Efficient resource allocation is important to realize ultra-reliability and low latency in the ultra-reliable and low-latency communication (URLLC) system with massive users. In this paper, resource allocation in a multi-user URLLC system is studied to improve the system reliability. A latency-constrained packet dropping rate (LCPDR) which jointly captures packet losses induced by transmission errors, timeout violations and buffer overflows during packet delivery, is used to comprehensively measure the end-to-end reliability. A joint optimization problem of subchannel, time slot and power allocation is formulated to minimize the LCPDR with the maximum delay constraint. To reduce complexity and resource consumption when there are a large number of users, user grouping is proposed and the problem is reformulated as two subproblems on two timescales, i.e., user grouping on a long-timescale and resource allocation within each group on a short-timescale. To solve the subproblems, a three-tier hierarchical deep reinforcement learning (HDRL) architecture is proposed, in which a novel event-triggered dynamic regrouping mechanism is embedded into the long-timescale user grouping and short-timescale intra-group resource allocation tiers to dynamically migrate abnormal users between groups. A distance-aware proximal policy optimization (PPO) algorithm named H-PPO-IL based on HDRL and imitation learning (IL) is proposed, in which IL is used to pretrain the PPO agents and accelerate the convergence. Simulation results show that the proposed algorithm has a lower LCPDR and a faster convergence speed, which is more effective compared with benchmark algorithms.

Intracerebral hemorrhage (ICH) segmentation is a critical step in the treatment of hemorrhagic stroke. Although many segmentation models have been developed for this task, their large parameter counts and high computational costs hinder deployment on mobile devices. To address this challenge, we propose a lightweight multi-axis interactive and mask-guided network for ICH segmentation, named DualMNet. DualMNet integrates a Multi-Axis Interactive Attention (MIA) module, which utilizes three-axis interactions to generate an attention map, extracting the inherent characteristics of CT images and enhancing feature representation with minimal computational cost. Furthermore, to overcome the difficulties of missed and false detections caused by low contrast between hemorrhagic and normal tissues, as well as the excessively small area of some bleeding regions, we design a Mask-Guided Feature Fusion (MGFF) module. This module uses intermediate mask predictions to guide the adaptive fusion of low-level and high-level features, making the model more sensitive to hemorrhagic regions. Experimental results on two public datasets, PHY and BHSD, demonstrate that DualMNet outperforms existing lightweight models in segmentation accuracy, achieving higher Dice scores while significantly reducing model parameters.

Abstract The orthogonal time frequency space (OTFS) modulation is a promising multicarrier waveform which can have good performance in high‐mobility scenarios. Among various receivers designed for the OTFS modulation, the message‐passing (MP) receiver has received much attention which can achieve a good bit error rate (BER) performance. However, the BER performance of MP receiver is usually obtained through simulations due to the lack of corresponding BER formula in the literature. This work studies the BER formula of OTFS modulation with MP receiver over doubly selective fading channel. Specifically, equivalent signal models of the MP receiver are first defined and the expression of the corresponding signal‐noise‐ratio (SNR) is derived, then the probability distribution function of the SNR is estimated by assuming the distribution of SNR as a lognormal distribution, then the methods to estimate the parameters of the distribution function are proposed, and finally the formula for calculating the BER of the OTFS modulation with MP receiver is obtained. Simulation results verify that the proposed BER formula is accurate.

Abstract This work studies the equalization methods for the reduced zero‐padding version of the orthogonal time frequency space modulation. Here, the focus is on applications in low data rate scenarios such as the Internet of Things, and the case where the number of subcarriers and time slots are small will be considered. First, the system equation is established to model the inter‐subcarrier and inter‐slot interference phenomena. Based on the system model, three sphere‐decoding based equalization methods are proposed. The first method is to directly apply the sphere‐decoding algorithm on the full‐scale system model; for the second method, the transmitter needs to perform the triangular precoding; and for the third method, the transmitter needs to perform the diagonal precoding. Simulation results show that the proposed sphere‐decoding based methods achieve better bit error rate performance than the state‐of‐the‐art methods like message passing, unitary approximate message passing, and variational approximate message passing, especially when the channel has severe inter‐slot interference.

New radio-based access to unlicensed bands (NR-U) multichannel access procedures are 3GPP standards for supporting 5G NR-U and WiFi coexistence in unlicensed frequency bands. NR-U provides a new solution to alleviating the severe spectrum scarcity problem of 5G cellular networks in licensed frequency bands for supporting diverse Internet of Things (IoT) services. This article studies the performance modeling of an NR-U and WiFi coexistence system using the NR-U Type B multichannel access procedure and the WiFi DCF procedure and focuses on a system model with two unlicensed channels: 1) a primary channel and 2) a secondary channel. 2-D Markov models are first built to describe the NR-U category-4 LBT procedure and the WiFi DCF procedure on the primary channel, respectively. A 3-D Markov model is then built to describe the WiFi DCF procedure on the secondary channel, taking into account the effect of the NR-U Type B multichannel access procedure. Based on the built Markov models, performance models are further derived for analyzing the system performances in terms of system throughput and packet delay on the primary channel and/or secondary channel. The derived performance models are validated through simulation results. The impacts of the number of NR-U gNBs/WiFi APs and major system parameters on the system performances are investigated based on the derived performance models.

