In spite of their commercial success, Cloud services are still subject to two major weak points: data security and infrastructure resiliency. In this paper, we propose an original Cloud network architecture aiming at improving the resiliency of Cloud network infrastructures interconnecting remote data centers. The main originality of this architecture consists in exploiting the principles of Software Defined Networking (SDN) in order to adapt the rerouting strategies in case of network failure according to a set of requirements. In existing Cloud networks configurations, network recovery after a fiber cut is achieved by means of the usage of redundant bandwidth capacity preplanned through backup links. Such an approach has two drawbacks. First, it induces at a large scale a non-negligible additional cost for the Cloud Service Providers (CSP). Second, the pre-computation of the rerouting strategy may not be suited to the specific quality of service requirements of the various data flows that were transiting on the failing link. To prevent these two drawbacks, we propose that CSPs deploy their services in several redundant data centers and make sure that those data centers are properly interconnected via the Internet. For that purpose, we propose that a CSP may use the services of multiple (typically two) Internet Service Providers to interconnect its data centers via the Internet. In practice, we propose that a set of “routing inflection points” may form an overlay network exploiting a specific routing strategy. We propose that this overlay is coordinated by a Software Defined Networking-based centralized controller. Thus, such a CSP may choose the network path between two data centers the most suited to the underlying traffic QoS requirement. The proposed approach enables this CSP a certain independency from its network providers. In this paper, we present this new Cloud architecture. We outline how our approach mixes concepts taken from both SDN and Segment Routing. Unlike the protection techniques used by existing CSPs, we explain how this approach can be used to implement fast rerouting strategy for inter-data center data exchanges.

Electric Vehicles (EV) appear as a clean transportation technology with clear environmental advantages and potential benefits for the grid, being a key element of future modern smart cities. However, autonomy of the vehicle and lack of recharging stations are barriers that need to be overcome in order to make the service reliable and broadly accepted. To tackle this problem, French GreenFeed project is working to define and implement an interoperable and universal architecture to allow EV-recharge roaming. In this work, we consider such architecture and focus on issues faced by the e-mobility operator (EMO) and identified by the GreenFeed project. The EMO, a key actor in the architecture, maintains contracts with EV users to allow EV recharge at any geographical place (recharge roaming) through agreements with charging infrastructure operators. The interest of the EMO is to schedule the recharge demands and fulfil the contracts while maximizing its revenue. We analyse different online EV recharge scheduling under different pricing models, agreed between the EV users and the EMO, to study their impact on the EMO's revenues. In this analysis, we assume recharges arriving at random times and requesting a certain amount of energy during a fixed time period. We simulate a scenario with real day-ahead hourly electricity prices over two years and different scheduling polices to illustrate the feasibility of online recharge scheduling.

Cloud Brokers enable interoperability and portability of applications across multiple Cloud Providers. On the other hand, incoming Cloud Providers start to support more and more unbundled Cloud Instances offerings. Thus, consumers may set up at their will the quantity of CPU, network bandwidth and memory or hard disk capacities their Cloud Instances will have. These facts enable the standardization of interoperable Cloud Instance configurations. In this paper, CompatibleOne is presented as an approach to bring Cloud Computing as a commodity. For this, the requirements to make of a product a commodity have been identified and have been mapped into the CompatibleOne architecture components. Our approach shows the practical feasibility of bringing Cloud Computing as a commodity.

