With the continuous expansion of the power Internet of Things (IoT) and the rapid increase in the number of Smart Devices (SDs), the data generated by SDs has exponentially increased. The traditional cloud-based smart grid cannot meet the low latency and high reliability requirements of emerging applications. By moving computing, data, and services from the centralized cloud to Edge Servers (ESs), edge computing exhibits excellent performance in communication delay and traffic reduction. Simultaneously, service caching also shows attractive advantages in handling the surge in data traffic. In this paper, we consider the joint optimization of computing offloading and service caching in edge computing-based smart grid, and formulate the problem as a Mixed-Integer Non-Linear Program (MINLP), aiming to minimize the task cost of the system. The original problem is decomposed into an equivalent master problem and sub-problem, and a Collaborative Computing Offloading and Resource Allocation Method (CCORAM) is proposed to solve the optimization problem, which includes two low-complexity algorithms. Specifically, a gradient descent allocation algorithm is first proposed to determine the computing resource allocation strategy, and then a game theory-based algorithm is proposed to determine the computing strategy. Simulation results show that CCORAM with low time complexity is very close to the optimal method, and performs much better than other benchmark methods.

Mobile edge computing (MEC) reduces data service latency by pushing data to the network edge. However, due to the dynamic and diverse requests of mobile users, the problem of mobile edge caching is more complex than cloud caching. Therefore, the existing model-based caching strategies cannot be directly used in the mobile edge caching environment. Besides, when taking the cooperative storage relationship between neighbor edge servers into consideration, the caching problem becomes more difficult. To this end, we formulate an mobile edge caching problem to minimize the total latency in mobile edge computing. Firstly, a heuristic caching strategy is proposed to solve the mobile edge caching problem in the single-time-slot scenario. Then, with the consideration of users’ mobility and the correlation of files, we propose a caching strategy for the multiple-time-slot scenario based on multi-agent deep reinforcement learning. To address the cold start problem in deep reinforcement learning, we adopt the proposed heuristic caching strategy used in the single-time-slot scenario to further optimize the training results. Extensive experiments on generated data and real-world datasets are conducted to verify that the proposed edge caching strategies can achieve the minimum latency compared with the state-of-the-art strategies.

Loosely coupled and highly cohesived microservices running in containers are becoming the new paradigm for application development. Compared with monolithic applications, applications built on microservices architecture can be deployed and scaled independently, which promises to simplify software development and operation. However, the dramatic increase in the scale of microservices and east-west network traffic in the data center have made the cluster management more complex. Not only does the scale of microservices cause a great deal of pressure on cluster management, but also cascading QoS violations present a substantial risk for SLOs (Service Level Objectives). In this paper, we propose a Microservice-Oriented Topology-Aware Scheduling Framework (MOTAS), which effectively utilizes the topologies of microservices and clusters to optimize the network overhead of microservice applications through a heuristic graph mapping algorithm. The proposed framework can also guarantee the cluster resource utilization. To deal with the dynamic environment of microservice, we propose a mechanism based on distributed trace analysis to detect and handle QoS violations in microservice applications. Through real-world experiments, the framework has been proved to be effective in ensuring cluster resource utilization, reducing application end-to-end latency, improving throughput, and handling QoS violations.

Vehicular CrowdSensing (VCS) has become a powerful sensing paradigm by selecting users driving vehicles to perform tasks. In most existing research, the platform centrally allocates tasks according to the collected user information. We argue that the information collection process results in user privacy leakage, and the centralized allocation leads to a heavy computation complexity. We propose to apply a distributed task allocation method in the widely-used route navigation system. The navigation system recommends several routes to a user and each route may cover some tasks. Then, the user distributively selects a route according to the route profit (task reward minus route cost). Since the task reward is shared by users, the route selections of users may influence each other. Hence, it remains unclear how to design a distributed route navigation approach to reach an equilibrium state (i.e., each user is satisfied with the selected route), while guaranteeing a good total profit. To this end, we formulate the problem as a multi-user potential game and propose a distributed route navigation algorithm. The trace-based simulation results verify that the proposed algorithm achieves a Nash equilibrium, while achieving a total user profit performance close to that of the optimal solution.