This article studies the performance modeling of an NR-U and WiFi coexistence system using the NR-U category-4 listen-before-talk (LBT) procedure and 802.11e enhanced distributed channel access (EDCA) mechanism in the presence of hidden nodes. 3-D Markov models are first developed to describe the NR-U category-4 LBT procedure for a priority-${p}$queue of an NR-U gNB and the 802.11e EDCA mechanism for an AC-${q}$queue of a WiFi AP, taking into account the effects of hidden nodes and the transmission of different-priority data. Based on the Markov models, theoretical models are derived for analyzing the mean throughput of a priority-${p}$queue in an NR-U gNB and that of an AC-${q}$queue in a WiFi AP as well as the mean transmission delay of an NR-U priority-${p}$data packet and that of a WiFi AC-${q}$data packet. The derived theoretical performance models are validated through simulation results. Based on the performance models, the impacts of major system parameters on the system performance are investigated, including the density of NR-U gNBs, the sensing distance and channel transmission rate of an NR-U gNB, and the data packet arrival rate at an NR-U gNB.

Abstract This work studies the precoding methods for the reduced cycle‐prefix (CP) orthogonal time frequency space (OTFS) modulation to simplify the receiver design. First, a system model is established which shows that there are both intra‐slot‐interference and inter‐slot‐interference in the reduced‐CP OTFS system. Then, the precoding operation is proposed to pre‐cancel the inter‐slot‐interference in the reduced‐CP OTFS system so that the bit error rate (BER) performance can be greatly improved. According to how much channel state information (CSI) needs to be fed back, precoding methods for two different cases are proposed: the partial CSI feedback case and the partial and statistical CSI feedback case. Among them, the partial and statistical CSI based precoding method has very small feedback burden. Simulation results of the BER performance of the proposed precoding methods for reduced‐CP OTFS are reported and compared under different system parameters.

This paper studies resource scheduling for wireless small cloud assuming that the execution of an application is modeled as a workflow graph. A scheduling method is proposed which consists of two parts. For the task-to-computing-unit assignment part which operates on the long-term time-scale, the solution is to partition the workflow into task groups which have loose data transfer constraint and each task group is assigned to the computing unit which can finish executing the task group at the earliest time. For the computing and wireless communication resource scheduling part which operates on the short-term time-scale, we simply use the round-robin principle. A part by part delay component estimation method is presented based on which the average delay performance can be analyzed. Simulation results validate the theoretical analysis and show that the proposed method outperforms existing methods. The impact of different parameter settings on the system performance is also investigated.

Licensed-Assisted Access Long-Term Evolution (LAA- LTE) framework can offload excessive data stream to the unlicensed spectrum. Devices from different operators will compete for the same band and have to endure a more complex and congested channel conditions. Therefore a rapid and efficient access mechanism should be introduced in the LAA framework. Based on the Type B LBT channel access scheme, this work proposes an Energy-Based Signal Assisted Access (EBSA) mechanism. The EBSA mechanism can realize the rapid information transfer between devices belonging to different operators. Then a Markov chain model is set up and used to analyze this mechanism theoretically. Simulation results verify the validity of the theoretical analysis, and in the ultra-dense employed LAA network, the proposed method is efficient compared with the existing access mechanisms.

This paper considers the BBU computing resource allocation problem in a C-RAN system. In particular, we consider the additional computing resource consumption introduced for coordination between different BBUs when the tasks of a cell is processed by different BBUs and consider load balancing among multiple BBUs in order to reduce the effect of a breakdown BBU on the system. The problem is formulated as an optimization problem with the objective to minimize the total amount of computing resources used for processing the tasks of all cells and meanwhile balance the allocation of the computing resources in all used BBUs. The formulated problem is equivalent to the classical bin-packing problem, which has proved to be NP-hard. To solve the problem, a heuristic genetic algorithm (HeuGA) is proposed to find an approximate solution to the problem. HeuGA combines a genetic algorithm (GA) with a first-fit (FF) algorithm, and consists of four steps: population initialization, fitness function construction, individual selection, and cross and mutation. Simulation results show that HeuGA outperforms the FF algorithm and the GA algorithm in terms of the number of BBUs used in the system and BBU resource allocation balancing without affecting the efficiency of the resource allocation.