In Cloud computing, resources such as CPU, RAM, and disk etc. are provided as a service via the Internet. Elasticity is a key feature of Cloud Computing. It aims to enabling a Cloud Service Provider (CSP) to negotiate the possibility to borrow external resources from other CSPs when its own facilities are not able to satisfy a client request. Reciprocally, a CSP may sell some of its unused resources to another CSP in case of under load. Today, Cloud Federation is a key approach considered for Cloud elasticity. In the context of the Easi-Clouds European ITEA 2 research project, we aim to develop Pricing-as-a-Service (PraaS) suited to the Federated Cloud environment. The aim of this paper is twofold. First, we provide a state of the art of the pricing and revenue sharing models in federated environment. We discuss the specifications of Cloud Federation as well as its different drivers and barriers. Three types of pricing strategies (on-demand, spot and reserved) are presented. In the second part, we present the problem of revenue sharing in the federation with some properties we are willing to fulfill. We propose and evaluate our revenue sharing model suited to the Federated environment with numerical analysis via simulation. We compare our approach to the proportional share and the Shapley value method. Finally, we provide and analyze the results of our simulations with our conclusion and perspectives.

Currently, Cloud brokers bring interoperability and portability of applications across multiple Clouds. In the future, Cloud brokers will offer services based on their knowledge of Cloud providers infrastructure to automatically and cost-effectively overcome performance degradation. In this paper, we present a Mixed-Integer Linear Program (MILP) that provides a cost-effective placement across multiple Clouds. Our MILP formulation considers parameters of Cloud providers such as price, configuration of VMs, network latency, and provisioning time. We evaluate the cost-effectiveness of deploying a Cloud infrastructure into a single or across multiple Cloud providers by using real prices and VM configurations. The results show that in some cases may be cost-effective to distribute the infrastructure across multiple Cloud providers. We also propose three placement policies for faulty multi-Cloud scenarios. The best of these policies minimizes the cost of the Cloud infrastructure under fixed provisioning time values.

Provisioning ubiquitous broadband wireless services necessitates the duplication of the radio equipment or/and the reduction of the size of the radio cells. However, both approaches are not economically viable for the operators. The concept of distributed antennas is proposed as an alternative solution offering low-cost broadband access. In this matter, we propose an innovative Radio Access Network (RAN) architecture called GeRoFAN for Generic Radio-over-Fiber Access Network. The GeRoFAN architecture is based on a transparent optical loop connecting low-cost 4G radio-system antennas using analog Radio-over-Fiber (RoF) technology. However, by transporting Radio Frequencies (RFs) over an Optical Channel (OC), the composite signal suffers from various physical layer impairments that may degrade the system capacity. Supported by a mathematical model reflecting the relevant optical layer limitations, we propose in this paper an innovative control plane that targets to optimize the global radio cellular capacity of GeRoFAN by efficiently mapping RFs onto OCs. This problem is formulated as an Integer Linear Program (ILP) with the objective to minimize the number of OCs subject to 4G cellular planning rules.

Radio-over-Fiber (RoF) emerges as a promising solution to achieve cost-effective radio-antennas backhauling. The limited dynamic range of analog optical transmission is known as one of the most penalizing limitations of analog RoF links. Thus, without a special care to RoF signal impairments, the wireless system capacity may be seriously degraded. In this paper, we propose an impairment-aware RoF-based cellular backhauling system called GeRoFAN aiming to preserve the capacity of the radio system while using efficiently the optical capacity. Supported by an analytical modeling of RoF physical impairments, we develop an evolutionary multi-objective heuristic called PaGeO (Pareto-based Genetic Optimization) exploiting the concept of Pareto optimization to determine the best backhauling strategy for the GeRoFAN control plane. We outline through numerical results that our impairment-aware backhauling solution computed by PaGeO, applied to LTE radio cellular system, outperforms alternative backhauling policies. We also highlight the impact of RF cellular planning on the tradeoff optimization achieved by PaGeO.