This paper jointly considers the optimization of multi-user computing offloading and service caching in Mobile Edge Computing (MEC), and formulates the problem as a Mixed-Integer Non-Linear Program (MINLP), aiming to minimize the task cost of the system. The original problem is decomposed into an equivalent master problem and sub-problem, and a Collaborative Computing Offloading and Resource Allocation Method (CCORAM) is proposed to solve the optimization problem, which includes two low-complexity algorithms. Simulation results show that CCORAM with low time complexity is very close to the optimal method, and performs much better than other benchmark methods.

Mobile Edge Computing (MEC) has been envisioned as a promising distributed computing paradigm, where mobile users offload their tasks to edge nodes to decrease the cost of energy and computation. However, most existing works only consider the congestion of wireless channels as the crucial factor influencing the strategy-making process, and ignore the impact of the offloading among edge nodes. In addition, centralized task offloading strategies result in heavy computation complexity in center nodes. Along this line, we take both the congestion of wireless channels and the offloading among multiple edge nodes into consideration to enrich users’ offloading strategies. To this end, we first formulate the offloading problem as a multi-user potential game, and then propose a distributed task offloading algorithm to reach an equilibrium state which can also protect individual privacy. Specifically, in the above task offloading algorithm, we propose two subalgorithms to select users for updating strategies: Parallel User Selection Algorithm (PUS) and Single User Selection Algorithm (SUS) in order to substantially accelerate the convergence. Extensive experiments on three real-world data sets validate that the proposed algorithm achieves a Nash equilibrium and effectively decreases the total user cost which is acceptable compared to the optimal solution.

In this preliminary study, we consider the server allocation problem for edge computing system deployment. Our goal is to minimize the average turnaround time of application requests/tasks, generated by all mobile devices/users in a geographical region. We consider two approaches for edge cloud deployment: the flat deployment, where all edge clouds co-locate with the base stations, and the hierarchical deployment, where edge clouds can also co-locate with other system components besides the base stations. In the flat deployment, we demonstrate that the allocation of edge cloud servers should be balanced across all the base stations, if the application request arrival rates at the base stations are equal to each other. We also show that the hierarchical deployment approach has great potentials in minimizing the system's average turnaround time. We conduct various simulation studies using the CloudSim Plus platform to verify our theoretical results. The collective findings trough theoretical analysis and simulation results will provide useful guidance in practical edge computing system deployment.

Mobile Edge Computing (MEC) has emerged as a promising computing paradigm in 5G networks, which can empower User Equipments (UEs) with computation and energy resources offered by migrating workloads from the UEs to the MEC servers. Although the issues of computation offloading and resource allocation in MEC have been studied with different optimization objectives, they mainly investigate quasi-static system environments, without considering the different resource requirements and time-varying system conditions in a dynamic system. In this paper, we exploit a multi-user MEC system, and investigate the task execution scheme for dynamic joint optimization of offloading decision and resource assignment. Our objective is to minimize the energy consumption of all UEs, with considering the delay constraint as well as the dynamic resource requirements of heterogeneous computation tasks. Accordingly, we formulate the problem as a mixed integer non-linear programming problem (MINLP), and propose a value iteration based Reinforcement Learning (RL) approach, named Q-Learning, to obtain the optimal policy of computation offloading and resource allocation. Simulation results demonstrate that the proposed approach can significantly decrease UEs' energy consumption in different scenarios, compared with other baseline methods.

