Cloud-Fog computing has emerged as an essential paradigm to support the growing demand for real-time data processing driven by the Internet of Things. By integrating the extensive computing capabilities of cloud data centres with the low-latency benefits of fog nodes, this architecture increases resource utilisation and improves quality of service. However, the dynamic and heterogeneous nature of cloud fog environments poses significant workflow scheduling challenges, especially when optimising multiple trade-offs such as latency, cost, energy consumption, and resource utilisation. This paper investigates the many-objective dynamic workflow scheduling problem under deadline constraints in container-based cloud-fog computing environments (MDWS-CoCF). Unlike existing studies that primarily focus on horizontal scaling, this work considers both vertical and horizontal scaling of containers, allowing for real-time adjustments of container configurations based on task-specific requirements. To address this complex problem, we first develop a dynamic workflow scheduling simulator that models real-world scenarios, including a variety of task categories and container scalability. Based on this simulator, we propose a Cooperative Coevolution Genetic Programming (CCGP) approach that evolves specialised heuristics for task selection, resource allocation, and container deployment to facilitate adaptive and efficient scheduling in MDWS-CoCF. Extensive simulations using real-world data traces show that the proposed CCGP approach significantly outperforms existing baseline algorithms, achieving superior performance as measured by the HyperVolume and Inverted Generational Distance metrics. The results show that the evolved heuristics are robust and effective under different dynamic scenarios, ensuring balanced optimisation of many objectives.

In virtual machine (VM) deployment, the physical machine (PM) usually first retrieves VM image files from the central image server through block transfer, and the VM image retrieval and communication are the two main factors that consume network bandwidth resources, In this paper, we propose a heuristic-based algorithm to minimize both image retrieval cost and communication cost for VM placement in a fat-tree network. It consists of three phases: PM clustering, VM partitioning, 1) We first cluster the PMs based on the possible longest communication distance, which is estimated by a pre 2) To reduce the traffic between PM clusters, a semidefinite programming algorithm is used to place the coarsened VMs to PM Here coarsening means packing the resources of smaller VMs as a whole, so as to accelerate the solving process. 3) In each PM cluster, the VMs are mapped to PMs one by one, and the VMs with common blocks and communication traffic between each other are more likely to be placed together. Extensive simulations show that our algorithm is more effective and efficient than the state-of-the-art.

Multi-cloud is a promising paradigm due to its advantages such as avoiding vendor lock-in and optimising costs. This article focuses on dynamic flexible workflow scheduling with minimum total monetary cost in multi-clouds, considering multiple categories of services for each cloud with different configurations and billing methods. Existing studies generally ignore the characteristics and states of each individual cloud when making schedules, which may be ineffective regarding cost savings and quality of service. To address this issue, we propose to introduce a cloud selection decision on top of the existing task selection and resource selection decisions to help us select appropriate resource for task in an overall cost-effective cloud. To automatically learn the task, cloud and resource selection rules simultaneously, we propose a new genetic programming with multi-tree representation based on a customised discrete event-driven dynamic workflow scheduling simulator. Simulation results based on two real-world data traces show that the proposed algorithm performs significantly better than the state-of-the-art algorithms in terms of reducing the rental costs and deadline deviation, and improving the success rate. The results also show that the superiority of the proposed algorithm lies in the ability to select an appropriate cloud resource for a task.

Smart grid is widely deployed in the modern smart city. However, the frequent network attacks on smart grid today have raised concerns about the privacy and security of individual fine-grained information collection. Moreover, once the secret key held by the data control center is disclosed, any adversary could easily break the indistinguishability of ciphertexts, thereby revealing users privacy. To this end, we propose an enabling efficient multidimensional encrypted data aggregation scheme for fog-cloud-based smart grid by using the improved Boneh-Goh-Nissim cryptosystem and the super-increasing vector technique, which can realize efficient multi-source multidimensional encrypted data aggregation in the fog-cloud architecture. The security analysis demonstrates that our scheme can satisfy security requirements of smart grid, i.e. it achieves the usage data privacy protection and confidentiality, multidimensional aggregation data integrity, and key-disclosure resistance. The performance evaluations show that our scheme is more efficient than related data aggregation schemes.

