Open Radio Access Network (O-RAN) provides the capability to efficiently distribute the RAN Network Functions (NFs) such as the Radio Unit (RU), Distributed Unit (DU) and Centralized Unit (CU) in O-RAN Cloud (O-Cloud) nodes using virtualization, automation, intelligence and open interface specifications. In addition, Carrier Aggregation (CA) technology enhances the throughput of the users by aggregating Component Carriers (CCs) and allocating one Primary Cell (PCell) and multiple Secondary Cells (SCell) to each user. Finding the DU NFs placement of each CC (PCell or SCell) while minimizing the number of used O-Cloud nodes and the average user end to end delay is our aim in this paper. Thus, we model the average delay and propose an algorithm using Deep Q Network (DQN) based Deep Reinforcement Learning (DRL) algorithm to find the solution to the problem. Simulation results demonstrate that our proposed scheme reduces the average end user delay and the number of employed O-Cloud nodes at least 90% and 20% with respect to the baselines.

In this work, we propose globally optimal power allocation strategies to maximize the users sum-rate (SR), and system energy efficiency (EE) in the downlink of single-cell multicarrier non-orthogonal multiple access (MC-NOMA) systems. Each NOMA cluster includes a set of users in which the well-known superposition coding (SC) combined with successive interference cancellation (SIC) technique is applied among them. By obtaining the closed-form expression of intra-cluster power allocation, we show that MC-NOMA can be equivalently transformed to a virtual orthogonal multiple access (OMA) system, where the effective channel gain of these virtual OMA users is obtained in closed-form. Then, the SR and EE maximization problems are solved by using very fast water-filling and Dinkelbach algorithms, respectively. The equivalent transformation of MC-NOMA to the virtual OMA system brings new theoretical insights, which are discussed throughout the paper. The extensions of our analyses to other scenarios, such as considering users rate fairness, admission control, long-term performance, and a number of future next-generation multiple access (NGMA) schemes enabling recent advanced technologies, e.g., reconfigurable intelligent surfaces are discussed. Extensive numerical results are provided to demonstrate the performance gaps among single-carrier NOMA (SC-NOMA), OMA-NOMA, and OMA.

As one of the key technologies in 5G networks, Carrier Aggregation (CA) is studied in this paper. In CA, Component Carriers (CCs) can be activated and deactivated depending on multiple factors, e.g., energy consumption and Quality of Service (QoS) demand of users. We propose CC management strategies where each User Equipment (UE) minimizes its average delay and at the same time minimizes its power consumption while considering that CCs can be activated and deactivated only at certain times, as in real-world CA implementations. We first model the problem as a centralized multi-objective optimum CC management problem. Since centralized approaches would impose a large overhead on the system, we then develop a semi-distributed solution by modeling the problem as a stochastic game and propose a multi-agent Double Deep Q-Network (DDQN) based CC management algorithm to solve the stochastic game. We finally compare the proposed approaches with single CC activation and all-CC activation baseline schemes. Simulation results show that our proposed algorithms outperform the all-CC algorithm in terms of UE power consumption and have the capability of transmitting a number of bits with delay close to the all-CC scheme. Meanwhile, our DDQN-based algorithm decreases the UE power consumption by about 20% with respect to the all-CC scheme.

Open Radio Access Networks (O-RAN) architecture is based on disaggregation, virtualization, openness, and intelligence. These features allow the RAN network functions (NFs) to be split into Central Unit (CU), Distributed Unit (DU), and Radio Unit (RU); and deployed on open hardware and cloud nodes as Virtualized Network Functions (VNFs) or Containerized Network Functions (CNFs). In this paper, we propose strategies for the placement of CU and DU network functions in the regional and edge O-Cloud nodes while jointly associating the users to RUs. The aim is to minimize the end-to-end delay of users and minimize the cost of O-RAN deployment. Thus, we first formulate the end-to-end delay, the cost, and the constraints. We then model the problem as a multi-objective optimization problem The optimization formulation consists of a huge number of constraints and variables. To provide a solution to the problem, we develop the corresponding Markov Decision Problem (MDP) and propose a Deep Q-Network (DQN)-based algorithm. The simulation results demonstrate that our proposed scheme reduces the average user delay up to 40% and the deployment cost up to 20% with respect to our baselines.