In Cloud environment, Cloud Providers (CP) grant access to computing and storage resources through Web-portals. Resource virtualization is a key enabler for providing such services. Thanks to virtualization, multiple Virtual Machines (VMs) can be hosted by the same Physical Machine (PM). Existing CPs, such as Amazon Web Services and Windows Azure, offer pre-configured VMs with fixed-size computing, storage and network capacities. In this context, end-users can only choose from a set of predefined VM instances offered by the CPs. However, it is expected that, in the near-future, end-users will be able to access the Cloud without any restriction on the size of the required resources. They will be charged according to the amount of resources used. In such scenario, the major problem faced by a CP is to select the appropriate PM that will host a new VM while still satisfying the end-user requirements. This resource allocation problem is similar to the well-known online Bin Packing problem. In this paper, we investigate several algorithms that were proposed to solve the Bin Packing problem, and compare them in a Cloud environment in terms of resource utilization and percentage of dropped VMs. The novel concept of Resource Over-Reservation (ROR) is introduced as a mean to reduce the percentage of dropped VMs and to improve resource utilization.

In Cloud environments, efficient resource provisioning and management present today a challenging issue because of the dynamic nature of the Cloud on one hand, and the need to satisfy heterogeneous resource requirements on the other hand. In such dynamic environments where end-users can arrive and leave the Cloud at any time, a Cloud service provider (CSP) should be able to make accurate decisions for scaling up or down its data-centers while taking into account several utility criteria, e.g., the delay of virtual resources setup, the migration of existing processes, the resource utilization, etc. In order to satisfy both parties (the CSP and the end-users), an efficient and dynamic resource allocation strategy is mandatory. In this paper, we propose an original approach for dynamic resource allocation in a Cloud environment. Our proposal considers computing job requests that are characterized by their arrival and teardown times, as well as a predictive profile of their computing requirements during their activity period. Assuming a prior knowledge of the predicted computing resources required by end-users, we propose and investigate several algorithms with different optimization criteria. However, prediction errors may occur resulting in some cases in the drop of one or several computing requests. Our proposed algorithms are compared in terms of various performance parameters including the rejection ratio, the dropping ratio, as well as the satisfaction of the endusers and the CSP.

Power-awareness in carriers' networks has become a major concern in the past few years due to the increasing power consumption required by high speed electronics. Today, the scalability of the Internet is not only limited by communication technology but also by operational expenditure induced by electrical power needed to feed and cool switching/routing and transmission equipment. Designing low power network equipment is certainly the most effective solution to reduce unnecessary power dissipation due to the Joule effect. In the short-term, alternative solutions targeting the existing infrastructures must be provided until the emergence of green-networking technologies. Recently, innovative approaches based on traffic rerouting and grooming have been proposed in the literature for core WDM networks. Transceivers used at the optical switching nodes for add-drop and grooming purposes are the key elements conditioning power consumption. In current networks, transceivers are powered-on permanently, whatever the activity of the data sources at the electrical layer. In this paper, we aim to minimize network's power consumption assuming transceivers with fixed data rates. In this context, we propose an optimal mapping of a set of scheduled traffic demands at the electrical layer onto the optical network. The number of optical channels per optical fiber being upper-bounded, this mapping also minimizes connections' rejection. This optimization problem is formulated as an Integer Linear Program.

Since the early 2000s, a real attention is paid to physical impairments arising in large-scale optical networks. One cost-effective solution to cope with transmission impairments is to deploy 3R (re-amplifying, re-shaping, and re-timing) regenerators in a limited number of network nodes (i.e., translucent networks). Taking into account the simultaneous effect of four transmission impairments (amplified spontaneous emission, chromatic dispersion, polarization mode dispersion, and nonlinear phase shift), we propose a novel exact approach for impairment-aware network planning. In contrast with previous works, we investigate the problem under pre-planed dynamic traffic. Our proposal takes advantage of the dynamics of the traffic pattern so that regeneration resources may be shared among non-concurrent requests. Given a network topology and a set of pre-planned requests, we target the minimum number of regenerators or/and regeneration sites. Obtained results emphasize the mutual impact between requests' time-correlation and regenerator concentration.

