Dataflow Models of Computation (MoCs) are widely used in embedded systems, including multimedia processing, digital signal processing, telecommunications, and automatic control. In a dataflow MoC, an application is specified as a graph of actors connected by FIFO channels. One of the first and most popular dataflow MoCs, Synchronous Dataflow (SDF), provides static analyses to guarantee boundedness and liveness, which are key properties for embedded systems. However, SDF and most of its variants lack the capability to express the dynamism needed by modern streaming applications. In particular, the applications mentioned above have a strong need for reconfigurability to accommodate changes in the input data, the control objectives, or the environment. We address this need by proposing a new MoC called Reconfigurable Dataflow (RDF). RDF extends SDF with transformation rules that specify how and when the topology and actors of the graph may be reconfigured. Starting from an initial RDF graph and a set of transformation rules, an arbitrary number of new RDF graphs can be generated at runtime. A key feature of RDF is that it can be statically analyzed to guarantee that all possible graphs generated at runtime will be consistent and live. We introduce the RDF MoC, describe its associated static analyses, and present its implementation and some experimental results.

Cloud-Radio Access Network is a promising mobile network architecture that centralizes the computing resources in the Baseband Unit pool which adds more flexibility and increases network performance. However, as computing resources are shared among the Radio Remote Heads (RRH) connected to the Baseband Unit pool, efficient scheduling algorithms should be explored to meet the deadlines requirements of RRHs' subframes and to increase the network throughput. In this paper, we propose optimal scheduling algorithms for computing resources along with three heuristics. We test the different algorithms as a function of three metrics. The evaluation is performed under a real traffic model and the results highlight the importance of choosing the appropriate scheduling algorithm to increase network performance.

Cloud-Radio Access Network (C-RAN) is a promising mobile network architecture that is becoming the foundation of 5G wireless network. It permits to centralize the computing resources in the Baseband Unit (BBU) pool which adds more flexibility and increases network performance. However, as computing resources are now shared among the Radio Remote Heads (RRHs) connected to the BBU pool, efficient scheduling algorithms should be explored in order to meet the deadlines requirements of RRHs' subframes and to increase network throughput. In this paper, we propose optimal scheduling algorithms for computing resources along with three heuristics. We test the different algorithms as a function of three performance metrics such as the offered throughput, the computing resources occupancy and the number of non-decoded subframes. The evaluation is performed under a real traffic model for the incoming subframes. The obtained results highlight the importance of choosing the appropriate scheduling algorithm and bring recommendations to mobile network operators on the best scheduling algorithm that should be adopted to increase network performance.

Dataflow Models of Computation (MoCs) are widely used in embedded systems, including multimedia processing, digital signal processing, telecommunications, and automatic control. In a dataflow MoC, an application is specified as a graph of actors connected by FIFO channels. One of the most popular dataflow MoCs, Synchronous Dataflow (SDF), provides static analyses to guarantee boundedness and liveness, which are key properties for embedded systems. However, SDF (and most of its variants) lacks the capability to express the dynamism needed by modern streaming applications. In particular, the applications mentioned above have a strong need for reconfigurability to accommodate changes in the input data, the control objectives, or the environment. We address this need by proposing a new MoC called Reconfigurable Dataflow (RDF). RDF extends SDF with transformation rules that specify how the topology and actors of the graph may be reconfigured. Starting from an initial RDF graph and a set of transformation rules, an arbitrary number of new RDF graphs can be generated at runtime. A key feature of RDF is that it can be statically analyzed to guarantee that all possible graphs generated at runtime will be consistent and live. We introduce the RDF MoC, describe its associated static analyses, and outline its implementation.

Cloud RAN (C-RAN) is a very promising architecture for future mobile network deployment, where the cloud-centric approach is useful in improving total processing load. In this context, radio and baseband network functions processing pose interesting problems that we try to expose and solve in this paper. A novel architecture for C-RAN and a first modeling of the system are proposed. Furthermore, we study the impact of many radio parameters on the processing time. Moreover, a mathematical and a deep learning model are proposed and evaluated for processing time prediction. Results show the feasibility of the proposed approaches.