Carrier aggregation (CA) is considered a key enabler technology for delivering higher rates to users of LTE and 5G networks. However, the increased transmission rate comes with the price of higher energy consumption which stems from users continuously monitoring the control channel of the active component carriers (CCs) whether data transmission is ongoing or not. In order to reduce energy consumption, we exploit the activation-deactivation procedure at the medium access control (MAC) layer of LTE/5G network. In this paper, we propose a reinforcement learning-based algorithm to improve energy-efficiency by dynamically activating-deactivating secondary component carriers (SCCs) with awareness of the user traffic profiles. The proposed algorithm aims to predict the arrival of data and identify SCCs to activate for each user. In addition, a traffic splitting approach and an intelligent exploration strategy are proposed to balance users' load among CCs and improve the convergence of the algorithm, respectively. Results of the proposed algorithm are compared with three baseline algorithms. The first baseline always activates all CCs for each user, the second baseline activates one carrier only (i.e., the primary carrier) and the third baseline algorithm relies on a reactive method, where the activation-deactivation decision is performed after observing the arrival of data. Results show that Q-learning outperforms the baseline algorithms by achieving the highest sum throughput (and lowest average delay) with the lowest number of activated SCCs, which is obtained by learning to dynamically activate SCCs according to the traffic pattern. Hence, Q-learning is considered the most energy-efficient compared to the baseline algorithms.

Carrier Aggregation (CA) has been a breakthrough in LTE that led to increased throughput for users, and is still one of the key technologies in 5G that helps to enhance spectrum utilization. In CA, Component Carriers (CCs) are dynamically activated and deactivated depending on several performance factors. Optimal selection of CCs has been studied in the literature. However, the latency associated with activation and deactivation of CCs, control channel overhead for switching CCs, as well as the energy consumed for monitoring the active CCs have not been a part of the optimal CC selection problem. Nevertheless, those become stringent design constraints in practice. In this paper, we address optimal CC selection and resource allocation in 5G networks, where the above constraints are considered and the 5G network supports several service types with different 5G QoS Identifiers (5QI). The proposed optimum joint CC selection and Radio Resource Block (RB) allocation schemes maximize average throughput of users and satisfy QoS of users in terms of delay. In addition, the proposed schemes take CC activation and deactivation burden into consideration and aim to minimize the number of activations and deactivations. The simulation results demonstrate that our proposed solution outperforms the state of the art solution while satisfying the QoS requirements and creating close to 95.5% reduction on the number of CCs activations and deactivations.

In delay-sensitive industrial Internet of Things (IIoT) applications, the age of information (AoI) is employed to characterize the freshness of information. Meanwhile, the emerging network function virtualization provides flexibility and agility for service providers to deliver a given network service using a sequence of virtual network functions (VNFs). However, suitable VNF placement and scheduling in these schemes is NP-hard and finding a globally optimal solution by traditional approaches is complex. Recently, deep reinforcement learning (DRL) has appeared as a viable way to solve such problems. In this paper, we first utilize single agent low-complex compound action actor-critic RL to cover both discrete and continuous actions and jointly minimize VNF cost and AoI in terms of network resources under end-to-end Quality of Service constraints. To surmount the single-agent capacity limitation for learning, we then extend our solution to a multi-agent DRL scheme in which agents collaborate with each other. Simulation results demonstrate that single-agent schemes significantly outperform the greedy algorithm in terms of average network cost and AoI. Moreover, multi-agent solution decreases the average cost by dividing the tasks between the agents. However, it needs more iterations to be learned due to the requirement on the agents' collaboration.

In this paper, the performance of fifth generation (5G) cellular networks under two promising technologies, namely dual connectivity and coordinated multi-point transmission (CoMP) is investigated. In this regard, a joint access and fronthaul radio resource allocation for a two tier downlink heterogeneous cloud radio access network (H-CRAN) is formulated. The main aim of the proposed problem formulation is to maximize the system energy efficiency (EE) by using both millimeter wave (mmW) and micro wave (μW) links in access and fronthaul. The proposed optimization is a mixed integer nonconvex problem with a high computational complexity solution. Therefore, existing convex optimization methods can not be used directly to solve this problem. To solve this issue, an iterative algorithm based on a successive convex approximation (SCA) approach with low complexity is exploited. As the numerical results show, via dual connectivity and CoMP EE is improved by approximately 90% compared to using only μW subcarriers.

In this paper, fifth-generation (5G) cellular networks under three promising technologies, namely, dual connectivity, coordinated multi-point transmission (CoMP), and power domain non orthogonal multiple access (PD-NOMA) are investigated. The main aim is to maximize the downlink energy efficiency (EE) by using both millimeter wave (mmW) and micro wave (μW) links in access and fronthaul, while employing CoMP and PD-NOMA. In this regard, joint access and fronthaul radio resource allocation for a downlink heterogeneous cloud radio access network is considered. The proposed optimization is a mixed integer non-convex problem with a high computational complexity solution, and hence, the alternate search method based on a successive convex approximation approach using fractional programming is exploited. Furthermore, the convergence of the proposed iterative resource allocation method is proved and its computational complexity is investigated. As the numerical results show, via dual connectivity through receiving signals from both mmW and μW transmitters, the system EE is improved by approximately 50%, in contrast to using only μW subcarriers (e.g., as in local thermal equilibrium). In addition, by applying both PD-NOMA and CoMP technologies on the μW subcarriers, the EE of the system increases by approximately 45%.

