arXiv subjects

Kin K. Leung

Publications and source records attributed to Kin K. Leung.

At least 55 records · Page 3Linked to original sources

arXiv2020

Energy-Efficient Resource Management for Federated Edge Learning with CPU-GPU Heterogeneous Computing

Edge machine learning involves the deployment of learning algorithms at the network edge to leverage massive distributed data and computation resources to train artificial intelligence (AI) models. Among others, the framework of federated edge learning (FEEL) is popular for its data-privacy preservation. FEEL coordinates global model training at an edge server and local model training at edge devices that are connected by wireless links. This work contributes to the energy-efficient implementation of FEEL in wireless networks by designing joint computation-and-communication resource management ($\text{C}^2$RM). The design targets the state-of-the-art heterogeneous mobile architecture where parallel computing using both a CPU and a GPU, called heterogeneous computing, can significantly improve both the performance and energy efficiency. To minimize the sum energy consumption of devices, we propose a novel $\text{C}^2$RM framework featuring multi-dimensional control including bandwidth allocation, CPU-GPU workload partitioning and speed scaling at each device, and $\text{C}^2$ time division for each link. The key component of the framework is a set of equilibriums in energy rates with respect to different control variables that are proved to exist among devices or between processing units at each device. The results are applied to designing efficient algorithms for computing the optimal $\text{C}^2$RM policies faster than the standard optimization tools. Based on the equilibriums, we further design energy-efficient schemes for device scheduling and greedy spectrum sharing that scavenges "spectrum holes" resulting from heterogeneous $\text{C}^2$ time divisions among devices. Using a real dataset, experiments are conducted to demonstrate the effectiveness of $\text{C}^2$RM on improving the energy efficiency of a FEEL system.

cs.IT↗

arXiv2020

Resource Sharing in the Edge: A Distributed Bargaining-Theoretic Approach

The growing demand for edge computing resources, particularly due to increasing popularity of Internet of Things (IoT), and distributed machine/deep learning applications poses a significant challenge. On the one hand, certain edge service providers (ESPs) may not have sufficient resources to satisfy their applications according to the associated service-level agreements. On the other hand, some ESPs may have additional unused resources. In this paper, we propose a resource-sharing framework that allows different ESPs to optimally utilize their resources and improve the satisfaction level of applications subject to constraints such as communication cost for sharing resources across ESPs. Our framework considers that different ESPs have their own objectives for utilizing their resources, thus resulting in a multi-objective optimization problem. We present an $N$-person \emph{Nash Bargaining Solution} (NBS) for resource allocation and sharing among ESPs with \emph{Pareto} optimality guarantee. Furthermore, we propose a \emph{distributed}, primal-dual algorithm to obtain the NBS by proving that the strong-duality property holds for the resultant resource sharing optimization problem. Using synthetic and real-world data traces, we show numerically that the proposed NBS based framework not only enhances the ability to satisfy applications' resource demands, but also improves utilities of different ESPs.

cs.GT↗

arXiv2020

Overcoming Noisy and Irrelevant Data in Federated Learning

Many image and vision applications require a large amount of data for model training. Collecting all such data at a central location can be challenging due to data privacy and communication bandwidth restrictions. Federated learning is an effective way of training a machine learning model in a distributed manner from local data collected by client devices, which does not require exchanging the raw data among clients. A challenge is that among the large variety of data collected at each client, it is likely that only a subset is relevant for a learning task while the rest of data has a negative impact on model training. Therefore, before starting the learning process, it is important to select the subset of data that is relevant to the given federated learning task. In this paper, we propose a method for distributedly selecting relevant data, where we use a benchmark model trained on a small benchmark dataset that is task-specific, to evaluate the relevance of individual data samples at each client and select the data with sufficiently high relevance. Then, each client only uses the selected subset of its data in the federated learning process. The effectiveness of our proposed approach is evaluated on multiple real-world image datasets in a simulated system with a large number of clients, showing up to $25\%$ improvement in model accuracy compared to training with all data.