This paper considers the optimal placement of Baseband Unit (BBU) function split in Cloud Radio Access Networks (C-RANs) which is an essential key technology in C-RAN deployment. In particular, the BBU function split is modeled as directed chains to be mapped to a network infrastructure. As such, we propose an Integer Linear Program (ILP) formulation for small and medium size networks. Alternatively, we introduce four heuristic algorithms with significantly less complexity. We then benchmark the four heuristic algorithms based on the construction of a multi-stage graph. The simulation results strongly confirm the efficiency and scalability of our algorithms as well as their ability to achieve an optimal solution.

Dataflow Models of Computation (MoCs) are widely used in embedded systems, including multimedia processing, digital signal processing, telecommunications, and automatic control. In a dataflow MoC, an application is specified as a graph of actors connected by FIFO channels. One of the most popular dataflow MoCs, Synchronous Dataflow (SDF), provides static analyses to guarantee boundedness and liveness, which are key properties for embedded systems. However, SDF (and most of its variants) lacks the capability to express the dynamism needed by modern streaming applications. In particular, the applications mentioned above have a strong need for reconfigurability to accommodate changes in the input data, the control objectives, or the environment. We address this need by proposing a new MoC called Reconfigurable Dataflow (RDF). RDF extends SDF with transformation rules that specify how the topology and actors of the graph may be reconfigured. Starting from an initial RDF graph and a set of transformation rules, an arbitrary number of new RDF graphs can be generated at runtime. A key feature of RDF is that it can be statically analyzed to guarantee that all possible graphs generated at runtime will be consistent and live. We introduce the RDF MoC, describe its associated static analyses, and outline its implementation.

Network Functions Virtualization (NFV) Point of Presence (PoP) Data Centers are often constrained by compute and storage capacity and the cost of energy required to operate the data centers. High energy cost is a general concern for NFV operators, and in particular for specific-purpose NFV PoP DCs such as those in mobile core networks. In this context, optimized resource management and workload distribution based on a domain-agnostic policy engine for driving energy efficiency in data centers is proposed. An open stack based solution is proposed to enable policy-based monitoring and energy management. The specified policies are used to enforce soft and hard constraints in the system with periodic event monitoring and dynamic resource management to minimize energy consumption.

More and more, users store their data in the cloud. While the content is then retrieved, the retrieval has to respect quality of service (QoS) constraints. In order to reduce transfer latency, data is replicated. The idea is make data close to users and to take advantage of providers home storage. However to minimize the cost of their platform, cloud providers need to limit the amount of storage usage. This is still more crucial for big contents. This problem is hard, the distribution of the popularity among the stored pieces of data is highly non-uniform: several pieces of data will never be accessed while others may be retrieved thousands of times. Thus, the trade-off between storage usage and QoS of data retrieval has to take into account the data popularity. This paper presents our architecture gathering several storage domains composed of small-sized datacenters and edge devices; and it shows the importance of adapting the replication degree to data popularity. Our simulations, using realistic workloads, show that a simple cache mechanism provides a eight-fold decrease in the number of SLA violations, requires up to 10 times less of storage capacity for replicas, and reduces aggregate bandwidth and number of flows by half.

It is estimated that the ICT industry contains not less than one billion personal computers (PC), with power consumption ranging from 100 to 200 W per workstation. Virtualization and Thin Client technologies, mediated by broadband network links will have a very significant impact on efforts to reduce energy consumption and to enhance data security throughout the ICT industry. Besides, virtualization is quickly getting into the small and medium-sized business (SMB) market, promising a better (centralized) energy management with means of consolidation and resource reuse. Indeed, multiple virtual machines (VM) or platforms, running Virtual Desktops (VD), can be hosted on a single physical server, requiring only hypervisors and adequate desktop delivery protocols to deliver the VD service to distant users on any suitable device. In this paper, we propose to deliver a VD service over a federation of data centers (DC), sparsely connected to some nodes of the core network. The proposed model aims to provide an adequate resource dimensioning at the DC and the network while minimizing the inherent energy consumption. We prove by means of simulations the importance of user profiling and activity prediction in avoiding over-dimensioning, especially at the DCs. We also show the interest of energy-aware modes at the servers.

