Search

Edge computing is an emerging promising computing paradigm that brings computation and storage resources to the network edge, hence significantly reducing the service latency and network traffic. In edge computing, many applications are composed of dependent tasks where the outputs of some are the inputs of others. How to offload these tasks to the network edge is a vital and challenging problem which aims to determine the placement of each running task in order to maximize the Quality-of-Service (QoS). Most of the existing studies either design heuristic algorithms that lack strong adaptivity or learning-based methods but without considering the intrinsic task dependency. Different from the existing work, we propose an intelligent task offloading scheme leveraging off-policy reinforcement learning empowered by a Sequence-to-Sequence (S2S) neural network, where the dependent tasks are represented by a Directed Acyclic Graph (DAG). To improve the training efficiency, we combine a specific off-policy policy gradient algorithm with a clipped surrogate objective. We then conduct extensive simulation experiments using heterogeneous applications modelled by synthetic DAGs. The results demonstrate that: 1) our method converges fast and steadily in training; 2) it outperforms the existing methods and approximates the optimal solution in latency and energy consumption under various scenarios.

As distributed ledgers, blockchains run consensus protocols which trade capacity for consistency, especially in non-ideal networks with incomplete connectivity and erroneous links. Existing studies on the tradeoff between capacity and consistency are only qualitative or rely on specific assumptions. This paper presents discrete-time Markov chain models to quantify the capacity of Proof-of-Work based public blockchains in non-ideal networks. The comprehensive model is collapsed to be ergodic under the eventual consistency of blockchains, achieving tractability and efficient evaluations of blockchain capacity. A closed-form expression for the capacity is derived in the case of two miners. Another important aspect is that we extend the ergodic model to analyze the capacity under strong consistency, evaluating the robustness of blockchains against double-spending attacks. Validated by simulations, the proposed model is accurate and reveals the effect of link quality and the distribution of mining rates on blockchain capacity and the ratio of stale blocks.

The existence of restricted Service Level Agreement (SLA) choices, which typically correspond with a couple of service tiers, can result in a customer accepting a service that may not effectively respond to their needs. From a service provider perspective, it is also a less than optimum business model, with capacity being reserved for customers who will not use it and subsequently being unavailable for customers who would. We, therefore, advocate the use of personalized SLAs to avoid such situations, which can ideally be set up without the assistance of a human operator. We suggest classifying customers according to their distinguishing features, one of which includes a customer’s propensity to have online devices in their home. Through the results presented in this paper, we are confident about the accuracy of our classification results; however, we recognize that there are opportunities for latency improvements in the efficiency of the process.

Multi-access edge computing (MEC) aims to extend cloud service to the network edge to reduce network traffic and service latency. A fundamental problem in MEC is how to efficiently offload heterogeneous tasks of mobile applications from user equipment (UE) to MEC hosts. Recently, many deep reinforcement learning (DRL) based methods have been proposed to learn offloading policies through interacting with the MEC environment that consists of UE, wireless channels, and MEC hosts. However, these methods have weak adaptability to new environments because they have low sample efficiency and need full retraining to learn updated policies for new environments. To overcome this weakness, we propose a task offloading method based on meta reinforcement learning, which can adapt fast to new environments with a small number of gradient updates and samples. We model mobile applications as Directed Acyclic Graphs (DAGs) and the offloading policy by a custom sequence-to-sequence (seq2seq) neural network. To efficiently train the seq2seq network, we propose a method that synergizes the first order approximation and clipped surrogate objective. The experimental results demonstrate that this new offloading method can reduce the latency by up to 25% compared to three baselines while being able to adapt fast to new environments., Accepted by IEEE Transaction on Parallel and Distributed Systems

The ten papers in this special section are devoted to the most recent developments and research outcomes addressing the related theoretical and practical aspects of computational intelligence solutions in sustainable computing and aims at presenting latest innovative ideas targeted at the corresponding key challenges, either from a methodological or from an application perspective.

MEC is an emerging paradigm that utilizes computing resources at the network edge to deploy heterogeneous applications and services. In the MEC system, mobile users and enterprises can offload computation-intensive tasks to nearby computing resources to reduce latency and save energy. When users make offloading decisions, the task dependency needs to be considered. Due to the NP-hardness of the offloading problem, the existing solutions are mainly heuristic, and therefore have difficulties in adapting to the increasingly complex and dynamic applications. To address the challenges of task dependency and adapting to dynamic scenarios, we propose a new DRL-based offloading framework, which can efficiently learn the offloading policy uniquely represented by a specially designed S2S neural network. The proposed DRL solution can automatically discover the common patterns behind various applications so as to infer an optimal offloading policy in different scenarios. Simulation experiments were conducted to evaluate the performance of the proposed DRL-based method with different data transmission rates and task numbers. The results show that our method outperforms two heuristic baselines and achieves nearly optimal performance.