Cloud services are based on the provisioning of computing, storage, and networking resources in order to satisfy requests generated by remote end-users. High speed Internet access and multi-core Virtual Machines (VMs) enable today the provisioning of diversified and enriched types of services in Cloud environment. In this paper, we consider several types of basic services and show how their orchestration may lead to the provisioning of more sophisticated services. For this purpose, we define four types of requests that cover the wide spectrum of possible services. We then formulate the resource provisioning problem as a Mixed Integer Linear Program (MILP). We assume that the underlying infrastructure is based on a set of end-to-end connections with guaranteed sustainable bandwidth such as Carrier-Grade Ethernet (CGE) circuits. We investigate the impact of two innovative services on resource allocation carried out by a Cloud Service Provider (CSP). These services correspond to distributed data storage and to multicast data transfer. For the former service, we consider the possibility of splitting a storage request onto different remote storage nodes. The latter service aims to distribute a same data sequence from one server towards multiple remote nodes assuming a limited number of network nodes have multicast capacities. These two innovative services provide a gain of 7% in terms of accepted requests when applied to the 18-node NSFnet backbone network.

Because of the very high optical fiber's capacity, prompt and unambiguous fiber cut detection is mandatory to guarantee carriers' network survivability. In this matter, the monitoring-cycle and the monitoring-trail mechanisms have been proposed in the recent literature. In this paper, we propose the innovative concept of monitoring-tree that enables to reduce considerably the monitoring cost while keeping unambiguous single fiber cut detection and localization. We propose an ILP formulation aiming at mapping a monitoring-tree onto a mesh topology. Applied to real networks, we show that our approach outperforms its counterparts.

Many studies have been carried out in the field of translucent network design during the last years. Translucent networks use sparse regeneration strategy where electrical regeneration is performed at a limited number of network nodes in order to improve the signal budget. The objective of translucent network design is to judiciously choose the regeneration sites to overcome the signal degradation and to maintain a predefined quality of transmission. Pachnicke et al. have recently proposed an approach for routing and regenerator placement. In this paper, we propose a novel heuristic called COR2P for translucent network design. Our concern is both to minimize the number of required regenerators and to concentrate them in a limited number of sites. In addition, we propose an improved version of Pachnicke's algorithm under the name of RP-CBR+ and we proceed to the comparison of these two approaches. On one hand, both COR2P and RP-CBR+choose, a priori, a set of regeneration sites, before the effective RWA phase. Our approach is both traffic and topology-driven since it considers the effective traffic matrix and the limited capacity of the network. Oppositely, regeneration sites placement in RP-CBR+ is only topology-driven since it considers a connection demand between any pair of nodes under unlimited network capacity. On the other hand, COR2P differs from RP-CBR+in three points: the order in which traffic demands are processed, the wavelength assignment strategy and the possibility or not to use additional regeneration sites. Simulation results show that unlike RP-CBR+, COR2P accepts all traffic demands under low traffic load, using the same number of regenerators. Under high traffic load, our algorithm presents much lower blocking ratios than RP-CBR+at the price of a larger number of regenerators.

The emergence of new grid services requires the definition of suited business models taking into account the expectations of both end-users and resource providers. Basically, such business models consider the problem of computing, storage and network resources virtualization. The optimization of resource utilization may be carried out either in discrete time or in continuous time. In previous works, we proposed a discrete time approach enabling a better utilization of both network and computing resources in the context of layer-2 virtual private networks (VPN). In this paper, we extend this analysis by integrating to our model a new grid entity, namely the grid resource broker (GRB) as a middleware entity between end-users and resource providers. We compare the economic gain of both the GRB and resource providers via several criteria for three different pricing strategies. In addition, we define the best adopted pricing strategy to the GRB and study its impact on resource providers.

