|
List of Abstracts
|
|
|
Cost effective All-optical Network QoS-monitoring
Gregor Bochmann SITE - School of Information Technology and Engineering, University of Ottawa |
|
|
Agile All-Photonic Networks (AAPN) is a Canadian research network (funded by NSERC and 6 industrial partners) exploring the use of very fast photonic switching for building optical networks that allow the sharing (multiplexing) of a wavelength between different information flows. The aim is to bring photonic technology close to the end-user in the residential or office environment. The talk gives an overview of the proposed overlaid star network architecture and describes new results on (a) bandwidth allocation algorithms, (b) the routing and protection of MPLS flows over an AAPN using the concept of OSPF areas, and (c) our evolving plans for building demonstration prototypes. |
|
|
Cost effective All-optical Network QoS-monitoring
Mark Coates, Yvan Pointurier and Michael Rabbat Department of Electrical and Computer Engineering, McGill University
|
|
|
In all-optical networks, per-link monitoring is very difficult as signals remain in the optical domain from end-to-end, and expensive due to the amount of hardware required. Is it possible to monitor the quality-of-service throughout the whole network, with reasonable accuracy, by just monitoring a small number or lightpaths (or links, if the technology is available)? Where should the optical monitors be placed? This talk will describe a procedure for estimating a full set of network path metrics from a limited number of measurements. The approach exploits the strong spatial and temporal correlation observed in path-level metric data, which arises due to shared links and stationary components of the observed phenomena. We derive diffusion wavelets from the routing matrix to generate a basis in which the signals are compressible (coefficients exhibit approximately power law decay). This allows us to exploit powerful non-linear estimation algorithms that strive for sparse solutions. |
|
|
Optical networks optimization, an enterprise perspective
Denis Fluet Director of VoIP and Video Business Solutions for Optical Networks, Nortel. |
|
|
Large enterprises have been deploying optical networks for many years. Keep being pressed for more bandwidth by their data replication and business continuity requirements, couple to the emerging requirement for further apart data centers, enterprise are being more and more careful about their bandwidth efficiency and cost-effectiveness of their optical network. This paper will cover the enterprise network attributes, and the innovative complementary ways to address the enterprise optical network optimization challenge. |
|
|
Fill Me Up! With or Without Quality of Service? Testing and Monitoring Advancements: Maximizing Quality of Service of Triple-Play Delivery
Nicholas Gagnon Business Development Manager, Access Networks EXFO Electro-Optical Engineering Inc., Montréal
|
|
|
Commercial and residential customers of multiple-service network operators clearly expect optimum quality for their voice, data and video services. Of course, providing triple-play services requires a lot of bandwidth. Although low-bandwidth applications such as voice services (VoIP and POTS) and multiplayer gaming applications are still extremely useful and popular, the industry has shifted. With the advent and ever-increasing adoption of image-rich, high-bandwidth, graphic and video applications, it is now data and video traffic that is occupying the lion’s share of the bandwidth on any network.
However, as bandwidth demand increases, so does bandwidth offer. The
availability of next-generation transmission equipment and networks is
steadily rising as network operators worldwide adapt their services to
meet demand. Fiber-to-the-home (FTTH) ─ be it in the form of
point-to-point (P2P) fiber services or passive optical networking (PON)
─ is definitely a technology of choice when it comes to satisfying this
unquenchable “thirst” for bandwidth. In order to succeed in this increasingly competitive environment, network providers need to do more than just build an architecture that will offer a “fat pipe – ready to be filled”. They also need to provide subscribers with the service assurance that will keep users satisfied. This presentation will provide an overview of the principal technologies that allow network operators to deliver multiple services over Internet protocol (IPTV/video, voice/VoIP, high-speed Internet, multiplayer gaming). Important parameters such as latency, packet drops, and program clock reference synchronization must be considered and monitored all over the network ─ from core to access ─ to ensure that delivery meets the quality of service (QoS) and quality of experience (QoE) levels expected by customers. Namely, real-life network impairments affecting the QoS and QoE of these services will be addressed. The proper way to test and monitor these parameters in the core (transport) and access (distribution) segments of the network will also be discussed, and a PON remote fiber testing and monitoring technique will be shared. Finally, the presentation will also examine the particularities of both the technological and the business dimensions. The business case facet will present best practices that allow network operators to reduce operation expenses (OPEX) and truck rolls, without affecting the QoS and QoE of their multiple IP services. |