cs.LG↗

arXiv2020

State Action Separable Reinforcement Learning

Reinforcement Learning (RL) based methods have seen their paramount successes in solving serial decision-making and control problems in recent years. For conventional RL formulations, Markov Decision Process (MDP) and state-action-value function are the basis for the problem modeling and policy evaluation. However, several challenging issues still remain. Among most cited issues, the enormity of state/action space is an important factor that causes inefficiency in accurately approximating the state-action-value function. We observe that although actions directly define the agents' behaviors, for many problems the next state after a state transition matters more than the action taken, in determining the return of such a state transition. In this regard, we propose a new learning paradigm, State Action Separable Reinforcement Learning (sasRL), wherein the action space is decoupled from the value function learning process for higher efficiency. Then, a light-weight transition model is learned to assist the agent to determine the action that triggers the associated state transition. In addition, our convergence analysis reveals that under certain conditions, the convergence time of sasRL is $O(T^{1/k})$, where $T$ is the convergence time for updating the value function in the MDP-based formulation and $k$ is a weighting factor. Experiments on several gaming scenarios show that sasRL outperforms state-of-the-art MDP-based RL algorithms by up to $75\%$.

cs.LG↗

arXiv2020

Adaptive Gradient Sparsification for Efficient Federated Learning: An Online Learning Approach

Federated learning (FL) is an emerging technique for training machine learning models using geographically dispersed data collected by local entities. It includes local computation and synchronization steps. To reduce the communication overhead and improve the overall efficiency of FL, gradient sparsification (GS) can be applied, where instead of the full gradient, only a small subset of important elements of the gradient is communicated. Existing work on GS uses a fixed degree of gradient sparsity for i.i.d.-distributed data within a datacenter. In this paper, we consider adaptive degree of sparsity and non-i.i.d. local datasets. We first present a fairness-aware GS method which ensures that different clients provide a similar amount of updates. Then, with the goal of minimizing the overall training time, we propose a novel online learning formulation and algorithm for automatically determining the near-optimal communication and computation trade-off that is controlled by the degree of gradient sparsity. The online learning algorithm uses an estimated sign of the derivative of the objective function, which gives a regret bound that is asymptotically equal to the case where exact derivative is available. Experiments with real datasets confirm the benefits of our proposed approaches, showing up to $40\%$ improvement in model accuracy for a finite training time.

cs.LG↗

arXiv2020

Resource Allocation in One-dimensional Distributed Service Networks

We consider assignment policies that allocate resources to users, where both resources and users are located on a one-dimensional line. First, we consider unidirectional assignment policies that allocate resources only to users located to their left. We propose the Move to Right (MTR) policy, which scans from left to right assigning nearest rightmost available resource to a user, and contrast it to the Unidirectional Gale-Shapley (UGS) matching policy. While both these policies are optimal among all unidirectional policies, we show that they are equivalent with respect to the expected distance traveled by a request (request distance), although MTR is fairer. Moreover, we show that when user and resource locations are modeled by statistical point processes, and resources are allowed to satisfy more than one user, the spatial system under unidirectional policies can be mapped into bulk service queuing systems, thus allowing the application of a plethora of queuing theory results that yield closed form expressions. As we consider a case where different resources can satisfy different numbers of users, we also generate new results for bulk service queues. We also consider bidirectional policies where there are no directional restrictions on resource allocation and develop an algorithm for computing the optimal assignment which is more efficient than known algorithms in the literature when there are more resources than users. Finally, numerical evaluation of performance of unidirectional and bidirectional allocation schemes yields design guidelines beneficial for resource placement.

cs.PF↗

arXiv2020

Let's Share: A Game-Theoretic Framework for Resource Sharing in Mobile Edge Clouds

Mobile edge computing seeks to provide resources to different delay-sensitive applications. This is a challenging problem as an edge cloud-service provider may not have sufficient resources to satisfy all resource requests. Furthermore, allocating available resources optimally to different applications is also challenging. Resource sharing among different edge cloud-service providers can address the aforementioned limitation as certain service providers may have resources available that can be ``rented'' by other service providers. However, edge cloud service providers can have different objectives or \emph{utilities}. Therefore, there is a need for an efficient and effective mechanism to share resources among service providers, while considering the different objectives of various providers. We model resource sharing as a multi-objective optimization problem and present a solution framework based on \emph{Cooperative Game Theory} (CGT). We consider the strategy where each service provider allocates resources to its native applications first and shares the remaining resources with applications from other service providers. We prove that for a monotonic, non-decreasing utility function, the game is canonical and convex. Hence, the \emph{core} is not empty and the grand coalition is stable. We propose two algorithms \emph{Game-theoretic Pareto optimal allocation} (GPOA) and \emph{Polyandrous-Polygamous Matching based Pareto Optimal Allocation} (PPMPOA) that provide allocations from the core. Hence the obtained allocations are \emph{Pareto} optimal and the grand coalition of all the service providers is stable. Experimental results confirm that our proposed resource sharing framework improves utilities of edge cloud-service providers and application request satisfaction.

