Search

Federated learning (FL) has been considered as an efficient yet privacy-preserving approach for enabling the distributed learning. There have been many studies investigating the applications of FL in different scenarios, e.g., Internet of Things, Internet of Vehicles, and UAV systems. However, due to delivering the trained model via wireless links, FL may suffer from a potential issue, i.e., some malicious users may intentionally overhear the trained model delivered through the wireless links. In this paper, we investigate the energy optimization for nonorthogonal multiple access (NOMA) assisted with secrecy provisioning. Specifically, we consider that the wireless devices (WDs) adopt NOMA to deliver their respectively trained local models to a base station (BS) which serves a parameter-server, and there exists a malicious node that overhears the parameter-server when delivering the aggregated global model to all WDs. We adopt the physical layer security to quantify the secrecy throughput under the eavesdropping attack and formulate an optimization problem to minimize the overall energy consumption of all the WDs in FL, by jointly optimizing the uplink time, the downlink time, the local model accuracy, and the uplink decoding order of NOMA. In spite of the non-convexity of this joint optimization problem, we propose an efficient algorithm, which is based on the theory of monotonic optimization, for finding the solution. Numerical results show that our proposed algorithm can achieve the almost same solutions as the LINGO's global-solver while reducing more than 90% computation-time than LINGO. Moreover, the results also show that our proposed NOMA decoding scheme can outperform some heuristic decoding schemes.

Equipped with machine learning (ML) models, unmanned aerial vehicle (UAV) swarms can execute various applications like surveillance and target detection. However, the connections between UAVs and cloud servers cannot be guaranteed, especially when executing massive data. Thus, traditional cloud-centric approach will not be suitable, since it may cause high latency and significant bandwidth consumption. In this work, we propose a federated learning (FL) framework via non-orthogonal multiple access (NOMA) for a UAV swarm which is composed of a leader-UAV and a group of follower-UAVs. Specifically, each follower-UAV updates its local model by using its collected data, and then all follower-UAVs form a NOMA-group to send their respectively trained FL parameters (i.e., the local FL models) to the leader-UAV simultaneously. We formulate a joint optimization of the uplink NOMA-transmission durations, downlink broadcasting duration, as well as the computation-rates of the leader-UAV and all follower-UAVs, aiming at minimizing the latency in executing the FL iterations until reaching a specified accuracy. Numerical results are presented to verify the effectiveness of our proposed algorithm, and demonstrate that the proposed algorithm can outperform some baseline strategies.

Real-time and accurate short-term traffic prediction can effectively improve traffic efficiency, reduce accidents, and facilitate relevant departments to take reasonable traffic guidance measures. Therefore, we propose a two-level data driven model for short-term traffic prediction in an edge computing environment. Firstly, a Deep Belief Network (DBN) is developed to extract the traffic characteristics between the road occupancy and road flow collected by the deployed detectors. Then, we predict the developed future road flow of each road segment based on the output of the DBN, which would be used as one of the inputs of a Hidden Markov Model (HMM). Finally, a HMM is developed to predict the future road speed of each road segment characterizing the statistical relationship between the road flow and road speed. To validate the effectiveness of our proposed model, the data from the Performance Measurement System (PeMS) of the California Department of Transportation is applied. Simulation results show that our proposed model has better prediction performance in short-term traffic prediction than other models.

Computation offloading via mobile edge computing (MEC) has been considered as one of the promising paradigms for enabling computation-intensive yet latency-sensitive services on resource-limited mobile terminals in future wireless systems. In this paper, we propose a channel sharing assisted multi-user computation offloading in MEC via non-orthogonal multiple access (NOMA), in which a group of edge-computing users (EUs) form a NOMA-group and further reuse the cellular user’s (CU’s) channel for computation-offloading transmission. Specifically, we formulate a joint optimization of the EUs’ offloaded workloads, NOMA-transmission duration, as well as the processing-rate allocations (for different EUs) at the edge computing server (ECS), with the objective of minimizing the overall latency for all EUs to complete their tasks. In spite of the non-convexity of the joint optimization, we identify the structural feature of the optimal offloading solution and propose a layered-algorithm for finding the solution efficiently. Numerical results are provided to validate the effectiveness of our proposed algorithm as well as the advantage of our multi-user offloading scheme via NOMA.