|
|
A Column Generation Approach for Design of Networks using Path-Protecting p-Cycles
Brigitte Jaumard, CIISE, Concordia University Caroline Rocha, DIRO, Université de Montréal
Dimitri Baloukov, TRLabs and ECE Department,
University of Alberta
|
|
|
We investigate a first column generation (CG) formulation for the design of failure independent path-protecting (FIPP) p-cycle survivable transport network. Previous work has proposed different formulations and heuristics for FIPP $p$-cycles which extend span-protecting p-cycles by adding the property of providing end-to-end failure independent path switching against either span or node failures. We develop a CG model that additionally allows the exploration of FIPP p-cycles without imposing mutual disjointness among working routes protected by the same cycle. The proposed CG model decomposes the FIPP p-cycle design problem into the master problem which takes care of the demand constraints, and the pricing problem which includes the constraints associated with the properties and the characteristics of a FIPP p-cycle. The key feature of a CG model lies in a generation of cycles motivated by the value of the reduced cost of the pricing problem, the key global indicator that is the driving element of the simplex algorithm. Results show a clear advantage of the CG model over the previous models, in particular in exploiting cycles that are not restricted to those satisfying a mutual disjointness condition on the working paths. Although it is not always possible to guarantee that the solutions obtained with the CG model are optimal, it can be shown that they are very close to the optimal ones.
|
|
|
SAIMD: A Novel False Congestion Detection Scheme in TCP over OBS Networks
Pin-Han Ho Department of Electrical and Computer Engineering, University of Waterloo
|
|
|
Due to the bufferless nature, TCP over Optical Burst Switched (OBS) networks suffer from false congestion detection, where a packet loss event due to random burst contention in a lightly-loaded OBS network may trigger TCP congestion control mechanisms, thereby degrading the performance of TCP. This phenomenon is particularly serious for the TCP flows with very high bandwidth, where returning to slow start due to false congestion detection would critically impair the TCP services and network performance. This paper introduces a novel congestion control scheme for high-bandwidth TCP flows over OBS networks, called Statistical Additive Increase Multiplicative Decrease (SAIMD). SAIMD maintains and analyzes a number of previous round trip time (RTTs) at the TCP senders in order to identify the confidence with which a packet loss event is due to network congestion. The confidence is derived by positioning short-term RTT in the spectrum of long-term historical RTTs. The derived confidence corresponding to the packet loss is then taken in the developed policy for TCP congestion window adjustment. We will show through extensive simulation that the proposed scheme can effectively solve the false congestion detection problem and significantly outperform the conventional TCP counterparts without losing fairness. The advantage gained in our scheme is at the expense of introducing more overhead in the SAIMD TCP senders. Based on the proposed congestion control algorithm, a throughput model is formulated, and is further verified by simulation results. |
|
|
Dynamic Multicast Traffic Grooming
Ahmad Khalil Electrical Engineering Department, The City University of New York
|
|
|
New and emerging applications in high-performance networks are of the multicast traffic type, where sessions include a single source transmitting to a number of destinations. Similar to the case of unicast traffic, multicast applications require only a fraction of the wavelength capacity. This makes the multiplexing of several application streams on the same wavelength a logical, and a cost effective approach. This multiplexing, however, must be done in a way that optimizes the use of resources, and minimizes the cost of the network. The allocation of resources should therefore take into account the session patterns, and their traffic requirements. However, with the use of IP over WDM, MPLS over WDM, or with NGS, traffic sessions now tend to exhibit a dynamic nature. Sessions arrive according to a certain arrival process, and they are characterized by holding times, which are taken from a certain distribution. Therefore, since it is practically impossible to design optical networks such that they accommodate all such dynamic sessions, most efforts have concentrated on devising call acceptance, and provisioning strategies that will try to reduce the session blocking probabilities. This talk will address the problem of dynamically provisioning both low-speed unicast and multicast connection requests in mesh-based WDM optical networks. Several routing/provisioning schemes to dynamically provision both unicast and multicast connection requests will be presented. A constraint-based grooming strategy is devised to utilize the overall network resources as efficiently as possible. Based on this strategy, several different sequential multicast grooming heuristics are first presented. Then, a hybrid grooming approach is devised and combined with sequential approaches to achieve a grooming scheme that is biased toward serving multicast traffic demands in comparison with all other sequential grooming approaches. |
