21 May 2024 in IEEE Transactions on Network and Service Management
Edge caching is a promising solution for next-generation networks by empowering caching units in small-cell base stations (SBSs), which allows user equipments (UEs) to fetch users’ requested contents that have been pre-cached in SBSs. It is crucial for SBSs to predict accurate popular contents through learning while protecting users’ personal information. Traditional federated learning (FL) can protect users’ privacy but the data discrepancies among UEs can lead to a degradation in model quality. Therefore, it is necessary to train personalized local models for each UE to predict popular contents accurately. In addition, the cached contents can be shared among adjacent SBSs in next-generation networks, thus caching predicted popular contents in different SBSs may affect the cost to fetch contents. Hence, it is critical to determine where the popular contents are cached cooperatively. To address these issues, we propose a cooperative edge caching scheme based on elastic federated and multi-agent deep reinforcement learning (CEFMR) to optimize the cost in the network. We first propose an elastic FL algorithm to train the personalized model for each UE, where adversarial autoencoder (AAE) model is adopted for training to improve the prediction accuracy, then a popular content prediction algorithm is proposed to predict the popular contents for each SBS based on the trained AAE model. Finally, we propose a multi-agent deep reinforcement learning (MADRL) based algorithm to decide where the predicted popular contents are collaboratively cached among SBSs. Our experimental results demonstrate the superiority of our proposed scheme to existing baseline caching schemes.
26 September 2016 in IEEE Transactions on Wireless Communications
This paper addresses the beam allocation problem in a switched-beam based massive multiple-input-multiple-output (MIMO) system working at the millimeter wave (mmWave) frequency band, with the target of maximizing the sum data rate. This beam allocation problem can be formulated as a combinatorial optimization problem under two constraints that each user uses at most one beam for its data transmission and each beam serves at most one user. The brute-force search is a straightforward method to solve this optimization problem. However, for a massive MIMO system with a large number of beams N, the brute-force search results in intractable complexity O(NK), where K is the number of users. In this paper, in order to solve the beam allocation problem with affordable complexity, a suboptimal low-complexity beam allocation (LBA) algorithm is developed based on submodular optimization theory, which has been shown to be a powerful tool for solving combinatorial optimization problems. Simulation results show that our proposed LBA algorithm achieves nearly optimal sum data rate with complexity O(K logN). Furthermore, the average service ratio, i.e., the ratio of the number of users being served to the total number of users, is theoretically analyzed and derived as an explicit function of the ratio N=K.
9 March 2017 in IEEE Journal on Selected Areas in Communications
Device-to-device (D2D) assisted offloading heavily depends on the participation of human users. The content preference and sharing willingness of human users are two crucial factors in the D2D assisted offloading. In this paper, with consideration of these two factors, the optimal content pushing strategy is investigated by formulating an optimization problem to maximize the offloading gain measured by the offloaded traffic. Users are placed into groups according to their content preferences and share content with intergroup and intragroup users at different sharing probabilities. Although the optimization problem is nonconvex, the closed-form optimal solution for a special case is obtained, when the sharing probabilities for intergroup and intragroup users are the same. Furthermore, an alternative group optimization (AGO) algorithm is proposed to solve the general case of the optimization problem. Finally, simulation results are provided to demonstrate the offloading performance achieved by the optimal pushing strategy for the special case and AGO algorithm. An interesting conclusion drawn is that the group with the largest number of interested users is not necessarily given the highest pushing probability. It is more important to give high pushing probability to users with high sharing willingness.
4 June 2013 in IEEE Journal on Selected Areas in Communications
In the orthogonal frequency division multiple access (OFDMA) system, one of the efficient and low complex methods to allocate radio resources among multiple users is chunk-based resource allocation, which groups a number of adjacent subcarriers into a chunk and allocates resources chunk by chunk. In this paper, performance analysis of chunk-based resource allocation is studied in the multi-cell OFDMA environment. Fractional frequency reuse (FFR) is considered in the cellular OFDMA. Basically, FFR divides each cell into central and edge areas where two different values of the frequency reuse factor are assumed. This paper analytically evaluates how spectral efficiency performance is affected by system parameters, including radius ratio of the central area to the whole cell, transmit signal to noise ratio (SNR), number of users, number of subcarriers per chunk, and coherence bandwidth. The numerical results show that there exists an optimal radius ratio to achieve the highest spectral efficiency in the proposed research. The optimal radius ratio is about 0.7, which is almost irrespective of the SNR, number of users, and number of subcarriers per chunk. In other words, the sizes of the central area and the edge area of the whole cell are almost equal when achieving the optimal performance.
26 September 2024 in IEEE Systems Journal
As the rise of information and communication technology, the cooperative work of vehicles has become crucial in realizing Internet of Vehicles (IoV). Federated learning (FL) is a promising technology to protect vehicles' privacy in IoV. Vehicles with limited computation capacity may face a large computational burden associated with FL. Federated edge learning (FEEL) systems are introduced to solve such a problem. In FEEL systems, vehicles adopt the cellular-vehicle to everything (C-V2X) mode 4 to upload encrypted data to road side units' (RSUs) cache queue. Then, RSUs train the data transmitted by vehicles, update the local model hyperparameters, and send back results to vehicles, thus, vehicles' computational burden can be released. However, each RSU has limited cache queue. To maintain the stability of cache queue and maximize the accuracy of model, it is essential to select appropriate vehicles to upload data. The vehicle selection method for FEEL systems faces challenges due to the random departure of data from the cache queue caused by the stochastic channel and the different system status of vehicles. This article proposes a vehicle selection method for FEEL systems that aims to maximize the accuracy of model while keeping the cache queue stable. Extensive simulation experiments demonstrate that our proposed method outperforms other baseline selection methods.
13 November 2024 in IEEE Transactions on Vehicular Technology
Delay-sensitive applications for autonomous vehicles (AVs) require a substantial amount of computational resources. However, the onboard computation resources may be insufficient, resulting in long processing latencies. To deal with this critical issue, we jointly consider roadside unit (RSU) and assistant vehicle offloading, along with resource allocation, to minimize latency for vehicular tasks. This approach also takes into account frequency reuse among sub-areas for assistant vehicle offloading. The latency minimization problem can be formulated as a mixedinteger non-linear programming (MINLP) problem. Given the inherent complexity of the MINLP problem, we propose a twostep solution. The first step focuses on the combined decision of assistant vehicle offloading and transmit power allocation. To solve this problem, we propose a particle swarm optimization (PSO) algorithm with low complexity and low average transmit power. The second step deals with RSU offloading/local computation decision, bandwidth allocation, and computation resource allocation. An iterative algorithm is proposed to achieve the optimal solution. Without adding additional computation resources, simulation results demonstrate that the proposed vehicular task offloading approach improves overall delay performance than the adaptive MEC offloading scheme and the pure MEC computing scheme