Non-orthogonal multiple access (NOMA) has been considered as a promising scheme for enabling ultra-high throughput transmission and massive-connectivity in next generation wireless systems. In this paper, we investigate the secrecy-based NOMA transmission for encountering the eavesdropping attack. Exploiting the NOMA-users simultaneous transmission as an artificial jamming, we investigate the joint optimization of NOMA-users’ power allocations and the secrecy-provisioning, with the objective of the effective secure throughput of NOMA-users while ensuring the fairness among them. Despite the non-convexity of the formulated joint optimization problem, we explore its hidden feature and design a search algorithm to compute the optimal solution. Numerical results are provided to validate the performance of our proposed algorithm.1

Mobile edge computing (MEC) has been considered as a promising approach for enabling computation-intensive Internet services in future wireless systems. In this paper, we investigate non-orthogonal multiple access (NOMA) assisted MEC, in which edge-computing users (EUs) adopt NOMA to simultaneously offload part of their computation-workloads to the edge-server (ES). To improve the spectrum-efficiency, we consider a paradigm of underlaying device-to-device (D2D) communications, namely, the EUs reuse a cellular user's (CU's) licensed channel for offloading transmission. We firstly characterize the transmit-powers of EUs and CU in this D2D approach, and then formulate a joint optimization of the EUs' computation- workloads offloading and the ES's computation-resource allocation, with the objective of minimizing the latency in completing the EUs' tasks. In spite of the non-convexity of the formulated problem, we exploit its layered structure and propose an efficient algorithm for computing the optimal solution. Numerical results are provided to validate the effectiveness and efficiency of our proposed NOMA assisted MEC via the D2D sharing 1.

In this paper, we investigate the revenue-sharing based computation-resource allocation for mobile Blockchain, in which conventional mobile devices (i.e., the edge-computing users, EUs) can acquire computation-resources from the edge-server (ES) for increasing their chances of winning the mining game in Blockchain. Meanwhile, as a compensation to the ES, the EUs adopt the revenue-sharing mechanism by sharing parts of their respective rewards to the ES. We formulate a joint optimization of the ES's computation-resource allocation to the EUs and the EUs' revenue-sharing to the ES, with the objective of maximizing the system-reward of all EUs and the ES. Despite the non-convexity of the joint optimization problem, we propose an algorithm based on the cyclic block coordinate descent (CBCD) to solve the problem. Our algorithm decouples the original problem into two subproblems and alternatively optimizes the EUs' revenue-sharing and ES's computation-resource allocation until reaching convergence. For each subproblem, we also propose an efficient algorithm for solving it. Numerical results are provided to validate the effectiveness and efficiency of our proposed algorithms, as well as the performance advantage of our proposed revenue-sharing based computation-resource allocation for mobile Blockchain.

With the rapid advancement of electric vehicle (EV) technology, EV has been emerging as a promising transportation due to the low carbon emission. However, the frequent and long time charging is indispensable to continue travelling. During peak hours, EVs further spend long time on the path routing because of the traffic congestion and queuing in the charging stations. Therefore, we study the EV charging scheduling problem that minimizes the total elapsed time which includes charging time for EVs through jointly optimizing the charging path routing and charging station selection in this paper. Considering the NP-hardness of this optimization problem, we propose an efficient EV charging scheduling method to obtain the optimal solution based on crowd sensing through considering the remaining energy in the battery, traffic condition, and the queue length of charging stations. Simulation results demonstrate that the proposed backtracking method based on crowd sensing can effectively reduce the total elapsed time, in comparison with the greedy algorithm.

In recent years, the 3rd generation partnership project (3GPP) has approved the narrowband Internet of Things (NB-IoT) system to support the low-data-rate machine- type communications. With the rapid development of NB- IoT technology, the NB-IoT traffic volumes have been experiencing the unprecedented growth. To this end, non-orthogonal multiple access (NOMA) and mobile edge computing (MEC) have been proposed as promising technologies for the NB-IoT system. In this paper, our goal is to minimize the total energy consumption subject to the computation capacity and execution latency limits by jointly optimizing the transmit power, computation resource allocation, and successive interference cancellation (SIC) ordering. Considering the NP-hardness of the joint optimization problem, we obtain the optimal solution in the coupled steps: the resource allocation optimization and the combinatorial SIC ordering optimization. By exploiting the convex nature of the resource allocation optimization problem, we obtain the optimal transmit power and computation resource allocation based on the KKT conditions and the idea of gradient descent method when fixing the SIC ordering. Considering the combinatorial optimization of SIC ordering, we further propose a tabu search based SIC ordering algorithm on the basis of the proposed resource allocation optimization algorithm. Finally, simulation results demonstrate that the proposed joint optimization of transmit power, computation resource allocation, and SIC ordering in the context of NOMA can effectively reduce the total energy consumption of MEC- aware NB-IoT system, in comparison with the frequency- division multiple access technique.

Mobile edge computing (MEC) has been envisioned as a promising scheme to address the explosive growth of computation-hungry mobile applications in future cellular systems. In this paper, we investigate the secrecy-driven energy-efficient computation offloading via MEC. Specifically, we take the secrecy-outage into account when an eavesdropper overhears the mobile terminal's (MT's) offloaded data to the edge server (ES) and formate a joint optimization of the MT's computation offloading, radio transmission, and secrecy-outage level, with the objective of minimizing the MT's energy consumption for completing its required workload. Despite the non-convexity of the joint optimization problem, we propose an efficient algorithm to find the optimal offloading solution. Based on the optimal solution for an arbitrary offloading pair of MT and ES, we further consider the scenario of multi-MT and multi-ES, and investigate the optimal pairing between the MTs and ESs for computation offloading, with the objective of minimizing all MTs' total energy consumption. Exploiting the matching structure of the pairing problem, we propose an efficient auction-based time-division scheduling algorithm to find the optimal pairing solution. Numerical results are provided to validate the effectiveness and efficiency of our proposed algorithms and the advantage of our computation offloading with secrecy-provisioning.