cs.NI↗

arXiv2019

MACS: Deep Reinforcement Learning based SDN Controller Synchronization Policy Design

In distributed software-defined networks (SDN), multiple physical SDN controllers, each managing a network domain, are implemented to balance centralised control, scalability, and reliability requirements. In such networking paradigms, controllers synchronize with each other, in attempts to maintain a logically centralised network view. Despite the presence of various design proposals for distributed SDN controller architectures, most existing works only aim at eliminating anomalies arising from the inconsistencies in different controllers' network views. However, the performance aspect of controller synchronization designs with respect to given SDN applications are generally missing. To fill this gap, we formulate the controller synchronization problem as a Markov decision process (MDP) and apply reinforcement learning techniques combined with deep neural networks (DNNs) to train a smart, scalable, and fine-grained controller synchronization policy, called the Multi-Armed Cooperative Synchronization (MACS), whose goal is to maximise the performance enhancements brought by controller synchronizations. Evaluation results confirm the DNN's exceptional ability in abstracting latent patterns in the distributed SDN environment, rendering significant superiority to MACS-based synchronization policy, which are 56% and 30% performance improvements over ONOS and greedy SDN controller synchronization heuristics.

cs.NI↗

arXiv2019

Energy-Efficient Radio Resource Allocation for Federated Edge Learning

Edge machine learning involves the development of learning algorithms at the network edge to leverage massive distributed data and computation resources. Among others, the framework of federated edge learning (FEEL) is particularly promising for its data-privacy preservation. FEEL coordinates global model training at a server and local model training at edge devices over wireless links. In this work, we explore the new direction of energy-efficient radio resource management (RRM) for FEEL. To reduce devices' energy consumption, we propose energy-efficient strategies for bandwidth allocation and scheduling. They adapt to devices' channel states and computation capacities so as to reduce their sum energy consumption while warranting learning performance. In contrast with the traditional rate-maximization designs, the derived optimal policies allocate more bandwidth to those scheduled devices with weaker channels or poorer computation capacities, which are the bottlenecks of synchronized model updates in FEEL. On the other hand, the scheduling priority function derived in closed form gives preferences to devices with better channels and computation capacities. Substantial energy reduction contributed by the proposed strategies is demonstrated in learning experiments.

cs.IT↗

arXiv2019

A Game-Theoretic Framework for Resource Sharing in Clouds

Providing resources to different users or applications is fundamental to cloud computing. This is a challenging problem as a cloud service provider may have insufficient resources to satisfy all user requests. Furthermore, allocating available resources optimally to different applications is also challenging. Resource sharing among different cloud service providers can improve resource availability and resource utilization as certain cloud service providers may have free resources available that can be ``rented'' by other service providers. However, different cloud service providers can have different objectives or \emph{utilities}. Therefore, there is a need for a framework that can share and allocate resources in an efficient and effective way, while taking into account the objectives of various service providers that results in a \emph{multi-objective optimization} problem. In this paper, we present a \emph{Cooperative Game Theory} (CGT) based framework for resource sharing and allocation among different service providers with varying objectives that form a coalition. We show that the resource sharing problem can be modeled as an $N-$player \emph{canonical} cooperative game with \emph{non-transferable utility} (NTU) and prove that the game is convex for monotonic non-decreasing utilities. We propose an $\mathcal{O}({N})$ algorithm that provides an allocation from the \emph{core}, hence guaranteeing \emph{Pareto optimality}. We evaluate the performance of our proposed resource sharing framework in a number of simulation settings and show that our proposed framework improves user satisfaction and utility of service providers.

cs.MA↗

arXiv2019

Online Collection and Forecasting of Resource Utilization in Large-Scale Distributed Systems