Delivering on-demand web content to end-users in order to carry out strict QoS metrics is not a trivial task for globally distributed network providers. This task becomes still harder when content popularity varies over the time and the SLA definitions have to include both transfer rate and latency metrics. Current worldwide content delivery approaches and datacenter infrastructures rely on cumbersome replication schemes that are agnostic to edge-network resources, and damage content provision. In this work we present AREN, an novel replication scheme for cloud storage on edge networks. AREN relies on a collaborative cache strategy and bandwidth reservation to adapt the replication degree according to strict SLA contracts and content popularity growth. We have evaluated the performances of replication schemes on edge networks using Caju, a content distribution system for edge networks. Compared to a non-collaborative caching, evaluations show that AREN prevents nearly 99.8% of all SLA violations when the storage system is heavily loaded. We also show that AREN provides a sevenfold decrease in the amount of storage usage for replicas, and it increases by roughly 20% the aggregate bandwidth, hence accelerating content delivery.

The explosion of user generated data along with the evolution of web 2.0 applications (e.g. social networks, blogs, podcasts, etc.) has resulted in a tremendous demand for storage. With cloud computing posing as a possible all-in-one solution, "storage clouds" focus on providing distributed storage capability. We discuss the creation of a storage cloud using edge devices, based on Peer-to-Peer resource provisioning. In this approach, mobile phones, PCs/Media Centers, Set-top-boxes, modems and networked storage devices can all contribute as storage within these storage clouds. Combining all end-user edge devices may result in a scalable, very flexible storage capability that keeps the data comparatively close to the user, increasing availability, while reducing latency. This work addresses the issue of Quality of Service (QoS)-aware scheduling in a P2P storage cloud, built with edge devices by designing an optimization scheme that minimizes energy from a system perspective and simultaneously maximizing user satisfaction from the individual user perspective.

The use of economic models to handle resource allocation in the grid is a reality. Auction mechanisms are used in grid and distributed testbeds to elicitate user's preferences while improving the performance of the system. Existing implementations rely on single auction mechanisms to allocate resources whilst grids are heterogeneous in nature. Many different applications cohabitate having different workflow requirements such as bag of tasks executions or real time interaction that cannot be dealt efficiently by a single mechanism. In that scenario it would be more realistic to allocate resources using the most suitable mechanism for each situation, enabling also its configuration, deployment and management at runtime. Therefore, the paper presents the design principles, architecture and implementation of a configurable auction server (CAS). The auction server offers support for the deployment, configuration and execution of different auction mechanisms, facilitating the task of market mechanism developers and enabling the execution of distributed marketplaces according to local demand.

The Grid is a promising concept to solve the dilemma of increasingly complex and demanding applications being confronted with the need for a more efficient and flexible use of existing computer resources. Even though Grid technologies have made progress within the context of large enterprises and academic projects, there has not yet been a widespread adoption by public institutions and small enterprises. One barrier to this adoption is the lack of economic paradigms which support the dynamic and efficient sharing of Grid resources by balancing resource scarcity and idle capacities. Economic algorithms promise to provide a good fit to the Grid's inherent strategic dimension by enabling users to express their valuation for computer resources. At the same time they provide incentives to contribute idle resources to the Grid in return for the market price.This paper presents a market-based approach for trading complex computational Grid services. The implemented combinatorial exchange aims at maximizing the social welfare of users. At its core, it provides a rich bidding language which is able to represent complex Grid services and simple workflows. The allocation mechanism is evaluated by means of a numerical experiment in order to gain detailed insights into the computational complexity of the underlying allocation problem. Our analysis provides input for the configuration of Grid markets.