Powering relays with harvested renewable ambient energy has been emerging as a promising solution to reduce the on-grid energy consumption and greenhouse gas emissions in green relaying communication networks. In this paper, we study the joint time scheduling and power allocation problem for the Decode-and-Forward energy-harvesting relay communication network. Particularly, our goal is to maximize the end-to-end throughput by a deadline subject to the finite data and energy storage. Due to the multi-slot optimization, the traditional deep reinforcement learning (RL) framework cannot be directly applied to obtain the optimal solution of maximizing the end-to-end throughput by a deadline in the online manner. To this end, we explore a novel deep reinforcement learning framework consisting of multiple computation units to obtain the online time scheduling and power allocation based on the current causal knowledge of energy arrivals and channel fading at each time slot. Simulation results show that the proposed deep reinforcement learning based algorithm can achieve more than 90% of maximum end-to-end throughput.

Blockchain has emerged as a decentralized and trustable ledger for recording and storing digital transactions. The mining process of blockchain, however, requires to consume heavy computation-resources to solve the proof-of-work puzzle, which is prohibitive from the perspective of the conventional mobile terminals (MTs). In this work, we investigate the mobile edge computing (MEC) assisted Blockchain, in which a group of MTs exploit MEC to enhance their computing-powers for solving the proof-of-work puzzle. In particular, we consider the recent advanced multiple-access MEC (MA-MEC) which enables each MT to acquire the computing-powers from several edge servers (ESs) simultaneously. We formulate a joint optimization problem which aims at maximizing the total reward of all MTs, while keeping the fairness among the MTs. Despite the non-convexity of the problem, we propose efficient layered algorithms to compute the optimal solution. Extensive numerical results have been provided to validate the efficiency of our proposed algorithms.

Narrow-band Internet of Things (NB-IoT), one of the emerging paradigms of low power wide area networks (LPWAN) for Internet of Things (IoT), has been envisioned as a promising solution to enable massive connectivity, cost-efficient, and highly reliable Internet of Thing (IoT) systems in future smart cities. In this work, we build up an indoor environment-temperature monitoring system based on NB-IoT. We present a detailed design of our system and illustrate the key technologies. Based on our system and the collected data (i.e., the temperature data), we further design an abnormality-detection mechanism based on the support vector machine (SVM). We provide experimental results to show the performance of our designed system and the proposed abnormality-detection mechanism.

This paper proposes an orthogonal frequency division multiplexing (OFDM) relaying wireless power transfer based protocol for energy-constrained unmanned aerial vehicle (UAV) communication network. At the first step of the proposed protocol, the energy-constrained UAV separates the received signals into two disjoint subcarrier groups to perform energy harvesting (EH) and information decoding (ID). At the second step, UAV forwards the received information signals with the energy harvested a priori. Based on the proposed protocol, this paper aims to find the joint optimization of subcarrier grouping and power allocation for maximizing the transmission rate under the EH constraint. We further solve such a joint resource allocation problem via dual decomposition after transforming it into an equivalent convex optimization problem. Simulation results indicate that the performance of our proposed protocol can be significantly improved for the OFDM relaying based UAV communication network.

We study status updates over wireless fading channels in terms of the age of information, which measures the freshness of the last received update since its generation. In this paper, we model the wireless transmission systems with Poisson arrivals, a First-Come-First-Served (FCFS) packet buffer, and two-state Markov modulated service process (MMSP) as M/MMSP/1/K queueing model. We derive closed-form expressions of average age of information in two extreme cases when the packet buffer size is either infinite or zero. Our analysis and simulation results show that the average age of information for M/MMSP/1/∞ queue greatly depends on channel variations and that M/MMSP/1/1 queue is more affected by packet arrival rate.

Dual-connectivity (DC), an emerging paradigm in the recent 3GPP specification, is envisioned as a promising solution to enhance mobile users' (MUs') traffic offloading by aggregating radio resources at both macro and small cells. In this paper, we investigate the energy-efficient DC-enabled traffic offloading through small cells which are powered by the on-grid power supply and harvesting renewable energy from nature. In spite of reducing the on-grid power consumption, powering traffic offloading by energy harvesting (EH) leads to the offloading outage due to the intermittency in EH power-supply, which degrades the offloading throughput. Therefore, to reap both the advantages of DC and the EH-supply, we propose a joint traffic scheduling and power allocation scheme that aims at minimizing the total on-grid power consumption, while accounting for the offloading outage and guaranteeing the MU's quality of service (QoS) requirement. In spite of the non-convexity nature of the joint optimization of traffic scheduling and power allocation, we propose an algorithm to efficiently compute the optimal offloading solution. Numerical results are provided to validate our proposed algorithm and the performance gain of the proposed DC-enabled traffic offloading scheme.