It is undeniable that rumors abound on online social networks and rumors can cause many disastrous consequences. Effective controlling of rumors is of great significance in social networks. However, the existing research only selects boosting users who are more likely to adopt the truth or select blocking users to terminate the spread of rumors. The former tends to correct the rumor after the spread is over but with high controlling cost, while the latter blocks the rumor without considering the truth transmission. In this article, we focus on how to select boosting–blocking users to control rumors when the rumor and truth are spreading together. We propose a boosting-truth blocking-rumor cascade (BTBRC) model. Under this model, given the rumor seed set and truth seed set, the boosting rumor controlling (BRC) problem aims to find a boosting–blocking seed set with$k$users such that the number of users influenced by the truth can be maximized. In order to solve it, we design a multihop neighbor boosting (MHNB) algorithm, which can get effective results with a data-parameter-dependent approximation ratio. Based on the above model, we also propose a positive boosting-truth blocking-rumor cascade (PBTBRC) model and design a connected multihop neighbor boosting (CMHNB) algorithm to solve the connected positive boosting rumor controlling (CPBRC) problem that requires a seed set to be connected under this model. Finally, extensive theoretical analysis and experimental results show the superiority of our algorithms over other comparison methods.

Cloud computing is an emerging computational infrastructure for cost-efficient workflow execution that provides flexible and dynamically scalable computing resources at pay-as-you-go pricing. Workflow scheduling, as a typical NP-Complete problem, is one of the major issues in cloud computing. However, in the cloud scenario with unlimited resources, how to generate an efficient and economical workflow scheduling scheme under the deadline constraint is still an extraordinary challenge. In this paper, we propose a hybrid heuristic algorithm called enhanced task type first algorithm (ET2FA) to solve deadline-constrained workflow scheduling in cloud with new features such as hibernation and per-second billing. The objectives to be minimized include the total cost and total idle rate. ET2FA involves three phases: 1) Task type first algorithm, which schedules tasks based on topological level and task types, and utilizes a compact-scheduling-condition based VM selection method to assign each task. 2) Delay operation based on block structure, which further optimizes total cost and total idle rate based on block structure properties. 3) Instance hibernate scheduling heuristic, which sets an instance to hibernate if idle for a duration. Extensive simulation experiments based on seven well-known real-world workflow applications show that ET2FA delivers better performance in comparison to the state-of-the-art algorithms.

Much work has already been studied on the interrelation between the epidemic spreading and awareness spreading to prevent infections in a social network. By selecting seed users to spread awareness, we can control epidemic spreading. However, selecting seed users with the maximum influential users may not be the best solution in location-based social networks. Therefore, it is challenging to determine users to spread the information (the awareness of prevention) in these networks. The minimized epidemic infection (MEI) problem aims to find a seed set with$k$seed users such that the infection users can be minimized. In this article, we propose a piecewise function to measure the probability of each user being infected, which considers the distance and time. Then, we propose an algorithm called location-infected-greedy (LIG) to solve theMEIproblem by finding the seed nodes that consider the probability of infection, time of check-in, location information, and influence of users. In the meantime, LIG can obtain an upper bound of the data-dependent approximate ratio, and it runs in$O(kn^{2})$, where$n$is the total number of nodes and$k$is the number of seed nodes. Finally, extensive contrast experiments on real-world location-based social networks show that our algorithm is efficient and effective.

Workflow scheduling is one of the most challenging problems in cloud computing. This paper proposes a heuristic algorithm task type first algorithm (T2FA) for solving deadline-constrained workflow scheduling in cloud with multicore resource (DWS_CMR). The objectives to be minimized are the maximal completion time (i.e., makespan) and the total costs. Firstly, resource model and workflow application model are introduced. Resource model has the configurations of multicore, processing capacity, bandwidth and leasing price, and workflow application model is described by directed acyclic graph (DAG). Based on above models, the mathematical model of DWS_CMR is established, which allows multiple tasks to run concurrently on multicore resources. Secondly, to exploit the characteristics of the problem, the structures of DAG are decomposed and formulated. Merging tasks conforming to the first structure into task blocks can simplify DAG. Four special types of tasks are extracted from the second and third structures, and are preferentially scheduled in task scheduling stage. Then, a new interrelated calculation method of estimated start time and actual start time of tasks is proposed, which can complete the task-to-resource mapping. Finally, T2FA is devised, which incorporates two important phases, including pre-processing and task scheduling. Experimental results show that T2FA can achieve significantly better schedules in most test cases compared to several existing algorithms.

As a critical problem in time series data mining, time series classification (TSC) has always been a challenging task for high-dimensional and high-frequency time sequences in internet of things (IoT), Recently, convolutional neural networks (CNNs) have exhibited great superiority on TSC tasks in deep learning models, but not effective in capturing the long-term temporal dependency of time series. In this paper, we transform time series into GM-images by Gramian Angular Field (GAF) and Markov Transition Field (MTF), which can well preserve the temporal dependency and transition statistics of raw time series, respectively. We further propose an efficient Adaptive Dila-DenseNet (ADDN) to extract various and discernable patterns from GM-images for TSC. In ADDN, we devise an adaptive feature aggregation method for combining static and dynamic information flexibly from GAF and MTF representations. Moreover, inspired by the dilated residual networks, we design the Dila-Dense block in ADDN to preserve local spatial information for GM-images. The significant decrease of GM-images resolution in common deep learning models may lead to performance degradation. Nevertheless, our Dila-Dense block can address this resolution issue without decreasing the receptive field. Experiments evaluated on 24 benchmark datasets demonstrate that our approach shows greater efficiency and efficacy over the compared deep learning baselines, in IoT.