Datacenters are critical to the commercial and social activities of modern society but are also major electricity consumers. To minimize their environmental impact, we must make datacentres more efficient whilst keeping the quality of service high. In this work we consider how a key datacenter component, Virtual Network Functions (VNFs), can be placed in the datacenter to optimise the conflicting objectives of minimizing service latency, packet loss and energy consumption. Multiobjective Evolutionary Algorithms (MOEAs) have been proposed to solve the Virtual Network Function Placement Problem (VNFPP), but state of the art algorithms are too slow to solve the large problems found in industry. Parallel Multiobjective Evolutionary Algorithms (PMOEAs) can reduce execution time by distributing the optimization process over many processes. However, this can hamper the search process as it is inefficient to share information between processes. This paper aims to determine whether PMOEAs can efficiently discover good solutions to the VNFPP. We found that PMOEAs can solve the VNFPP 5–10\(\times \) faster than a sequential MOEA without harming solution quality. Additionally, we found that one parallel algorithm, PPLS/D, found better solutions than other MOEAs and PMOEAs in most test instances. These results demonstrate that PMOEAs can solve the VNFPP faster, and in some instances better, than sequential MOEAs.

Mobile Cloud Computing (MCC) is regarded as a promising method to increase the data storage and enhance the processing power of mobile devices. Technologies such as Software Defined Networking (SDN) and Network Function Virtualisation (NFV) will be deployed in MCC to simplify the network management and accelerate mobile service deployment. In order to achieve a deeper understanding of future MCC, we developed a comprehensive analytical model to investigate the performance of MCC in the presence of both NFV service chains and SDN networks. The model is capable of capturing the interactions between SDN and NFV when they share the same underlying physical infrastructure. The end-to-end latency is derived for different scales of service deployments and network configurations. Comprehensive simulation experiments are conducted and the results demonstrate that the proposed analytical model corresponds well with the simulation experiments. In addition, we show how the analytical model can be a useful tool to investigate the impact of centralised SDN control on the performance of NFV traffic transmission.

The ever increasing demand for computing resources has led to the creation of hyperscale datacentres with tens of thousands of servers. As demand continues to rise, new technologies must be incorporated to ensure high quality services can be provided without the damaging environmental impact of high energy consumption. Virtualisation technology such as network function virtualisation (NFV) allows for the creation of services by connecting component parts known as virtual network functions (VNFs). By optimising the placement and routing of VNFs this technique can be used to maximally utilise available datacentre resources, to maintain a high quality of service whilst minimising energy consumption. Current research on this problem has focussed on placing VNFs and considered routing as a secondary concern. In this work we argue that the opposite approach, a routing-led approach is preferable. We propose a novel routing-led algorithm and analyse each of the component parts over a range of different topologies on problems with up to 16000 variables and compare its performance against a traditional placement based algorithm. Empirical results show that our routing-led algorithm can produce significantly better solutions to large problem instances on a range of datacentre topologies.

Human embodiment/representation in virtual environments (VEs) similarly to the human body in real life is endowed with multimodal input/output capabilities that convey multiform messages enabling communication, interaction and collaboration in VEs. This paper assesses how effectively different types of virtual human (VH) artefacts enable smooth communication and interaction in VEs. With special focus on the REal and Virtual Engagement In Realistic Immersive Environments (REVERIE) multi-modal immersive system prototype, a research project funded by the European Commission Seventh Framework Programme (FP7/2007-2013), the paper evaluates the effectiveness of REVERIE VH representation on the foregoing issues based on two specifically designed use cases and through the lens of a set of design guidelines generated by previous extensive empirical user-centred research. The impact of REVERIE VH representations on the quality of user experience (UX) is evaluated through field trials. The output of the current study proposes directions for improving human representation in collaborative virtual environments (CVEs) as an extrapolation of lessons learned by the evaluation of REVERIE VH representation.

Ranging from web caches to firewalls, network functions play a critical role in modern networks. Network function virtualisation (NFV) has gained significant interests from both industry and academia, thus making the study of their placement an active research topic. Due to multiple criteria that must be considered by stake holders, e.g. the minimisation of the end-to-end latency and overall energy consumption, the NFV placement problem is in principle a multi-objective optimisation problem. This paper develops a formal model for the NFV placement problem based on queuing theory. By using the popular NSGA-II as the optimiser, the effectiveness of the proposed model is validated through a series of proof-of-concept experiments. In particular, some genetic operators have been developed to match the characteristics of the problem.

Content caching is a promising approach in edge computing to cope with the explosive growth of mobile data on 5G networks, where contents are typically placed on local caches for fast and repetitive data access. Due to the capacity limit of caches, it is essential to predict the popularity of files and cache those popular ones. However, the fluctuated popularity of files makes the prediction a highly challenging task. To tackle this challenge, many recent works propose learning based approaches which gather the users' data centrally for training, but they bring a significant issue: users may not trust the central server and thus hesitate to upload their private data. In order to address this issue, we propose a Federated learning based Proactive Content Caching (FPCC) scheme, which does not require to gather users' data centrally for training. The FPCC is based on a hierarchical architecture in which the server aggregates the users' updates using federated averaging, and each user performs training on its local data using hybrid filtering on stacked autoencoders. The experimental results demonstrate that, without gathering user's private data, our scheme still outperforms other learning-based caching algorithms such as m-epsilon-greedy and Thompson sampling in terms of cache efficiency.