This paper studies the best-effort transmission policy that maximizes the sum rate by the joint power and rate allocation in the buffer-aided NOMA relay network. In particular, we first provide the sum-rate maximization formulation under the constraints of maximum transmit power and buffer size. We then derive a closed-form expression of optimal power and rate allocation that solves this optimization problem. Based on the derived closed-form expression, we further obtain the lower bound of the probability of using NOMA when adopting the best-effort transmission policy.

The advanced non-orthogonal multiple access (NOMA) has been considered as a promising scheme to satisfy the ultimate goals of future 5G cellular networks for providing ultra-high throughput and ultra-dense connections. By enabling a group of mobile users (MUs) to simultaneously share a same frequency channel and adopting successive interference cancellation to mitigate the co-channel interference, the NOMA can significantly improve the spectrum efficiency compared with the conventional orthogonal multiple access (OMA). However, due to cellular operators' limited and crowded spectrum resources, a critical question is how to properly size the channel bandwidth for the NOMA- enabled transmission to satisfy all MUs' traffic demands. In this paper, we propose a joint optimization scheme of bandwidth and power allocations for the NOMA- enabled downlink transmission, with the objective of minimizing the overall resource consumption cost that accounts for both the spectrum consumption cost and power consumption cost. In spite of the non-convexity nature of the joint optimization problem, we propose an efficient algorithm to compute the optimal bandwidth allocation and power allocation. Numerical results validate the proposed algorithm and the performance advantage of the proposed NOMA-enabled transmission in saving the overall resource consumption cost.

Offloading data traffic from the conventional macro- cell base stations to densely deployed small-cell base stations (SBSs) has been emerging as a promising technique to support the explosion of data traffic with reduced energy consumption and improved quality of service provision. In this paper, we study the joint spectrum allocation and power allocation problem for the data offloading in energy-harvesting downlink small-cell networks. First, we formulate the resource allocation problem under the revenue maximization criterion, which is expressed as the difference between the total utility across users and the total power payment. By proving the convexity of the problem, we can compute the solution efficiently. Numerical results show that the energy-efficiency can be improved while alleviating the burden of macro-cell base station through using the resource allocation scheme proposed for data offloading.

Ambient energy harvesting has emerged as a promising technique to improve the energy efficiency and reduce the total greenhouse gas emissions for green relay networks. In this paper, we study the joint relay selection and power control problem for the Decode-and-Forward energy-harvesting wireless relay network. In particular, the problem formulation is to maximize the end-to-end system throughput by a deadline under the limitations of data and energy storage. To solve the problem, we decompose such an optimization problem into two subproblems: the joint time scheduling and power control subproblem and the relay selection subproblem. Due to the convex nature of the joint time scheduling and power control subproblem, we derive the optimal solution via the primal decomposition. Based on the obtained system throughput, we can quickly select the best relay that achieves the maximum throughput. Simulation results show that the proposed algorithm can guarantee the maximum system throughput, in comparison with some existing algorithms.

The limited backhaul capacity has been considered as an important issue that influences the performance of traffic offloading in small-cell networks. We thus investigate how mobile users (MUs) efficiently offload traffic to different small cells with the new paradigm of dual-connectivity, subject to different backhaul capacities of small cells. Specifically, we formulate an optimization problem that jointly optimizes the MUs' access-selections, traffic scheduling, and transmit-power allocations, with the objective of minimizing all MUs' mobile data cost. To tackle with the difficulty due to the mixed binary nonconvex optimization problem, we decompose the joint optimization problem into a subproblem that optimizes the MUs' traffic scheduling and power allocations and a top problem that optimizes the MUs' access-selections. Based on our previous study [1], we can efficiently solve the subproblem by exploiting its hidden monotonicity. With the solution of the subproblem, we further propose two effective algorithms (namely, a simulated annealing based algorithm and a fast access-selection algorithm) to solve the top problem. The key idea of the fast access-selection algorithm is to exploit the information of limited backhaul capacities and properly remove the infeasible selection to reduce the search space. Numerical results validate the effectiveness of our proposed algorithms and the performance advantage of the proposed joint resource allocations for traffic offloading.

This paper considers an energy harvesting sensor node with finite data and energy storage, which transmits data packets with different reward values to its corresponding receiver node. In this regard, we propose an optimal threshold-based transmission scheduling policy for maximizing the long-term average transmission reward. In particular, we first analyze the performance of the proposed threshold-based transmission scheduling policy by studying the steady states of the energy harvesting sensor node and derive its expected transmission reward. We then propose a polynomial-time algorithm to compute these optimal reward value thresholds that maximize the expected reward. Numerical results show that the system expected reward increases with the increase of data and energy storage capacity. Our analysis further shows that the expected reward increases exponentially with the increase of data or energy storage capacity.