Large-scale distributed computing systems often contain thousands of distributed nodes (machines). Monitoring the conditions of these nodes is important for system management purposes, which, however, can be extremely resource demanding as this requires collecting local measurements of each individual node and constantly sending those measurements to a central controller. Meanwhile, it is often useful to forecast the future system conditions for various purposes such as resource planning/allocation and anomaly detection, but it is usually too resource-consuming to have one forecasting model running for each node, which may also neglect correlations in observed metrics across different nodes. In this paper, we propose a mechanism for collecting and forecasting the resource utilization of machines in a distributed computing system in a scalable manner. We present an algorithm that allows each local node to decide when to transmit its most recent measurement to the central node, so that the transmission frequency is kept below a given constraint value. Based on the measurements received from local nodes, the central node summarizes the received data into a small number of clusters. Since the cluster partitioning can change over time, we also present a method to capture the evolution of clusters and their centroids. As an effective way to reduce the amount of computation, time-series forecasting models are trained on the time-varying centroids of each cluster, to forecast the future resource utilizations of a group of local nodes. The effectiveness of our proposed approach is confirmed by extensive experiments using multiple real-world datasets.

cs.DC↗

arXiv2019

Dynamic Service Migration in Mobile Edge Computing Based on Markov Decision Process

In mobile edge computing, local edge servers can host cloud-based services, which reduces network overhead and latency but requires service migrations as users move to new locations. It is challenging to make migration decisions optimally because of the uncertainty in such a dynamic cloud environment. In this paper, we formulate the service migration problem as a Markov Decision Process (MDP). Our formulation captures general cost models and provides a mathematical framework to design optimal service migration policies. In order to overcome the complexity associated with computing the optimal policy, we approximate the underlying state space by the distance between the user and service locations. We show that the resulting MDP is exact for uniform one-dimensional user mobility while it provides a close approximation for uniform two-dimensional mobility with a constant additive error. We also propose a new algorithm and a numerical technique for computing the optimal solution which is significantly faster than traditional methods based on standard value or policy iteration. We illustrate the application of our solution in practical scenarios where many theoretical assumptions are relaxed. Our evaluations based on real-world mobility traces of San Francisco taxis show superior performance of the proposed solution compared to baseline solutions.

cs.DC↗

arXiv2019

Adaptive Federated Learning in Resource Constrained Edge Computing Systems

Emerging technologies and applications including Internet of Things (IoT), social networking, and crowd-sourcing generate large amounts of data at the network edge. Machine learning models are often built from the collected data, to enable the detection, classification, and prediction of future events. Due to bandwidth, storage, and privacy concerns, it is often impractical to send all the data to a centralized location. In this paper, we consider the problem of learning model parameters from data distributed across multiple edge nodes, without sending raw data to a centralized place. Our focus is on a generic class of machine learning models that are trained using gradient-descent based approaches. We analyze the convergence bound of distributed gradient descent from a theoretical point of view, based on which we propose a control algorithm that determines the best trade-off between local update and global parameter aggregation to minimize the loss function under a given resource budget. The performance of the proposed algorithm is evaluated via extensive experiments with real datasets, both on a networked prototype system and in a larger-scale simulated environment. The experimentation results show that our proposed approach performs near to the optimum with various machine learning models and different data distributions.

cs.DC↗

arXiv2019

Optimal Energy Consumption with Communication, Computation, Caching and QoI-Guarantee

Energy efficiency is a fundamental requirement of modern data communication systems, and its importance is reflected in much recent work on performance analysis of system energy consumption. However, most works have only focused on communication and computation costs, but do not account for caching costs. Given the increasing interest in cache networks, this is a serious deficiency. {In this paper, we consider the problem of energy consumption in data communication, compression and caching (C$3$) with a Quality of Information (QoI) guarantee in a communication network. {Our goal is to identify the optimal data compression rate and data placement over the network to minimize the overall energy consumption in the network.} he formulated problem is a \emph{Mixed Integer Non-Linear Programming} (MINLP) problem with non-convex functions, which is NP-hard in general. } {We} propose a variant of spatial branch and bound algorithm (V-SBB), that can {provide} the $ε$-global optimal solution to {our problem}. {We numerically show that our C3 optimization framework can improve the energy efficiency up to 88\% compared to any C2 optimization between communication and computation or caching. Furthermore, for our energy consumption problem, V-SBB {provides comparatively better solution than some other MINLP solvers.}}

cs.NI↗

arXiv2019

Learning the Optimal Synchronization Rates in Distributed SDN Control Architectures