Inspired by the recent discoveries in neuroscience, the study of the sparse binary projection model started to attract people's attention, shedding new light on image retrieval. Different from the classical work that tries to reduce the dimension of the data for faster retrieval speed, the model projects dense input samples into a higher-dimensional space and outputs sparse binary data representations after winner-take-all competition. Following the work along this line, this paper designed a new algorithm which obtains a high-quality sparse binary projection matrix through unsupervised training. Simple as it is, the algorithm reported significantly improved results over the state-of-the-art methods in both search accuracy and retrieval speed in a series of empirical evaluations on large-scale image retrieval tasks, which exhibited its promising potential in industrial applications.

In a cloud data center, servers are always over-provisioned in an active state to meet the peak demand of requests, wasting a large amount of energy as a result. One of the options to reduce the power consumption of data centers is to reduce the number of idle servers, or to switch idle servers into low-power sleep states. However, the servers cannot process the requests immediately when transiting to an active state. There are delays and extra power consumption during the transition. In this paper, we consider using state-of-the-art servers with multi-sleep modes. The sleep modes with smaller transition delays usually consume more power when sleeping. Given the arrival of incoming requests, our goal is to minimize the energy consumption of a cloud data center by the scheduling of servers with multi-sleep modes. We formulate this problem as an integer linear programming (ILP) problem during the whole period of time with millions of decision variables. To solve this problem, we divide it into sub-problems with smaller periods while ensuring the feasibility and transition continuity for each sub-problem through a Backtrack-and-Update technique. We also consider using DVFS to adjust the frequency of active servers, so that the requests can be processed with the least power. Our simulations are based on traces from real world. Experiments show that our method can significantly reduce the power consumption for a cloud data center.

In this paper, we propose a scheduling scheme called GreenFlowing to reduce the electricity cost for a cloud data center by leveraging heterogeneous ESDs. In our model, the data center can be powered by intermittent green energy like wind and solar, and the electricity with time-varying prices from the power grid. The energy from different sources can also choose to flow into long-term or short-term ESDs for later use. Note that, the former can sustain energy for a long time but with low charging/discharging rate, while the latter can charge/discharge very fast but with high energy leakage that can sustain energy only for a few hours. By combining them together, the energy cost can further be reduced. However, it is hard to decide when and how much energy from different sources should be used to power the data center directly or charged into different types of ESDs. We formulate our scheduling into a large-scale linear programming (LP) problem, which can be solved using CPLEX. Numerical experiments show that our scheduling can significantly reduce the total electricity cost for a cloud data center.

In tradition, the problem of Multi-Agent Path Finding is to find paths for the agents without conflicts, and each agent execute one-shot task, a travel from a start position to its destination. However, making just one planning for the agents may not satisfy the requirement in dynamic environments such as logistics sorting center, where the paths of the agents may constantly need to be adjusted according to the incoming tasks. The challenging issue is to dynamically adjust the already planned paths while make planning for the agents ready to execute new incoming tasks. In this paper, we formulate it into Dynamic Multi-Agent Path Finding (DMAPF) problem, the goal of which is to minimize the cumulative cost of paths. To solve this problem, we propose an algorithm called Lifelong Planning Conflict-Based Search (LPCBS), which can efficiently and optimally make planning for the new incoming tasks while adjusting the already planned paths. Experiment results show that the LPCBS performs much better than the existing works in each planning.

In recent years, automated guided vehicle (AGV) is becoming increasingly important for logistic center, which usually has tens of thousands of express packages to sort and transport every day. In order to enhance the efficiency of sorting, multiple AGVs have been used to transport as many express packages as possible in a given time. However, there may exist collisions and deadlocks when two or more AGVs are trying to pass through a shared area at the same time. In this paper, we study collision-free and deadlock-free path planning of multiple AGVs with the objective of minimizing the makespan, that is, the maximum time of the AGVs to finish a round of delivery. To solve the problem, we first propose a candidate path generation algorithm to generate several paths for each AGV, from which a combination of the paths is selected based on optimal path combination selection (OPCS) algorithm, such that the collisions and deadlocks can be minimized. Then, we construct a colored resource-oriented timed Petri net (CROTPN) model for dynamic changing AGV routes after slicing the scheduling period into timeslots. Through Slot-Control policy, the collision and deadlock can be effectively avoided. Our experiment is simulated based on CPN tools. Experiment results show that our method can achieve a high sorting throughput using the least AVGs.