We investigate mobile user's (MU's) uplink traffic offloading based on the new paradigm of small-cell dual-connectivity (DC) that allows a MU to simultaneously offload traffic to a small-cell base station (sBS) and send data to a macro base station (mBS). In spite of the benefit of reducing the MU's power consumption, traffic offloading via DC leads to a security issue, i.e., a malicious node close to the sBS might intentionally eavesdrop the MU's offloaded data by receiving the MU's radio signal. Therefore, we formulate an optimization problem that jointly determines the MU's traffic scheduling and transmit-powers to the sBS and mBS, with the objective of minimizing the MU's total power consumption, while guaranteeing both the MU's secrecy requirement and traffic demand requirement. To solve the formulated joint optimization problem, we propose an effective layered-algorithm (which exploits the hidden convexity of the problem) to solve it and compute the MU's optimal offloading solution. Numerical results validate the efficiency of our algorithm and the significant influences of the MU's secrecy requirement.

In this paper, we investigate the energy-aware data- offloading of mobile user (MU) which schedules its traffic demand to a macro Base Station (BS) and a small-cell access point (AP) simultaneously. For saving the usage of licensed spectrum, we consider that the MU uses unlicensed spectrum to offload data. The open access of unlicensed spectrum, however, results in that the MU's data offloading suffer from uncontrollable interference, which comprises the benefit of data offloading. We propose an outage-probability to quantify such an adverse influence and formulate a joint rate-splitting and power allocation problem to minimize a system-wise cost accounting for both the MU's power consumption and the BS's licensed channel usage. Despite the non-convexity of the joint optimization problem, we transform it into three rate- allocation problems under different cases and derive the respective optimal solutions, which yield the globally optimal solution for the original problem. Numerical results are provided to validate the optimal offloading-solution.

With the rapid development of smart handled devices and mobile internet services, mobile network operators (MNOs) have experienced an explosive growth in traffic demand in cellular access networks. Intelligently offloading traffic through small-cell networks has been widely considered as an efficient approach for MNOs to relieve traffic congestion in cellular access networks and accommodate more mobile users (MUs) with satisfactory quality of service (QoS). However, offloading traffic to small-cell networks might incur co-channel interference among the MUs. Such interference, if without a proper control, will lead to significant power consumptions of the MUs, which undermines the benefit of traffic offloading. In this paper, we are motivated to investigate the joint access-selection and power allocation problem, in which the MUs are appropriately selected to offload their traffic demands to different small-cell networks with proper transmit-powers. Our objective is to maximize a system-reward that takes into account both the MNO's economic reward for serving the MUs and the MUs' transmit-power consumption costs. The formulated problem corresponds to a mixed binary and non-convex optimization problem. We exploit the decomposable structure of the problem and propose an efficient algorithm to solve it. Numerical results are provided to show the performance of the proposed algorithm as well as the benefits of the proposed traffic offloading scheme.

Driven by the intermittent nature of energy harvesting, this paper aims to maximize the long-term average transmission reward of energy harvesting sensor nodes subject to the finite data and energy storage. In particular, we first propose a sub-optimal threshold-based transmission scheduling policy for energy harvesting sensor nodes through maximizing conditional expected rewards with respect to data storage and energy storage, respectively. We then study the properties of the conditional expected rewards and further propose a low complexity algorithm to efficiently compute these reward value thresholds. Numerical results show that the system expected reward increases with the increase of data and energy storage capacity. Our analysis further shows that the proposed transmission scheduling can guarantee the same reward by increasing the data storage space when the energy storage knowledge is imperfect.

In this paper, we investigate how the mobile users (MUs) can effectively offload traffic by taking advantage of the capability of dual-connectivity, which enables an MU to simultaneously communicate with a macro base station (BS) and a small-cell access point (AP) via two radio-interfaces. We formulate an optimization problem that jointly determines each MU's traffic schedule (between the BS and AP) and power allocations (between two radio-interfaces), with the objective to minimize all MUs' total cost. We first propose an effective scheme to characterize the feasibility of the joint optimization problem. Then, by exploiting the layered structure, we propose an efficient layered scheme to solve it. Numerical results are provided to validate the proposed schemes and show the performance gain via properly offloading MUs' traffic via dual-connectivity.

Dynamic spectrum access via active spectrum sharing has been considered as a promising approach to improve the spectrum utilization for future wireless systems. In this paper, based on our recent study on the optimal transmit-power allocation for an active spectrum sharing system comprised of single primary-user (PU) and multiple secondary-users (SUs) [7], we move a further step to investigate a more challenging scenario comprised of multiple PUs and multiple SUs. Specifically, we formulate a joint SU-selection and power allocation problem, in which the PUs properly select different groups of SUs to share channels with and the PUs and SUs then determine the proper transmit- powers. We aim at maximizing a system reward for serving the SUs' traffic while trading off the PUs' additional power consumptions to guarantee their required quality of service (QoS). We exploit the layered structure of the joint optimization problem and propose an efficient algorithm to solve it. Numerical results are provided to validate performances of the proposed algorithm and show advantages of performing the joint SU-selection and power allocation in active spectrum sharing.