Grid services provisioning requires the definition of suited business models taking into account the expectations of both end-users and service providers. Traditional resource management techniques mainly deal with end-users requirements but neglect the benefit that service providers may expect from the sharing of their equipments. In this paper, we propose a fairer management algorithm defending the interests of both end-users and service providers. We consider scheduled jobs in the context of layer-2 Virtual Private Networks (VPN). In practice, the instant of activation of a scheduled job may vary over a certain time window. Depending on the requirements and the budgets of end-users, the size of this time window may fluctuate, the job computing delay being fixed a priori. Our motivation is two-fold. First, we expect to serve a higher number of end-users while respecting each user's job utility. Second, we aim to increase as best as possible the benefit of service providers. For that purpose, we introduce a weighted cost function enabling a service differentiation relying on time constraints disparity of the submitted jobs. Two approaches are considered to solve this problem: an exact Integer Linear Programming (ILP) formulation and a metaheuristic inspired from the Simulated Annealing (SA) algorithm. The obtained numerical results outline the impact of the flexible time window on the system performance, namely the number of executed jobs and the gain of the service provider.

Grid networks have captured a lot of attention in recent years because of their potential to generate new applications thanks to network, computing and storage resources virtualization. Quality of service (QoS) is a key issue for grid services provisioning. Current grid services are provided on multi service networks such as the Internet. Thus, QoS architectures originally developed for the Internet such as DiffServ (DS) have been tested in grid environment. Since grid network services based on Internet networks will be developed in the next years, we propose in this paper to investigate the potentialities of an innovative Internet QoS architecture known as flow-aware networking (FAN). FAN appears as a promising alternative to DS for QoS provisioning in IP networks. DS proceeds to traffic differentiation and QoS provisioning through IP packet marking whereas FAN consists in implicit IP flow differentiation and a flow-based admission control. A grid session may be seen as a succession of parallel TCP flows with voluminous data transfers. In this paper, we compare by means of computer simulations the performance of FAN and DS architectures under grid environment. Two metrics are adopted for that purpose: the average transit delay and the average goodput of a grid session in an IP access router.

Abstract A point‐to‐point quantum key distribution (QKD) system takes advantage of the laws of quantum physics to establish secret keys between two communicating parties. Compared to the classical methods, such as public‐key infrastructures, QKD offers unconditional security, which makes it attractive for very high security applications. However, this unprecedent level of security is mitigated by the inherent constraints of quantum communications, such as the limited rates and ranges of an individual point‐to‐point QKD link. A QKD network, which can be built by combining multiple point‐to‐point QKD devices, can alleviate the constraints and enable point‐to‐multi‐point key distribution based on QKD technology. The European project, secure communication based on quantum cryptography (SeCoQC) aims at deploying a prototype QKD network, which will be demonstrated in September 2008, by developing the architecture and the protocols, as well as the specific hardware for long‐range QKD networks. This paper discusses the important aspects of the architecture and the network layer protocols of the SeCoQC QKD network. Copyright © 2008 John Wiley & Sons, Ltd.

This paper describes extensive research in the field of the Multipoint Control Protocol (MPCP) for the Ethernet Passive Optical Network (EPON). More in particular the dynamic bandwidth allocation (DBA) algorithm IPACT (Interleaved Polling with Adaptive Cycle Time) is analytically modeled. The main focus has been on analyzing cycle times and packet delay in detail. The analysis starts from a simple traffic model with Poisson arrival rates, symmetric load and fixed packet size. Based on the obtained insights, more general cases with asymmetric traffic load, packet size distribution, self-similar traffic and differentiated services are further elaborated.

Recently several investigations in the field of multilayer traffic engineering (TE) mainly focusing on random traffic demands have been carried out. In this paper, we propose different TE strategies for EXC/OXC networks including electrical traffic grooming and lightpath rerouting. Unlike existing investigations that consider only random demands for TE, we consider simultaneously scheduled and random traffic demands. Scheduled demands correspond to guaranteed services whereas random demands correspond to best-effort services. In order to outline the impact of traffic predictability on TE efficiency, we introduce the new concept of semi-random traffic demand that corresponds to random demands with predictable holding time. For a given network and a given traffic scenario, numerical simulations outline the huge gap between the rejection ratio of random demands and that of semi-random demands. In order to improve the rejection ratio of random demands, we propose to proceed for such demands to a time limited resource reservation (TLRR). The choice of the value of reservation window Delta strongly depends on the arrival and life duration processes of random demands. We show that it is possible to achieve comparable rejection ratios for both random and semi-random demands