Summary Infrastructure‐as‐a‐Service cloud providers always charge users in 2 ways: on‐demand billing and reservation billing with different discounts. On‐demand billing is more flexible for short‐term use but with higher cost, while reservation billing is much cheaper for long renting period but may cause inefficiency because of underutilized capacity. As middle agent, cloud broker always tries to reserve a large number of virtual machine (VM) instances from cloud providers to meet the requirement of all users, while making high profit through reducing its cost by leveraging long‐term reservations and time‐multiplexing of instances. Different from existing work, we believe that the rental time of each user's request should be served consecutively in one VM instance, rather than partitioned into different VM instances by cloud broker to enhance time‐multiplexing ratio. In this paper, we first try to smooth users' unordered and time‐overlapping requests using demand graph, based on which we propose two reservation algorithms for cloud broker, one for off‐line and the other for online. We also consider cloud broker with multiple service providers, finding that much more cost can be saved by leveraging the various reservation discounts of those providers. Extensive simulations have been done on the traces from real cloud platforms. Experiment results show that the broker cost can be substantially saved using our off‐line and online algorithms.

In this paper, we study the issues of scheduling of user requests to geographically distributed data centers. The data centers are powered by three types of energy: wind, solar, and brown energy. We consider using energy storage devices (ESDs) to store energy when its supply is abundant or the price is low, and discharge it when the energy supply is in short or expensive. By introducing ESDs in data centers, it brings some new challenges regarding the scheduling of user requests, such as when and how much to charge the energy into ESDs and when to discharge it, how this charge and discharge of ESDs affects the QoS and total energy cost. We focus on two optimization problems: 1) Schedule the servers and energy usage from different energy sources, such that the QoS requirement is met and the total energy cost is minimized. 2) Minimize total carbon emissions for given a budget of energy cost. The problems are formulated as mixed integer linear programming problem (MILP). We use Cplex to solve the formulated problems. We have run extensive experiments on traces from real world. Experiments show that our scheduling methods using ESDs can significantly reduce both carbon emissions and total cost.

The high power consumption of cloud data centers has aroused great concern on environmental implications such as global warming. Therefore, cloud service providers like Facebook and Green House Data have built their own wind or solar farms to power the data centers to reduce both energy cost and carbon emissions. In this paper, we propose an optimization- based framework to make planing for green cloud data centers, where the objective functions range from minimizing energy cost to minimizing carbon emissions. Our planning is based on the optimized scheduling of the users' requests to each data center, while considering time-varying and location-varying electricity prices and the supply of sustainable energy under different weather conditions. Our plan includes: 1) How many servers each data center should have. 2) How many wind turbines and solar panels should be used to power each data center. 3) What capacity of energy storage device (ESD) should be equipped for each data center. We formulate each problem as a constraint optimization problem, and solve it using Cplex. Extensive experiments have been done on traces from real world. As far as we know, our work is the first to explore the issue of planning for green cloud data centers through optimized scheduling methods.

In this paper, we study the issue of minimizing the total energy cost for green cloud data centers with time varying and location varying electricity prices and supply of renewable energy. Given the budget of energy cost, schedule the requests, servers, and power usage of different sources, such that the total cost can be minimized. We formulate the problem during the whole period of time as an MILP problem. We use Cplex to solve the problem. Experiment shows that our method can significantly reduce total cost after using ESDs.

Servers and network contribute about 60% to the total cost of data center in cloud computing. How to efficiently place virtual machines so that the cost can be saved as much as possible, while guaranteeing the quality of service plays a critical role in enhancing the competitiveness of service cloud provider. Considering the heterogeneous servers and the random property of multiple resources requirements of virtual machines, the problem is formulated as a multi-objective nonlinear programming in this paper. Virtual machine cluster with higher traffic is made staying together. This reduces the communication delay while saving the inter-server bandwidth consumption, especially the relatively scarce higher level bandwidth, by exploiting the topology information of data center. At the same time, statistic multiplex and newly defined “similarity” techniques are leveraged to consolidate virtual machines. The violation of resource capacity is kept at any designated minimal probability. Thus the quality of service will not be deteriorated while saving servers and network cost. An offline and an online algorithms are proposed to address this problem. Experiments compared with several baseline algorithms show the validity of the new algorithms: more cost is cut down at less computation effort.