In this paper, we model the wireless fading channels with Poisson arrivals and three-state Markov modulated service processes (MMSP) as M/MMSP/1 queues. We derive the generalized Pollaczek-Khinchin formula that precisely characterizes the delay performance of three-state Markov channels. Our key contribution mainly lies in the state transition factor defined in this paper that clearly describes the impact of channel varying on the delay performance of these Markov channels. Moreover, we resort to an innovative linear approximation of the start-service probability in terms of server utilization to obtain closed-form expressions, which are all verified by simulations.

We study a joint user association and resource allocation problem for device-to-device communications underlaying cellular networks. In particular, the problem formulation is to maximize the sum rate across cellular users (CUs) and device-to-device users (DUs) under the constraints of the minimum data rate requirements of individual users. To solve the problem, we decompose such optimization problem into two subproblems: the resource allocation subproblem for any pair of DU and CU, and the user association subproblem. Due to the convex nature of the resource allocation subproblem, we derive the optimal solution in the closed-form expression. We further relax the user association subproblem, although as an integer programming, to a linear programming, and then obtain its optimal solution via interior-point methods. Our analysis shows that the proposed algorithm has low overhead signaling during its implementation, and our simulation results further show that the proposed algorithm can guarantee the maximum system sum rate.

This paper studies the optimal power allocation required by the source and destination nodes for dual-hop decode-and-forward (DF) relay networks over Rayleigh fading channels. In particular, we first provide the problem formulation that minimizes the expected total transmit power consumption of DF cooperative transmission under the constraints of maximum tolerable outage probability and limited power. We then derive the optimal power allocation due to the convex nature. Meanwhile, the approximately optimal power allocation is provided in the closed-form expression. Based on these results, we further analyze the performance of DF cooperative transmission in comparison to existing transmission schemes.

In this paper, we investigate the revenue optimization of a cellular system, which intelligently provides access services to device-to-device users (DUs) by reusing the resource-blocks (RBs) of cellular users (CUs). While charging the DUs for services, the cellular system compensates for the additional power consumption costs of the CUs to meet their required quality of service (QoS), and hence aims at achieving the best tradeoff between charging the DUs and affording the CUs' costs to maximize its own revenue. We formulate this energy-aware revenue optimization problem as a joint RB-reuse and power control problem, which we further decompose into a power control problem for each individual CU-DU pair and a DU-selection problem for selecting appropriate DUs to reuse the CUs' RBs. For each CU-DU pair, we derive the optimal power allocation in closed form and the maximum gain of the cellular system from this pair. Based on the gains of all CU-DU pairs, we then consider the DU-selection problem as maximum weighted matching on a bipartite graph and solve it by using linear relaxation. Numerical results validate our analysis regarding the optimal power allocation for each CU-DU pair and the BS's optimal selection of the DUs to reuse the CUs' RBs such that the BS's revenue is maximized.

In this paper, we investigate the optimal spectrum leasing and sharing for dynamic spectrum access (DSA) network which actively shares its spectrum resource with a group of secondary users (SUs) for reaping revenue. The DSA aims at maximizing its total charge to the SUs based on their achieved throughput, while taking into account the additional power consumption of the primary user (PU) to overcome the SUs' interference such that quality of service (QoS) of the PU is guaranteed. The interference between the PU and the SUs as well as the interference within the SUs themselves make the spectrum leasing problem strictly nonconvex and thus difficult to solve in general. To tackle with this difficulty, we exploit the decomposition structure of the concerned spectrum leasing problem and propose a layered monotonic optimization approach to solve it efficiently. Numerical results are presented to validate our proposed approach and its computational efficiency.

In this paper, we investigate optimal management of local energy trading in future smart micro-grid (SMG) via pricing. In SMG, energy consumers and providers, in addition to trading with utility company, can also perform local energy trading controlled by a local trading manager (LTM) for reaping benefits. We first quantify the benefits achieved by the consumers and providers from local trading and then formulate a two-layered optimization framework to investigate i) how the energy consumers and providers maximize their benefits via appropriately adjusting their local trading decisions in response to the LTM's pricing, and ii) how the LTM adjusts its price in local market to benefit the consumers and providers as much as possible while guaranteeing a required gain for itself. We propose two algorithms to solve the layered optimization problem and perform numerical experiments with practical data set to validate the proposed local trading model and the algorithms.