In this paper, decentralized heuristic access policies are designed for two secondary users (SUs) in an underlay cognitive radio network in which the PU employs Type-I Hybrid ARQ. Exploiting the redundancy in PU retransmissions, each SU receiver applies interference cancellation (IC) to remove a successfully decoded PU message in the subsequent PU retransmissions. There is no central control unit and we consider two different scenarios. In the first scenario, each SU transmitter only knows the PU message knowledge state of its corresponding SU receiver. In the second scenario, the SU transmitters are unaware of any PU message knowledge states at the SU receivers. We design access policies in both offline and online modes, where in an online mode each SU can learn from its own local information. Using heuristic approaches for the considered scenarios, decentralized access policies are proposed to maximize the average sum throughput of SUs under a PU throughput degradation constraint. The results also show that heuristic access policies using the online mode have a performance gain close to that obtained using the offline mode, while providing increased robustness and flexibility.

This paper investigates distortion-power tradeoffs in transmission of quasi-stationary sources over delay and buffer limited block fading channels by studying encoder and decoder buffering techniques to smooth out the source and channel variations. Four source and channel coding schemes that consider buffer and power constraints are presented to minimize the reconstructed source distortion. The first one is a high performance scheme, which benefits from optimized source and channel rate adaptation. In the second scheme, the channel coding rate is fixed and optimized along with transmission power with respect to channel and source variations; hence this scheme enjoys simplicity of implementation. The two last schemes have fixed transmission power with optimized adaptive or fixed channel coding rate. For all the proposed schemes, closed form solutions for mean distortion, optimized rate, and power are provided and in the high SNR regime, the mean distortion exponent and the asymptotic mean power gains are derived. The proposed schemes with buffering exploit the diversity due to source and channel variations. Specifically, when the buffer size is limited, fixed channel rate adaptive power scheme outperforms an adaptive rate fixed power scheme. Furthermore, analytical and numerical results demonstrate that with limited buffer size, the system performance in terms of reconstructed signal SNR saturates as transmission power increases, suggesting that appropriate buffer size selection is important to achieve a desired reconstruction quality.

In this paper, an underlay cognitive radio network that consists of an arbitrary number of secondary users (SU) is considered in which the primary user (PU) employs type-I hybrid automatic repeat request (HARQ). Exploiting the redundancy in PU retransmissions, each SU receiver applies forward interference cancelation to remove a successfully decoded PU message in the subsequent PU retransmissions. The knowledge of the PU message state at the SU receivers and the ACK/NACK message from the PU receiver are sent back to the transmitters. With this approach and using a constrained Markov decision process (CMDP) model and constrained multi-agent MDP (CMMDP), centralized and decentralized optimum access policies for SUs are proposed to maximize their average sum throughput under a PU throughput constraint. In the decentralized case, the channel access decision of each SU is unknown to the other SU. Numerical results demonstrate the benefits of the proposed policies in terms of sum throughput of SUs. The results also reveal that the centralized access policy design outperforms the decentralized design especially when the PU can tolerate a low average long term throughput. Finally, the difficulties in decentralized access policy design with partial state information are discussed.

In this paper, delay-limited transmission of quasi-stationary sources over block fading channels is considered. Considering distortion outage probability as the performance measure, two source and channel coding schemes with power adaptive transmission are presented. The first one is optimized for fixed rate transmission, and hence enjoys simplicity of implementation. The second one is a high performance scheme, which also benefits from optimized rate adaptation with respect to source and channel states. In high SNR regime, the performance scaling laws in terms of outage distortion exponent and asymptotic outage distortion gain are derived, where two schemes with fixed transmission power and adaptive or optimized fixed rates are considered as benchmarks for comparisons. Various analytical and numerical results are provided which demonstrate a superior performance for source and channel optimized rate and power adaptive scheme. It is also observed that from a distortion outage perspective, the fixed rate adaptive power scheme substantially outperforms an adaptive rate fixed power scheme for delay-limited transmission of quasi-stationary sources over wireless block fading channels. The effect of the characteristics of the quasi-stationary source on performance, and the implication of the results for transmission of stationary sources are also investigated.

This paper addresses network code design for robust transmission of sources over an orthogonal two-hop wireless network with a broadcasting relay. The network consists of multiple sources and destinations in which each destination, benefiting the relay signal, intends to decode a subset of the sources. Two special instances of this network are orthogonal broadcast relay channel and the orthogonal multiple access relay channel. The focus is on complexity constrained scenarios, e.g., for wireless sensor networks, where channel coding is practically imperfect. Taking a source-channel and network coding approach, we design the network code (mapping) at the relay such that the average reconstruction distortion at the destinations is minimized. To this end, by decomposing the distortion into its components, an efficient design algorithm is proposed. The resulting network code is nonlinear and substantially outperforms the best performing linear network code. A motivating formulation of a family of structured nonlinear network codes is also presented. Numerical results and comparison with linear network coding at the relay and the corresponding distortion-power bound demonstrate the effectiveness of the proposed schemes and a promising research direction.