The problem is to obtain the optimal number of base stations (BSs) and their locations on minimized power consumption with requirements of coverage rate and cell load. In this paper, we propose a mathematical model of evaluation index for cell planning and identify the optimization problem. The system model characterizes the total power consumption of planning area, taking into account the radio power, baseband power and cooling system loss. Two other problems, base station number search algorithm and cell sites planning with fixed cell number, are provided to solve the original problem. Simulation result proves that the algorithm enables to find a suboptimal solution, which relatively decreases the total power consumption under the planning constraints.

This paper proposes an adaptive time division scheduling (ATDS) based resource allocation algorithm for device-to-device (D2D) communication underlaying cellular networks. The proposed ATDS algorithm introduces an adaptive TDS strategy, in which all D2D pairs in a cellular system are adaptively scheduled into a series of timeslots for communication based on an improved proportional fairness algorithm. For a particular timeslot, the scheduling priority of each D2D pair is first determined based on both the data transmission rate and the rate satisfaction ratio of each D2D pair. Then, a set of D2D pairs are assigned to the timeslot based on the priorities determined and only those D2D pairs with higher priorities are assigned to the timeslot. After that, the spectrum resources of cellular users are allocated for each D2D pair assigned to the timeslot. Simulation results show that the proposed ATDS algorithm can not only achieve a larger system throughput but also achieve a better fairness as compared with a couple of resource allocation algorithms that do not use time division scheduling.

This paper considers the resource allocation problem for device-to-device (D2D) communication underlaying cellular networks. In particular, we consider a system scenario where the number of D2D users is larger than that of cellular users. To address this scenario, we propose to not only allow one D2D pair to share the resources of multiple cellular users, but also allow one cellular user to share its resources with multiple D2D pairs. We formulate the resource allocation problem as a mixed integer non-linear programming (MINLP) problem with the objective to maximize the system capacity. To solve the problem, we propose a heuristic Graph-coloring resOurce ALlocation (GOAL) algorithm based on a graph-coloring approach, where the D2D pairs in the system are viewed as a set of vertexes and the resources of cellular users are viewed as a set of colors. To support resource allocation, GOAL introduces the concept of the interference negligible distance (INS) to identify those D2D pairs can simultaneously share the same spectrum resources of cellular users, and the concept of the signal to interference ratio (SIR) limited area (SLA) to identify a set of D2D pairs that cannot share the spectrum resources of a particular cellular user. Simulation results show that the proposed GOAL algorithm can significantly improve the system capacity and accommodate more D2D users.

This paper considers the resource allocation problem in device-to-device (D2D) communication underlaying cellular networks and proposes a time division scheduling (TDS) resource allocation algorithm to efficiently exploit the downlink spectrum resources of cellular users to support more D2D communication. The proposed TDS algorithm introduces a time division scheduling framework, in which the scheduling period of a base station is divided into a set of timeslots and the D2D pairs in the system are assigned to different timeslots for communication in a balanced manner in order to accommodate more D2D users in the system. In the D2D pair assignment for each timeslot, it follows a location dispersion principle in order to reduce the interference from D2D users to cellular users and thus increase the system throughput. Moreover, the minimum required data rate of each D2D pair is taken into account in the allocation of cellular resources in each timeslot. Simulation results show that the proposed TDS algorithm can significantly improve the system performance in terms of the system throughput and D2D user satisfaction ratio of a cellular system.

This study investigates the scheduling problem for a single‐cell downlink distributed antenna systems (DASs) with a massive number of remote access units (RAUs). To reduce signalling overhead under limited backhaul capacity, the authors make use of local long‐term channel state information (CSI) in coordinated scheduling design. They first derive the ergodic rate expressions for both the single RAU transmission and the cooperative RAU transmission modes as functions of long‐term CSI. Then greedy scheduling algorithms (GSAs) aiming for the maximum ergodic sum rate for the massive DAS using long‐term CSI are proposed. To mitigate the intra‐cell interference, a two‐stage GSA with hybrid transmission mode is devised. Asymptotic analysis reveals that as the number of RAUs goes to infinity, intra‐cell interference can be effectively mitigated. Simulation results verify the analysis and demonstrate that the two‐stage GSA exhibits a higher ergodic sum‐rate.

This work studies the dynamic resource allocation for device‐to‐device communication and proposes a service time prediction‐based dynamic resource allocation mechanism for a cellular system. The proposed allocation mechanism introduces the concept of ‘utility’ as a metric to reflect the quality of experience of users with different service requirements, and takes into account service time prediction in spectral resource allocation. Specifically, it enables a base station to predict the service time of a new user and the remaining service time of existing users in the system, and takes into account the predicted service time and remaining service time in calculating the effective utility gain caused by the arrival of a new user. Based on the calculated effective utility gain, the base station will select the physical resource block (PRB) with the largest effective utility gain every time it allocates a PRB to the new user. Simulation results show that the proposed dynamic allocation mechanism can significantly improve the system performance in terms of the overall average utility and the number of users successfully admitted to the system.