Existing Data Center Network (DCN) architectures are classified into two categories: switch-centric and server-centric architectures. In switch-centric DCNs, routing intelligence is placed on switches; each server usually uses only one port of the Network Interface Card (NIC) to connect to the network. In server-centric DCNs, switches are only used as cross-bars, and routing intelligence is placed on servers, where multiple NIC ports may be used. In this paper, we formally introduce a new category of DCN architectures: thedual-centricDCN architectures, where routing intelligence can be placed on both switches and servers. The dual-centric philosophy can achieve various tradeoffs in designing DCN architectures. We propose three novel dual-centric DCN architectures: FCell, FRectangle, and FSquare, all of which are based on the folded Clos topology. FCell is a power-efficient DCN architecture, with a larger diameter and lower bisection bandwidth than FSquare and FRectangle. FSquare is a high performance DCN architecture, in which the diameter is small and the bisection bandwidth is large; however, the DCN power consumption per server in FSquare is high. FRectangle significantly reduces the DCN power consumption per server, compared to FSquare, at the sacrifice of some networking performances. By investigating FCell, FRectangle and FSquare, and by comparing them with existing architectures, we demonstrate that, the three novel dual-centric architectures enjoy the advantages of both switch-centric designs and server-centric designs, have various nice properties for practical data centers, and provide flexible tradeoff choices in designing DCN architectures.

Modern cloud computing systems have been using multiple processing units on servers to increase their processing capability. Recently, applications with multiple parallelization options have been witnessed, and serve as a promising model for efficiently utilizing the processing capacity of the system. In this paper, we consider utility-based scheduling for periodic multisegment tasks with multiple parallelization options on platforms with multiple homogeneous processing units. Our goal is to maximize the system's overall utility achieved by scheduling the tasks. We show that the problem is closely related to another problem, which minimizes the density of each task separately. We consider two typical types of utility models, namely, a uniform utility model, where all tasks have equal utility, and a general utility model, where all tasks have different utility values. For the uniform utility model, we give the optimal solution for selecting and scheduling tasks. For the general utility model, we prove that the problem can be reduced to the classic 0-1 knapsack problem, and thus is NP-complete, we then provide the Fully Polynomial Time Approximation Scheme (FPTAS) for the problem. FPTAS algorithms are known for high time complexity, especially if we want to achieve near-optimal solutions. We then provide a simple 1/2 approximation algorithm based on a greedy strategy with significantly reduced time complexity. Simulations show that tasks with multiple parallelization options can improve system utility significantly, comparisons show that the 1/2 approximation algorithm can achieve near-optimal solutions under general conditions.

We propose a novel Data Center Network (DCN) architecture, named FCell, which is a tradeoff design in three aspects. First, FCell reflects a tradeoff between DCN power consumption and network performances, which mainly include end-to-end delays and bisection bandwidth. We propose a unified path length definition to characterize end-to-end delays in general DCNs. Comparisons with existing DCN architectures reveal that FCell consumes a moderate amount of power, and achieves generally low end-to-end delays and a satisfiable bisection bandwidth. Second, FCell reflects a tradeoff between switch-centric and server-centric designs. Two basic routing schemes are proposed to show that FCell can place routing intelligence on both servers and switches; thus, FCell can be regarded as a dual-centric architecture, which enjoys both the fast switching capability of switches and the high programmability of servers. Third, FCell reflects a tradeoff between scalability and flexibility. Scalability of FCell comes from its regularity; FCell also supports flexible growth of network size with minimal modifications on its original architecture. Through simulations, we evaluate the performances of the two routing schemes in different traffic conditions in FCell, and verify that our unified path length definition is a useful metric to characterize end-to-end delays in general DCNs.