In this paper, we aim to obtain the optimal joint power and rate control that maximizes the interference-aware secondary system utility subject to the interference temperature constraints of primary users (PUs) on the basis of the information-theoretic interference processing technique. In particular, we first draw the capacity-approaching rate region of secondary users (SUs) with given transmit power and channel information of SUs based on the near optimal interference cancellation and power splitting technique. Then, we formulate the power controlled secondary system utility maximization problem under the interference temperature constraints of PUs in the capacity-approaching rate region. To solve the optimization problem, we decompose it into one convex subproblem over the rate region with fixed power allocation and the other partitioning subproblem of the power allocations. Finally, we present an efficient algorithm, which can quickly achieve the utility-maximized power and rate control based on the idea of simulated annealing.

In the heterogeneous cellular network, each mobile station perceives different channel gains to different base stations. Therefore, it is important to associate a mobile station with the right base station with transmit power control so as to achieve substantial improvement in spectrum-efficiency and energy-efficiency. Despite its importance, the problem that maximizes the overall system revenue with minimum total transmit power consumption through joint BS association and power control has remained largely unsolved for the large-scale heterogeneous cellular network, mainly due to its non-convex and combinatorial nature. To solve this problem, this paper first proposes a single-stage formulation that captures two interactive objectives, i.e., the maximization of overall system revenue and the minimization of total transmit power consumption. The single-stage optimization problem is then efficiently solved by the proposed POSEM algorithm based on the theory of coalition formation game and the idea of simulated annealing. Finally, our analysis shows that the proposed algorithm is guaranteed to converge to the global optimal solution fast.

Random medium access control (MAC) is extensively used in wireless communication systems as a means to allocate the shared radio resource in a distributed manner. However, some of the ultimate performance limits, including the maximum throughput, are not well understood due to the non-convex nature of the utility optimization problem. In this paper, we introduce the monotonic optimization method to tackle the intractability of nonconvex utility maximization problems in random-access networks. We show that the optimal random MAC problem can be transformed to a canonical monotonic optimization and thus can be efficiently solved using the polyblock outer approximation algorithm. We also show that the conventional polyblock outer approximation algorithm can be significantly simplified by exploiting the unique problem structure of the optimal random MAC. Based on the proposed method, we perform extensive evaluations of the optimal performance in both fully interfered and partially interfered networks. Our results here would help to understand the fundamental limits of random-access networks, and spur the study on the practical random MAC protocols to find the optimal parameters that maximize a system utility.

In this paper, we consider the network-welfare optimization for a Cognitive Radio Network (CRN) consisting of multiple Primary-Users (PUs) and multiple Secondary-Users (SUs). For each pair of PU and SU, the interaction between them is modeled as a Stackelberg game where the PU shares its licensed channel with the SU by charging the interference, and the SU is allowed to access the PU channel by paying the interference fee. The objective of the CRN thus is to maximize the network-welfare of all PUs and SUs by forming optimal pairs among them, which can be considered as a Maximum Weight Matching (MWM) problem on a bipartite graph. We first propose an efficient algorithm to quantify the optimal social welfare associated with each pair of PU and SU. Then, by exploiting the special structure of the MWM problem, we propose two algorithms, namely, a modified Belief Propagation (BP) algorithm and a heuristic algorithm based on Simulated Annealing (SA) to determine the optimal matching.

Future smart grid has been widely conceived to be featured by its flexible supply-demand management and great exploitation of renewable energy. However, due to the volatility of renewable sources, exploitation of renewable energy introduces uncertainty in energy supply. Therefore, it is of practical importance to quantify the optimal exploitation of renewable energy based on the supply-demand framework for future smart grid, where the energy-providers (or the energy-users) adaptively adjust their energy-provisioning (or energy-demands) according to the system information that takes account of the volatility of renewable energy. In this work, we consider a residential smart grid equipped with a centralized renewable source as a supplement to the electricity acquired from the grid. Our model aims at: i) quantifying the optimal exploitation of renewable energy that trades off between the system-wide benefit from using the renewable energy and the associated cost due to its volatility, and ii) characterizing how the volatility of renewable energy influences its optimal exploitation. An efficient distributed algorithm is proposed to determine the optimal utilization of renewable energy as well as the associated energy scheduling decisions based on the supply-demand model.

Spectrum sensing and compulsory idling are two important functionalities of the overlaid Cognitive Radio (CR) to facilitate the opportunistic spectrum access and incumbent protection for primary user systems. Since both the spectrum sensing and compulsory idling introduce energy-overheads to CR users, it is critical for CR to balance between these additional energy-overheads and its energy consumption for effective data transmission. We thus propose energy-efficient transmission policies for CR that aims at minimizing its total cost (including both energy consumption and delay cost) to deliver a target traffic payload. We formulate this problem as a stochastic shortest path Markov decision process and obtain the optimal rate-adaptation policy via the value iteration. We further propose a low-complexity rate-adaptation policy through the certainty equivalent control with approximation to save computational complexity. Numerical results show the low-complexity policy achieves comparable performance as the optimal policy.