This paper considers the resource allocation problem for device-to-device (D2D) communication underlaying cellular networks and proposes a service-aware resource allocation (SARA) scheme for D2D communication to improve the network performance. SARA takes into account the different service requirements of D2D users. It consists of two allocation phases: on-demand resource allocation and secondary resource allocation. In the first phase, cellular user resources are allocated on demand to meet the different service requirements of D2D pairs. In the second phase, the remaining cellular user resources after the first phase are allocated to D2D users based on an objective function in order to improve the resource utilization. Simulation results show that the proposed SARA scheme can significantly improve the performance of a cellular system in terms of the satisfaction ratio of D2D pairs and the system throughput.

Resource allocation is a critical issue in device-to-device (D2D) communication underlying cellular networks. Existing work only considers static resource allocation, which cannot well address the dynamic resource allocation problem in a real cellular network. This paper studies dynamic resource allocation for D2D communication and proposes a service time prediction based dynamic resource allocation mechanism for a cellular network system. The proposed allocation mechanism considers different service types, and introduces the concept of "utility" as a metric to reflect the performance of a user or the system, and takes into account service time prediction in spectral resource allocation. Specifically, it enables the base station to predict the service time of a new user and the remaining service time of existing users in the system, and take into account the predicted service time and remaining service time in calculating the utility gain caused by a new user. Based on the estimated utility gain, the base station will select the physical resource block (PRB) with the largest utility gain every time it allocates a PRB to the new user. Simulation results show that the proposed dynamic allocation mechanism can significantly improve the system performance in terms of the system utility and the number of users successfully admitted to the system.

This paper proposes a Capacity Oriented Resource ALlocation (CORAL) algorithm for resource allocation in device-to-device (D2D) communication underlaying mobile cellular networks. To achieve high system capacity, CORAL allows a D2D pair to share the resources of more than one cellular user. Meanwhile, it introduces the concept of a Capacity-Oriented REstricted (CORE) region for a D2D pair to determine a candidate cellular user set for the D2D pair in resource allocation, which excludes those cellular users that introduce a negative system capacity gain, and thus can help increase the system capacity and reduce the computational complexity of CORAL as well. Moreover, CORAL attempts to avoid the situation where different D2D pairs simultaneously choose the same optimal cellular user, which would otherwise cause severe inteferences between the different D2D pairs and larger interference on the optimal cellular user. In addition, CORAL can ensure that each D2D pair can be allocated the resources of at least one cellular user. Simulation results show that CORAL can achieve a good performance in terms of both the system capacity and the rate loss of all cellular users.

This paper investigates the scheduling problem for a single-cell distributed multiple-input multiple-output (d-MIMO) downlink system with a massive number of remote access units (RAUs), N. We first derive the ergodic rate expressions for both the single RAU transmission (SRT) and the cooperative RAU transmission (CRT) modes as functions of long-term channel state information (CSI). Then, greedy scheduling algorithms aiming for maximizing the ergodic sum rate for the massive d-MIMO system using local long-term CSI are proposed. To mitigate intra-cell interference, a two-stage greedy scheduling algorithm (GSA) is developed to further improve the ergodic sum rate. Asymptotic analysis reveals that with infinite N intra-cell interference can be efficiently mitigated. Simulation results verify the derived expressions and demonstrate that the two-stage GSA exhibits a higher ergodic sum rate.

This paper proposes limited feedback spatial division multiple access (SDMA) schemes for transmit correlated channels by using statistical channel state information (SCSI) and instantaneous channel state information (ICSI). Different from conventional codebook-based MU-MIMO scheme, the proposed multi-beam selection with single channel quality indicator (CQI) feedback (MBS-SCF) scheme determines the preferred beam vector by exploiting the SCSI and only feeds back CQI at each timeslot. The performance of the MBS-SCF scheme is nearly the same as the conventional scheme. In order to further improve the sum rate, we propose multi-beam selection with dual CQIs feedback (MBS-DCF) scheme, which determines dual statistical eigen-directions and feeds back dual CQIs at each timeslot. It will increase the opportunity to exploit multiuser diversity and multiplexing gain. Simulation results demonstrate that the MBS-DCF scheme exhibits a higher sum rate than the conventional scheme does.

The IEEE 802.16 mesh mode coordinated distributed scheduling has been expected to be collision free so that only one node can transmit at any transfer opportunity (TO) in the two-hop neighborhood. We will show in this paper that this collision free property is not unconditional. Assuming the common interference modeling, necessary and sufficient conditions under which the IEEE 802.16 mesh mode coordinated distributed scheduling can be collision free are identified. As the IEEE 802.16 standard defines only a framework for the mesh mode coordinated distributed scheduling, our study could provide a suggestion on how to guarantee the correct functionality.