In grid systems, users compete for different types of resources such that they may execute their applications. Traditional grid systems are formed of organisations that join together for the purpose of collaborative projects. Resources of each of the participating organisation are pooled such that members of individual organisations may access the shared infrastructure. In general, each participant is both a provider and a consumer of resources. Whilst such systems address large organisations, in this paper we address democratic grid systems to satisfy needs of small organisations and even individuals, where on-demand grids may be formed by drawing idling resources available on the Internet. Whilst traditional grid systems resort to allocations that satisfy system specific objectives such as maximization of the resource utilisation, market mechanisms try to obtain allocations that are efficient economically. Economic mechanisms permit to achieve equilibrium between supply and demand and furthermore provide incentives for providers. Combinatorial auction has been argued as an effective mechanism to address the problem of resource allocation within grid systems. Auctions within which multiple types of resources in varying quantities may be traded eliminate the exposure problem by addressing co-allocation. In this paper, we describe a combinatorial exchange where multiple providers and multiple consumers may participate. We describe the winner determination problem that incorporates the time dimension, i.e. resource bundles may be requested for different time ranges, and describe a set of heuristics that have been designed to be fast. We show that these achieve a high level of efficiency as compared to exact solutions. The second part focusses on the pricing problem. The objective is to compute prices that represent the state of the market and bring trustworthy feedback to participants. Drawing on the approach taken by Kwasnica et al. (Manage Sci 51(3):419–434, 2005 ), we propose a pricing model that computes per-item pricing. Per-item pricing allows users to deduce the price of bundles that they require by linear summation. Furthermore, we propose a model that computes prices as a function of time, thus permitting users, in particular consumers to adjust their demand trading off price and time of execution.

Large scale systems such as the Grid need scalable and efficient resource allocation mechanisms to fulfil the requirements of its participants and applications while the whole system is regulated to work efficiently. Economics inspired models have shown to efficiently allocate resources and services, scaling up well as they are decentralized. However most of existing implementations rely on a single market mechanism as a mean to handle resource allocation. Besides, in most of those approaches the allocation mechanism itself is neither flexible nor configurable, mostly designed for specific purposes such as scheduling one type of tasks. Nowadays, Grids are heterogeneous systems composed of multiple cohabitating resources and applications. Such heterogeneity requires complex mechanisms and usually cannot be achieved by a single type of allocation mechanism. Therefore, we aim to develop a generic approach to resource allocation in Grids able to support multiple cohabitating auctions. We claim that resource allocation frameworks may deal with heterogeneity by means of flexibility and configurability and they have to provide functionalities by which the allocation mechanism should be configured and adapted to application requirements and resource providers needs. This paper presents a configurable auction server architecture that enables dynamic configuration of markets so as to adapt them to the requirement of their initiators.

Economic models have shown their suitability to allocate resources efficiently, considering an unbalanced supply and demand. As the use of the Grid is extending, a numerous set of distributed resource allocation frameworks have been developed to attain efficient resource management while keeping the scalability of the Grid. However, those frameworks make use of either simple double auction mechanisms or complex approximations to the NP-complete problem of the combinatorial auction. The problem of those mechanisms is that of its generality, that is, they have not been specially designed for the trading of time-leased computational resources. In this paper we present a novel variant of the double auction that has been specially adapted to trade time-differentiated resources as Grid resources can be considered. The paper presents the data structures, algorithms and architecture of the economic mechanism as well as it presents the evaluation of the mechanism through simulation. Simulated results are compared with the main double auction implementations found in the literature. The paper constitutes an approach to improve efficiency of resource allocation in the Grid from the point of view of the economic model and not from architectural aspects addressed by most of the contributions found in the literature.

This chapter aims at discussing issues concerning the advertisement and semantic discovery of Web services in a democratized Grid environment: an environment in which users are agnostic of the low-level details for managing the services offered and requested. This type of environment poses new requirements, and thus, it affects the functionality of a service advertisement/discovery system. In the context of this aim, the chapter presents the motivation and the technologies developed towards a semantic information system in the Grid4All environment. The chapter emphasizes on bridging the gap between Semantic Web and conventional Web service technologies, supporting developers and ordinary users to perform resources’ and services’ manipulation tasks, towards a democratized Grid.

Large scale systems such as the Grid need scalable and efficient resource allocation mechanisms to fulfil the requirements of its participants and applications while the whole system is regulated to work efficiently. Economics inspired models have shown ability to handle efficiently the allocation of resources and services, scaling up well as they are decentralized. Our model considers the arbitration of decisions at the local scope and short term, the regulation of the system at global scope, and the sharing of information between global and local environments. This paper presents a scalable model and evaluates by simulation a system where global market information circulates in aggregated and scalable form, the rate of demand by participants is globally regulated by a currency mechanism, preference is regulated by a reputation mechanism, and local regulation among competing participants is resolved by auction mechanisms. The paper shows how scalable systems benefit from distributed marketplaces supporting global information flow to regulate and optimize local and global behavior.