Information-centric networking (ICN) is an emerg- ing paradigm that decouples content from the host to achieve fast and cost-efficient communication and content distribution in the future Internet. A key feature of ICN is the deployment of ubiquitous in-network caching to speed up service delivery and improve network resource utilisation. ICN caching has been widely studied in terms of caching strategies and caching performance. However, the economic aspect of ICN has received marginal consideration so far, although it is vital to understand the potential cost- efficiency of ICN before its wide deployment in service provider network. To address this issue, we propose a cost- aware caching scheme to study the Quality-of-Service (QoS) and cost of ICN and investigate the inner association between them. Two new models are designed to characterise the cost and QoS of ICN with arbitrary topology under heterogeneous bursty content requests. A multi- objective evolution algorithm is adopted to find the optimal cache resource allocation. Numerical results show the effectiveness of the proposed scheme in achieving cost- efficiency and QoS guarantee in ICN caching.

Large network operator environments are composed of thousands of nodes and devices capable of performing multiple roles. This network infrastructure is multi-layered, multi-vendor and underpinned by a high capacity and complex optical transport network. Managing this network requires millions of lines of configuration files and hundreds of Operational Support Systems. Typically, the management data structure uses a hierarchical namespace containing tens of thousands of object identifiers (OID). Each OID identifies a variable that can be read, modified or set via management protocol. The British Telecom network collects many millions of OIDs every 10 minutes, and executes many thousands of configuration changes per month via many tools, and multiple generations of protocols, data models and software components. Blending Software Defined Network (SDN) model-driven management and Network Functions Virtualisation (NFV), for on-demand (scale-in and scale-out) virtual network functions (Big Data nodes, network heuristics and analytics), provides an exciting opportunity for significant operational savings: reduced outage impact, simplification of management stack, fault correlation and network healing, and network usage trending for efficient resource allocation and scaling. This paper and talk outlines the management plane challenges and use cases for complex tier-1 optical environments. It discusses how we need to rethink network analytics and embrace streaming telemetry for real-time resource adaptation. It outlines a guiding network telemetry framework being developed by leading operators and the enabling community-driven technologies, and how they may be used to facilitate the programmable management plane for optical transport infrastructure.

The use of serious games in education and their pedagogical benefit is being widely recognized. However, effective integration of serious games in education depends on addressing two big challenges: the successful incorporation of motivation and engagement that can lead to learning; and the highly specialised skills associated with customised development to meet the required pedagogical objectives. This paper presents the Westminster Serious Games Platform (wmin-SGP) an authoring tool that allows educators/domain experts without games design and development technical skills to create bespoke roleplay simulations in three dimensional scenes featuring fully embodied virtual humans capable of verbal and non-verbal interaction with users fit for specific educational objectives. The paper presents the wmin-SGP system architecture and it evaluates its effectiveness in fulfilling its purpose via the implementation of two roleplay simulations, one for Politics and one for Law. In addition, it presents the results of two types of evaluation that address how successfully the wmin-SGP combines usability principles and game core drives based on the Octalysis gamification framework that lead to motivating games experiences. The evaluation results shows that the wmin-SGP: provides an intuitive environment and tools that support users without advanced technical skills to create in real-time bespoke roleplay simulations in advanced graphical interfaces; satisfies most of the usability principles; and provides balanced simulations based on the Octalysis framework core drives. The paper concludes with a discussion of future extension of this real time authoring tool and directions for further development of the Octalysis framework to address learning.

The rapid development in multimedia services has shifted the major function of the current Internet from host-centric communication to service-oriented content dissemination. Motivated by this significant change, Information-Centric Networking (ICN) has emerged as a new networking paradigm, which aims at providing natural support for efficient information retrieval over the Internet. As a crucial characteristic of ICN, in-network caching enables users to efficiently access popular content from ubiquitous caches to improve the Quality-of-Experience (QoE). Therefore, in-network caching for ICN has received considerable attention in recent years and many cache schemes and models have been proposed. However, there is a lack of research into ICN cache models under practical environments such as arbitrary topology and multimedia services exhibiting bursty nature. To bridge the gap, this paper proposes a new analytical model to gain valuable insight into the caching performance of ICN with arbitrary topology and bursty content requests. The accuracy of the proposed model is validated by comparing the analytical results with those obtained from simulation experiments. The analytical model is then used as a cost-efficient tool to investigate the impact of key network and content parameters on the performance of caching in ICN.

Networks today integrate multiple technologies, allowing network infrastructure to deliver a variety of services to support the different characteristics and dynamic demands of applications. There is an increasing goal to make the network responsive to service requests issued directly from the application layer and high-layer client interfaces. This differs from the established model where services in the network are instantiated in response to management commands driven by a human user using a wide variety of Operational Support Systems (OSS), and where networks are typically over-provisioned to ensure minimal traffic loss, even at peak traffic periods.

Editorial for the Personal Ubiquitous Computing: Special Issue on Virtual Environments and Advanced Interfaces