Since the early development of Software-Defined Network (SDN) technology, researchers have been concerned with the idea of physical distribution of the control plane to address scalability and reliability challenges of centralized designs. However, having multiple controllers managing the network while maintaining a "logically-centralized" network view brings additional challenges. One such challenge is how to coordinate the management decisions made by the controllers which is usually achieved by disseminating synchronization messages in a peer-to-peer manner. While there exist many architectures and protocols to ensure synchronized network views and drive coordination among controllers, there is no systematic methodology for deciding the optimal frequency (or rate) of message dissemination. In this paper, we fill this gap by introducing the SDN synchronization problem: how often to synchronize the network views for each controller pair. We consider two different objectives; first, the maximization of the number of controller pairs that are synchronized, and second, the maximization of the performance of applications of interest which may be affected by the synchronization rate. Using techniques from knapsack optimization and learning theory, we derive algorithms with provable performance guarantees for each objective. Evaluation results demonstrate significant benefits over baseline schemes that synchronize all controller pairs at equal rate.

cs.NI↗

arXiv2018

DQ Scheduler: Deep Reinforcement Learning Based Controller Synchronization in Distributed SDN

In distributed software-defined networks (SDN), multiple physical SDN controllers, each managing a network domain, are implemented to balance centralized control, scalability and reliability requirements. In such networking paradigm, controllers synchronize with each other to maintain a logically centralized network view. Despite various proposals of distributed SDN controller architectures, most existing works only assume that such logically centralized network view can be achieved with some synchronization designs, but the question of how exactly controllers should synchronize with each other to maximize the benefits of synchronization under the eventual consistency assumptions is largely overlooked. To this end, we formulate the controller synchronization problem as a Markov Decision Process (MDP) and apply reinforcement learning techniques combined with deep neural network to train a smart controller synchronization policy, which we call the Deep-Q (DQ) Scheduler. Evaluation results show that DQ Scheduler outperforms the antientropy algorithm implemented in the ONOS controller by up to 95.2% for inter-domain routing tasks.

cs.NI↗

arXiv2018

Joint Data Compression and Caching: Approaching Optimality with Guarantees

We consider the problem of optimally compressing and caching data across a communication network. Given the data generated at edge nodes and a routing path, our goal is to determine the optimal data compression ratios and caching decisions across the network in order to minimize average latency, which can be shown to be equivalent to maximizing the compression and caching gain under an energy consumption constraint. We show that this problem is NP-hard in general and the hardness is caused by the caching decision subproblem, while the compression sub-problem is polynomial-time solvable. We then propose an approximation algorithm that achieves a $(1-1/e)$-approximation solution to the optimum in strongly polynomial time. We show that our proposed algorithm achieve the near-optimal performance in synthetic-based evaluations. In this paper, we consider a tree-structured network as an illustrative example, but our results easily extend to general network topology at the expense of more complicated notations.

cs.NI↗

arXiv2017

How Better is Distributed SDN? An Analytical Approach

Distributed software-defined networks (SDN), consisting of multiple inter-connected network domains, each managed by one SDN controller, is an emerging networking architecture that offers balanced centralized control and distributed operations. Under such networking paradigm, most existing works focus on designing sophisticated controller-synchronization strategies to improve joint controller-decision-making for inter-domain routing. However, there is still a lack of fundamental understanding of how the performance of distributed SDN is related to network attributes, thus impossible to justify the necessity of complicated strategies. In this regard, we analyze and quantify the performance enhancement of distributed SDN architectures, influenced by intra-/inter-domain synchronization levels and network structural properties. Based on a generic weighted network model, we establish analytical methods for performance estimation under four synchronization scenarios with increasing synchronization cost. Moreover, two of these synchronization scenarios correspond to extreme cases, i.e., minimum/maximum synchronization, which are, therefore, capable of bounding the performance of distributed SDN with any given synchronization levels. Our theoretical results reveal how network performance is related to synchronization levels and inter-domain connections, the accuracy of which are confirmed by simulations based on both real and synthetic networks. To the best of our knowledge, this is the first work quantifying the performance of distributed SDN analytically, which provides fundamental guidance for future SDN protocol designs and performance estimation.

cs.DC↗