Existing Data Center Network (DCN) architectures are classified into two categories: switch-centric and server-centric architectures. In switch-centric DCNs, routing intelligence is placed on switches, each server usually uses only one port of the Network Interface Card (NIC) to connect to the network. In server-centric DCNs, switches are only used as cross-bars, and routing intelligence is placed on servers, where multiple NIC ports may be used. In this paper, we formally introduce a new category of DCN architectures: the dual-centric DCN architectures, where routing intelligence can be placed on both switches and servers. We propose two typical dual-centric DCN architectures: FSquare and Rectangle, both of which are based on the folded Clos topology. FSquare is a high performance DCN architecture, in which the diameter is small and the bisection bandwidth is large, however, the DCN power consumption per server in FSquare is high. Rectangle significantly reduces the DCN power consumption per server, compared to FSquare, at the sacrifice of some performances, thus, Rectangle has a larger diameter and a smaller bisection bandwidth. By investigating FSquare and Rectangle, and by comparing them with existing architectures, we demonstrate that, these two novel dual-centric architectures enjoy the advantages of both switch-centric designs and server-centric designs, have various nice properties for practical data centers, and provide flexible choices in designing DCN architectures.

During the past decade, various novel data center network (DCN) architectures have been proposed to meet various requirements of large scale data centers. In existing works that consider server-centric DCN architectures, the lengths of a server-to-server-direct hop and a server-to-server-via-a-switch hop are assumed to be equal. We propose the concept of Normalized Switch Delay (NSD) to distinguish a server-to-server-direct hop and a server-to-server-via-a-switch hop, to unify the design and analysis of server-centric DCN architectures for interconnecting servers with two network interface cards. In [1], the authors claim that BCN is the largest known architecture to interconnect dual-port servers, with diameter 7, given a switch port number. We notice that the existing DPillar [2] architecture accommodates more servers than BCN does under the same configurations. Motivated by this fact, we consider a fundamental problem: maximizing the number of dual-port servers, given network diameter and switch port number; and give an upper bound on this maximum number. Then, we propose three novel architectures that try to achieve this upper bound: SWCube, SWKautz, and SWdBruijn, based on the generalized hypercube, Kautz graph, and de Bruijn graph, respectively. The number of servers that SWCube can accommodate is comparable to that of DPillar. The number of servers that SWKautz and SWdBruijn can accommodate is generally greater than that of DPillar. Compared with three existing architectures, the three proposed architectures, SWCube, SWKautz and SWdBruijn demonstrate various advantages. Analysis and simulations on the newly proposed architectures also show that they have nice properties for DCNs, such as low diameter, high bisection width, good fault-tolerance, and the capability of efficiently handling network congestion.

Heterogeneous multiprocessors have been widely used in modern computational systems to increase the computing capability. As the performance increases, the energy consumption in these systems also increases significantly. Dynamic Voltage and Frequency Scaling (DVFS) is considered an efficient scheme to achieve the goal of saving energy, because it allows processors to dynamically adjust their supply voltages and/or execution frequencies to work on different power/energy levels. In this paper, we consider scheduling non-preemptive frame-based tasks on DVFS-enabled heterogeneous multiprocessor platforms with the goal of achieving minimal overall energy consumption. We consider three types of heterogeneous platforms, namely, dependent platforms without runtime adjusting, dependent platforms with runtime adjusting, and independent platforms. For these three platforms, we first formulate the problems as binary integer programming problems, and then, relax them as convex optimization problems, which can be solved by the well-known interior point method. We propose a Relaxation-based Iterative Rounding Algorithm (RIRA), which tries to achieve the task set partition, that is closest to the optimal solution of the relaxed problems, in every step of a task-to-processor assignment. Experiments and comparisons show that our RIRA produces a better performance than existing methods and a simple but naive method, and achieves near-optimal scheduling under most cases. We also provide comprehensive complexity, accuracy and scalability analysis for the RIRA approach by investigating the interior-point method and by running specially designed experiments. Experimental results also show that the proposed RIRA approach is an efficient and practically applicable scheme with reasonable complexity.