In WDM optical networks connection requests can be classified into three different types: Permanent Lightpath Demands (PLDs), Scheduled Lightpath Demands (SLDs), and Random Lightpath Demands (RLDs). These three types of traffic demands are considered for network planning, network engineering and traffic engineering respectively. In this paper we investigate the Routing and Wavelength Assignment (RWA) problem in all-optical WDM networks operating under the wavelength continuity constraint by jointly considering PLDs, SLDs and RLDs. We aim to model a realistic situation where an operator wishes to employ its optical network initially designed for PLDs and/or SLDs to offer on the fly lightpath provisioning. To the best of our knowledge, our global approach has not already been investigated in the literature. According to the considered timescale, various types of optimization tools are used. Numerous numerical results are presented and discussed to outline the benefits of such a global perspective.

Most of the studies dealing with dynamic lightpath establishment (DLE) consider a single type of traffic demand. These demands are in general generated randomly without any information about their life duration. In this paper, one considers for DLE a mix of preplanned traffic and of random traffic. Preplanned traffic corresponds for instance to optical virtual private networks (OVPN). Two approaches for the acceptance of random traffic demands are compared. The first approach consists in evaluating the availability of the required resources only at the instant of arrival of the random demand. The second approach consists in accepting the new random demand only if sufficient resources are available over a given period of time Delta. This time limited reservation imposes to consider at call setup preplanned resources over the period Delta. Indeed, such resources cannot be used by a random demand. The life duration of a random demand being unknown, either its extinction date occurs during the period Delta, or later. In this last case, a rerouting procedure may be necessary to enable the random demand survivability. In summary, the originality of our approach consists in providing a time-limited guaranteed service to random traffic demands instead of a pure best effort service, as it is the case in general.

Over the last decade, numerous routing and wavelength assignment (RWA) algorithms have been developed for WDM optical networks planning. Most of these algorithms neglect the feasibility of the obtained lightpaths. In this paper, we propose a new algorithm, called LERP (Lightpath Establishment with Regenerator Placement), that enables to solve the problem of RWA in guaranteeing the feasibility of the obtained solution. A lightpath is said admissible if the BER (Bit Error Rate) at its destination node remains acceptable (remains under a given threshold). In the case of BER non-admissibility, one or more electrical regenerators may be placed along the lightpath. The LERP algorithm aims at minimizing the number of regenerators necessary to guarantee the quality of transmission along the lightpath. The originality of the our approach consists in considering simultaneously four physical layer impairments, namely, chromatic dispersion, polarization mode dispersion, amplified spontaneous emission and non-linear phase shift. The efficiency of the LERP algorithm is demonstrated via a numerical comparison with one of the alternative solutions proposed in the literature. Numerical simulations have been carried out in the context of the 18-node NSF network.

In this paper, we investigate the routing problem for dynamic traffic demands providing differentiated types of service across the network. Two classes of traffic are considered: QoS traffic and best-effort traffic. We propose a comprehensive offline/online routing solution that enables service providers to satisfy the demands of customer traffic flows while respecting the traffic priority. As a first step, an offline global optimization tool is used for network dimensioning. It aims to minimize the number of optical and electrical ports required in the network in order to satisfy a given set of high priority demands. Once the network's nodes are dimensioned, the network can be over-dimensioned to carry additional best-effort traffic requests. The latter set of traffic is routed by means of an online sequential optimization tool. Low priority traffic are divided into two subsets according to the requested capacity of the demands. This capacity may correspond either to an optical channel data rate (lightpath demands) or to a fraction of that rate (electrical demands). Our sequential optimization tool proceeds automatically to the grooming of the electrical demands into grooming demands. The network performance is expressed as the ratio of the number of rejected/blocked requests to the total number of requests