Transmit power control in wireless networks has long been recognized as an effective mechanism to mitigate co-channel interference. The optimal transmit power control problem is typically non-convex if a system utility is to be maximized. In our recent work [1], we proposed a distributed algorithm, referred to as GLAD, that obtains the global optimal solution to the non-convex power control problem. Same as other existing distributed power control algorithms, GLAD requires extensive message passing among all users in the network, which leads to high signaling overhead and high processing complexity. To reduce the message passing, this paper proposes a variant of the GLAD algorithm, referred to as I-GLAD, where the prefix “I” stands for infrequent message passing. The convergence of I-GLAD can be proved regardless of the reduction in the message passing rate. To further reduce the processing complexity at each transmitter, we develop an enhanced version of I-GLAD, referred to as NI-GLAD, where only the control messages from the neighboring links are processed. Our simulation results show that I-GLAD approximately converges to the global optimal solution regardless of the type of the system utility function. Meanwhile, the optimality of the solution obtained by NI-GLAD depends on the selection of the neighborhood size.

We consider a distributed power control algorithm for infrastructureless ad hoc wireless networks, where each link distributively and asynchronously updates its transmission power with limited message passing among links. This algorithm provably converges to the set of global optimal solutions despite the non-convexity of the power control problem. In contrast with existing distributed power control algorithms, our algorithm makes no stringent assumptions on the system utility functions. In particular, the utility function is allowed to be concave or non-concave, differentiable or non-differentiable, continuous or discontinuous, and monotonic or non-monotonic.

In multi-cell networks where mobile stations perceive different channel gains to different base stations (BS), it is critical to associate a mobile station with the proper BS to maintain good communication quality with limited bandwidth resources. Oftentimes, the already-challenging BS association problem is further complicated by the need of transmission power control, which is an essential component to manage co-channel interference in many wireless communications systems. Despite its importance, joint BS association and power control (BAPC) problem has remained largely open, mainly due to its non-convex nature that makes the global optimal solution difficult to obtain. In this paper, we propose a novel algorithm, referred to as BARN, to solve the joint BAPC problem efficiently and optimally in the sense that the number of mobile stations in service is maximized and the total transmission power is minimized in the same time. In particular, we first propose a single-stage formulation that captures the two objectives simultaneously. Then, the problem is transformed in a way that can be efficiently solved using the BARN algorithm that is derived from the standard Benders' Decomposition. Finally, we derive a close-form analytical formula to characterize the effect of the termination criterion of the algorithm on the gap between the obtained solution and the optimal one. By carefully choosing the termination rule, the BARN algorithm can always converge to the global optimal solution.

Maximizing system utility corresponding to different performance measures through power control has been a long standing open problem in interference-limited multiuser multicarrier wireless networks. The complicated coupling between the mutual interference of links on each subcarrier gives rise to a series of non-convex power control optimization problems, for which the global optimal solution is hard to obtain. This paper proposes a novel algorithm, MARL, to efficiently solve the non-convex power control problem in multiuser multicarrier wireless networks. The algorithm is guaranteed to converge to a global optimal solution, as long as the utility function of each link is monotonically increasing with its data rate. The MARL algorithm is designed based on three key observations of the power control problems considered in this paper: (1) the objective function is increasing in (1+SINR) (SINR: signal to interference- plus-noise ratio); (2) the feasible set of the corresponding equivalent reformulated problem is always "normal", although not necessarily convex; and (3) the two former observations imply that the power control problem can be transformed into a monotonic optimization (MO) problem, where the optimal solution always occurs at the upper boundary of the feasible (1+SINR) region. The MARL algorithm finds the desired optimal power control solution by constructing a series of polyblocks that approximate the feasible (1+SINR) region with an increasing precision. Furthermore, by tuning the error tolerance in MARL, we could engineer a desirable tradeoff between optimality and convergence time. MARL provides an important benchmark for performance evaluation of other heuristic algorithms targeting the same problem. With the help of MARL, we evaluate the performance of a state-of-the-art algorithm through extensive simulations.

This paper studies optimal power control for throughput maximization in wireless ad hoc networks. Optimal power control problem in ad hoc networks is known to be non-convex due to the co-channel interference between links. As a result, a global optimal solution is difficult to obtain. Previous work either simplified the problem by assuming that the signal- to-interference-and-noise-radio (SINR) of each and every link is much higher than 1, or settled for suboptimal solutions. In contrast, we propose a novel methodology to compute the global optimal power allocation in a general SINR regime. In particular, we formulate the problem into an equivalent linear multiplicative fractional programming (LMFP). A global optimization algorithm, referred to as LMFP-based power allocation (LBPA) algorithm, is proposed to solve the LMFP with reasonable computational complexity. Our analysis proves that the LBPA algorithm is guaranteed to converge to a global optimal solution. Through extensive simulations, we show that the proposed algorithm significantly improves the throughput of wireless networks compared with existing ones.