As the performance of modern multi-core processors increases, the energy consumption in these systems also increases significantly. Dynamic Voltage and Frequency Scaling (DVFS) is considered an efficient scheme for achieving the goal of saving energy. In this paper, we consider scheduling a set of independent aperiodic tasks, whose release times, deadlines and execution requirements are arbitrarily given, on DVFS-enabled multi-core processors. Our goal is to meet the execution requirements of all the tasks, and to minimize the overall energy consumption on the processor. Instead of seeking optimal solutions with high complexity, we aim to design lightweight algorithms suitable for real-time systems, with good performances. By applying a subinterval-based method, we come up with a simple algorithm to allocate tasks' available execution times during a heavily overlapped subinterval based on their desired execution requirement during that subinterval. Based on the allocated available execution times, we further consider the final frequency setting and task scheduling, which guarantee that all tasks meet their execution requirements, and tries to minimize the overall energy consumption. Extensive simulations for various platform and task characteristics and evaluations using a practical processor's power configuration indicate that our proposed algorithm has a good performance in terms of saving processor energy, though it has low complexity. Besides, the proposed algorithm is easy to be implemented in practical systems.

We consider the design and analysis of Data Center Network (DCN) architectures for interconnecting dual-port servers. Unlike existing works, we propose the concept of Normalized Switch Delay (NSD) to distinguish a server-to-server-direct hop and a server-to-server-via-switch hop, to unify the design of DCN architectures. We then consider a fundamental problem: maximizing the number of dual-port servers, given network diameter and switch port number; and give an upper bound on this maximum number. Two novel architectures are proposed: SWCube and SWKautz, based on the generalized hypercube and Kautz graph, respectively, which in most cases accommodate more servers than BCN [1], which was claimed to be the largest known architecture. Compared with three existing architectures, SWCube and SWKautz demonstrate various advantages. Analysis and simulations also show that SWCube and SWKautz have nice properties for DCNs, such as low diameter, good fault-tolerance, and capability of efficiently handling network congestion.

Two important components that consume the majority of IT power in data centers are the servers and the Data Center Network (DCN). Existing works fail to fully utilize power management techniques on the servers and in the DCN at the same time. In this paper, we jointly consider VM placement on servers with scalable frequencies and flow scheduling in the DCN, to minimize the overall system's power consumption. Due to the convex relation between a server's power consumption and its operating frequency, we prove that, given the number of servers to be used, computation workloads should be allocated to severs in a balanced way, to minimize the power consumption on servers. To reduce the power consumption of the DCN, we further consider the flow requirements among the VMs during VM allocation and assignment. Also, after VM placement, flow consolidation is conducted to reduce the number of active switches and ports. We notice that, choosing the minimum number of servers to accommodate the VMs may result in high power consumption on servers, due to servers' increased operating frequencies. Choosing the optimal number of servers purely based on servers' power consumption leads to reduced power consumption on servers, but may increase power consumption of the DCN. We propose to choose the optimal number of servers to be used, based on the overall system's power consumption. Simulations show that, our joint power optimization method helps to reduce the overall power consumption significantly, and outperforms various existing state-of-the-art methods in terms of reducing the overall system's power consumption.

Modern computational systems have adopted heterogeneous multiprocessors to increase their computation capability. As the performance increases, the energy consumption in these systems also increases significantly. Dynamic Voltage and Frequency Scaling (DVFS), which allows processors to dynamically adjust their supply voltages and execution frequencies to work on different power/energy levels, is considered an efficient scheme to achieve the goal of saving energy. In this paper, we consider scheduling frame-based tasks on DVFS-enabled heterogeneous multiprocessor platforms with the goal of achieving minimal overall energy consumption. We consider three types of heterogeneous platforms, namely, dependent platforms without runtime adjusting, dependent platforms with runtime adjusting, and independent platforms. For all of these three platforms, we first introduce a Relaxation-based Naive Rounding Algorithm (RNRA), which can produce good solutions for some cases, but may be unstable under other situations. Then, we propose a Relaxation-based Iterative Rounding Algorithm (RIRA). Experiments and comparisons show that our RIRA produces a better performance than RNRA and other existing methods, and achieves near-optimal scheduling under most cases.