In this article we assess the economic gain obtained by introducing multi-granularity switching optical cross-connects (MG-OXCs) in an optical transport network. An MG-OXC is an equipment able to switch wavelengths, to aggregate (disaggregate) wavelengths into (from) groups of wavelengths called wavebands and to switch wavebands. We evaluate the economic gain by comparing the cost of instantiating (i.e., provisioning) a set of Scheduled Lightpath Demands (SLDs) in a network of wavelength-switching optical cross-connects (WXCs) against the cost of instantiating them in a network of MG-OXCs. We formulate the problems of instantiating SLDs in both a network of WXCs and a network of MG-OXCs as two combinatorial optimization problems and compute approximate solutions to these problems using Tabu Search (TS) based algorithms. Finally, we define a method to determine the conditions (traffic distribution, cost of switching matrices and ports, etc.) under which a network of MG-OXCs is more economical than a network of WXCs

Virtual concatenation (VCAT) is an SDH/SONET network functionality recently standardized by the ITU-T. VCAT provides the flexibility required to efficiently allocate network resources to Ethernet, fiber channel, ESCON and other important data traffic signals. The aim of this paper is to quantify the savings in network resources provided by VCAT, with respect to contiguous concatenation (CCAT), in SDH/SONET mesh transport networks bearing protected scheduled connection demands (SCDs). We define mathematical models to quantify the add/drop and transmission resources required to instantiate a set of protected SCDs in either a VCAT-capable or a CCAT-capable network. Quantification of transmission resources requires a routing and slot assignment (RSA) problem to be solved. We formulate the RSA problem in VCATand CCAT-capable networks as two different combinatorial optimization problems: RSAv and RSAc, respectively. Protection of the SCDs is considered in the formulations using a shared backup path protection (SBPP) technique. We propose a simulated annealing (SA) based meta-heuristic algorithm to compute approximate solutions to these problems (i.e., solutions whose cost is close to the cost of the optimal ones). The gain in transmission resources and the cost structure of add/drop resources making VCAT-capable networks more economical are analyzed for different realistic traffic types (ESCON, GbE, fibre channel, voice).

Two main classes of traffic demands may be considered in data networks: virtual private networks (VPN) and random traffic requests. In this paper, we investigate traffic engineering for such traffic classes in the context of hybrid WDM optical networks. A hybrid optical network is made of a set of multi-granularity switches, each electrical cross-connect (EXC) of the network being coupled to an optical cross-connect (OXC). Our approach differs from existing approaches by combining global optimization tools with iterative optimization tools to solve the routing and wavelength assignment (RWA) problem. For this reason, a simulated annealing meta-heuristic has been developed to route lightpath demands associated to VPN traffic. On the other hand, random demands are routed sequentially by means of a Dijkstra algorithm. Hence, space-time correlation between demands as well as sub-wavelength traffic grooming have been considered in our RWA strategy.

We present algorithms that compute the routing and wavelength assignment (RWA) for scheduled lightpath demands in a wavelength-switching mesh network without wavelength conversion functionality. Scheduled lightpath demands are connection demands for which the setup and teardown times are known in advance. We formulate separately the routing problem and the wavelength assignment problem as spatio-temporal combinatorial optimization problems. For the former, we propose a branch and bound algorithm for exact resolution and an alternative tabu search algorithm for approximate resolution. A generalized graph coloring approach is used to solve the wavelength assignment problem. We compared the proposed algorithms to an RWA algorithm that sequentially computes the route and wavelength assignment for the scheduled lightpath demands.

