radio network (CRN) over cascaded channels. These channels are utilized in highly mobile networks such as cognitive vehicular networks (CVN). In addition, an eavesdropper aims to intercept the communications between secondary users (SUs). The SU receiver has full-duplex and energy harvesting capabilities to generate jamming signals to confound the eavesdropper and enhance security. Moreover, the SU transmitter extracts energy from ambient radio frequency signals in order to power subsequent transmissions to its intended receiver. To optimize the privacy and reliability of the SUs in a CVN, a deep Q-network (DQN) strategy is utilized where multiple DQN agents are required such that an agent is assigned at each SU transmitter. The objective for the SUs is to determine the optimal transmission power and decide whether to collect energy or transmit messages during each time period in order to maximize their secrecy rate. Thereafter, we propose a DQN approach to maximize the throughput of the SUs while respecting the interference threshold acceptable at the receiver of the primary user. According to our findings, our strategy outperforms two other baseline strategies in terms of security and reliability.

Maximizing Reliability in Overlay Radio Networks with Time Switching and Power Splitting Energy Harvesting

Cognitive radio networks (CRNs) are acknowledged for their ability to tackle the issue of spectrum under-utilization. In the realm of CRNs, this paper investigates the energy efficiency issue and addresses the critical challenge of optimizing system reliability for overlay CRN access mode. Randomly dispersed secondary users (SUs) serving as relays for primary users (PUs) are considered, in which one of these relays is designated to harvest energy through the time switching-energy harvesting (EH) protocol. Moreover, this relay amplifies-and-forwards (AF) the PU's messages and broadcasts them along with its own across cascaded $\kappa$-$\mu$ fading channels. The power splitting protocol is another EH approach utilized by the SU and PU receivers to enhance the amount of energy in their storage devices. In addition, the SU transmitters and the SU receiver are deployed with multiple antennas for reception and apply the maximal ratio combining approach. The outage probability is utilized to assess both networks' reliability. Then, an energy efficiency evaluation is performed to determine the effectiveness of EH on the system. Finally, an optimization problem is provided with the goal of maximizing the data rate of the SUs by optimizing the time switching and the power allocation parameters of the SU relay.

Federated Learning-based MARL for Strengthening Physical-Layer Security in B5G Networks

This paper explores the application of a federated learning-based multi-agent reinforcement learning (MARL) strategy to enhance physical-layer security (PLS) in a multi-cellular network within the context of beyond 5G networks. At each cell, a base station (BS) operates as a deep reinforcement learning (DRL) agent that interacts with the surrounding environment to maximize the secrecy rate of legitimate users in the presence of an eavesdropper. This eavesdropper attempts to intercept the confidential information shared between the BS and its authorized users. The DRL agents are deemed to be federated since they only share their network parameters with a central server and not the private data of their legitimate users. Two DRL approaches, deep Q-network (DQN) and Reinforce deep policy gradient (RDPG), are explored and compared. The results demonstrate that RDPG converges more rapidly than DQN. In addition, we demonstrate that the proposed method outperforms the distributed DRL approach. Furthermore, the outcomes illustrate the trade-off between security and complexity.

Performance Optimization of Energy-Harvesting Underlay Cognitive Radio Networks Using Reinforcement Learning

In this paper, a reinforcement learning technique is employed to maximize the performance of a cognitive radio network (CRN). In the presence of primary users (PUs), it is presumed that two secondary users (SUs) access the licensed band within underlay mode. In addition, the SU transmitter is assumed to be an energy-constrained device that requires harvesting energy in order to transmit signals to their intended destination. Therefore, we propose that there are two main sources of energy; the interference of PUs' transmissions and ambient radio frequency (RF) sources. The SU will select whether to gather energy from PUs or only from ambient sources based on a predetermined threshold. The process of energy harvesting from the PUs' messages is accomplished via the time switching approach. In addition, based on a deep Q-network (DQN) approach, the SU transmitter determines whether to collect energy or transmit messages during each time slot as well as selects the suitable transmission power in order to maximize its average data rate. Our approach outperforms a baseline strategy and converges, as shown by our findings.

Fast Grant Learning-Based Approach for Machine Type Communications with NOMA

In this paper, we propose a non-orthogonal multiple access (NOMA)-based communication framework that allows machine type devices (MTDs) to access the network while avoiding congestion. The proposed technique is a 2-step mechanism that first employs fast uplink grant to schedule the devices without sending a request to the base station (BS). Secondly, NOMA pairing is employed in a distributed manner to reduce signaling overhead. Due to the limited capability of information gathering at the BS in massive scenarios, learning techniques are best fit for such problems. Therefore, multi-arm bandit learning is adopted to schedule the fast grant MTDs. Then, constrained random NOMA pairing is proposed that assists in decoupling the two main challenges of fast uplink grant schemes namely, active set prediction and optimal scheduling. Using NOMA, we were able to significantly reduce the resource wastage due to prediction errors. Additionally, the results show that the proposed scheme can easily attain the impractical optimal OMA performance, in terms of the achievable rewards, at an affordable complexity.