We investigate the problem of routing a set of lightpath demands for which the start and end dates may be planned. We call this type of requests foreseeable lightpath demands or FLDs. In a transport network, FLDs correspond, for example, to clients' requests for pre-provisioned bandwidth capacity such as fixed-bandwidth pipes for bulk data transfers during the night, extra VPN bandwidth used during peak office working time, etc. Since in some cases the FLDs are not all simultaneous in time, it is possible to reuse physical resources to realize time-disjoint demands. We propose a routing algorithm that takes into account this property to minimize the number of required WDM channels in the physical links of the network. The gain (in term of saved resources) provided by the algorithm, when compared to a shortest path routing strategy, depends both on the spatial and temporal structure of the set of traffic demands and on the structure of the physical network. The routing problem is formulated as a combinatorial optimization problem. A tabu search meta-heuristic algorithm is developed to solve this problem.

The provision of broad-band services based on either the Internet or the asynchronous transfer mode (ATM) technique requires a new generation of access networks. In the short term, solutions such as x-digital subscriber line (DSL) and hybrid fiber coaxial (HFC), allowing the reuse of existing infrastructures look very promising. For the longer term, new infrastructures based either on radio or on optical access links seem preferable. Three x-DSL techniques aiming at high-bit-rate transmission over twisted pairs are presented: high-bit-rate DSL, asymmetrical DSL, and very-high-bit-rate DSL. An extension of existing cable television networks known as HFC is also described. Two other prospective approaches, wireless in the loop (WITL) and fiber in the loop (FITL), are then presented. Several techniques are considered for WITL: digital enhanced cordless telecommunication local multipoint distribution service, wireless ATM, and low Earth orbit satellite constellations. The various architectural alternatives for the FITL approach are discussed, a special interest being dedicated to the synchronous digital hierarchy self-healing loop and the ATM over a passive optical network. We mention the main experiments and the standardization activities inherent to the domain.

Cell delay variation is one of the major disadvantages of the ATM technique. In broadband ISDN, cell propagation delay along a virtual connection fluctuates because of the variable delays introduced in the network by the queues at switches and multiplexers. The authors consider the case of a constant bit rate source (CBR) accessing B-ISDN from a DQDB metropolitan area network. They first underline the necessity of a peak cell rate enforcement entity at the UNI between the DQDB network and the ATM access. They then investigate the suitability of a new traffic shaper so-called spacer-controller to police peak cell rate at the output of a DQDB/B-ISDN gateway. The case of a DQDB-ATM multi-connection environment is also considered.>

In B-ISDN, a policing function such as usage parameter control (UPC) is mandatory to constraint the cell flow of a new connection to respect its contract. The case of a user accessing B-ISDN via a DQDB metropolitan area network (MAN) is considered. One assumes that this user generates a variable bit rate (VBR) traffic towards the user network interface (UNI). We investigate the influence of the internal DQDB traffic on the time structure of this VBR cell flow when it enters the DQDB/B-ISDN gateway. We then consider two UPC functions which are the leaky bucket and the spacer-controller to police the VBR connection. The efficiency of each of these two UPC mechanisms is evaluated in terms of the cell delay variation and dynamics. The influence of several source parameters such as the mean bit rate, the peak bit rate and the burstiness on UPC efficiency is investigated.>

The Metaring High speed local area network (HSLAN) is based on a full-duplex slotted ring using spatial reuse, which allows throughputs several times higher than the medium capacity. The Metaring access protocol includes a simple and reliable global fairness mechanism which prevents a station from any starvation. By means of simulations, the fairness of the Metaring gigabit/s network is analyzed. Both symmetrical and asymmetrical traffic, as well as underload and overload traffic conditions, are investigated. It is shown that dimensioning the global fairness mechanism parameters strongly influences equity between nodes and the medium utilization ratio.

A bandwidth allocation scheme for the distributed-queue dual-bus (DQDB) network, the adaptive threshold algorithm (ATA), is proposed. The ATA mechanism is able to prevent with short response time any saturation or any underload of the network. The medium capacity is fairly shared between active nodes by keeping permanently the network global load under a given threshold. By means of simulation results, the performances of ATA and of bandwidth balancing are compared in terms of response time and of medium utilization ratio.>