|
|
Evaluation of Quality of Service Mechanisms in Optical Burst Switched Networks
Miroslaw Klinkowski National Institute of Telecommunications, Department of Transmission and Optical Technologies, Poland & Universitat Politècnica de Catalunya, Departament d'Arquitectura de Computadors, Spain Marian Marciniak Department of Transmission and Optical Technologies, National Institute of Telecommunications, Poland Davide Careglio Universitat Politècnica de Catalunya, Departament d'Arquitectura de Computadors, Barcelona, Catalunya, Spain Josep Solé-Pareta Universitat Politècnica de Catalunya, Departament d'Arquitectura de Computadors, Barcelona, Catalunya, Spain
|
|
|
In this paper, we address the problem of Quality of Service (QoS) provisioning in Optical Burst Switching (OBS). When examining the literature on OBS we can find several proposals of mechanisms dealing with QoS. As these mechanisms are evaluated in specific node/network scenarios each one, the results are usually not comparable. In this work we intend to confront the performance of the most frequently referenced QoS mechanisms in the same evaluation scenario consisting of a single isolated node. We show that the best performance of both high priority (HP) and low priority (LP) class is achieved with Burst Preemptive (BP) mechanism. The burst preemption mechanism suffers the problem of resource overbooking in case of successful preemption. Such problem can be avoided when applying a preemption window mechanism with additional preemption offset introduced. We evaluate the preemption window mechanism and find the offset which guarantees the performance of classical pre-emption scheme. |
|
|
A Bandwidth Estimation Method for Adaptive Capacity Allocation in Optical Networks
Jorg Liebeherr Nortel Chair in Network Architecture and Services Department of Electrical and Computer Engineering, University of Toronto Markus Fidler
Technische Universität Darmstadt, Germany Department of Electrical and Computer Engineering, University of Toronto |
|
|
Much research has been dedicated to methods that estimate the available bandwidth in a network from traffic measurements, yet little progress has been made on achieving a foundational understanding of the bandwidth estimation problem. In this talk, we develop a min-plus system theoretic formulation of bandwidth estimation. We show that the problem as well as previously proposed solutions can be concisely described and derived using min-plus system theory, thus establishing the existence of a strong link between network calculus and network probing methods. We relate difficulties in network probing to potential non-linearities of the underlying systems, and provide a justification for the distinctive treatment of FIFO scheduling in network probing. |
|
Evolutionary Upgrades of Optical Access and Metro Networks
Martin Maier INRS, Montreal |
|
|
In this talk, we elaborate on the question "WDM EPON - what's next?" Arguing that the short-term goals of cost reduction, design of colorless ONUs, and capacity increase by means of WDM will be addressed successfully in the near term, we together take a look at the impact of emerging applications on future optical access and metro network architectures and services. We introduce novel performance-enhancing techniques and lift the veil of all-optically integrated Ethernet-based access and metro networks, paying attention to their multifailure resilience and evolutionary long-lifetime capacity upgrades for unpredictable traffic due to events that are hard to predict, e.g., breaking news or DoS attacks. |
|
|
Planning GMPLS Transport Networks
H.T. Mouftah School of Information Technology and Engineering, University of Ottawa Nabil Naas School of Information Technology and Engineering, University of Ottawa
|
|
|
With the explosive traffic growth of WDM-based transport networks, the development of GMPLS-based transport networks becomes essential to cope with the network scalability problems. This paper defines a novel problem of planning GMPLS-based transport networks by (1) considering the whole traffic hierarchy defined in GMPLS; (2) allowing optical signal conversion at all granularity levels; (3) imposing optical reach constraint on the length of all-optical paths. The objective of the problem is to minimize the weighted port count in the transport network. Due to the computational complexity of the problem, only very-small-sized problems can be solved exactly through Mixed Integer Linear Programming (MILP) optimization. In this work, we propose novel heuristics that are capable of solving large-sized problems in a reasonable amount of time. |
|
|
Designing Robust Telecom Networks Biswanath Mukherjee University of California, Davis |
|
|
This talk will focus on the design of robust network architectures, which can recover quickly and efficiently from overloads, failures, and attacks. Some specific applications include provisioning of restorable bandwidth-guaranteed tunnels in MPLS networks, IP flows, and lightpaths in optical networks. Topics to be covered include the benefits of multi-path routing, re-provisioning, etc. |
|
|
Resource Reservations in the Physical Layer
Vassilis Prevelakis Computer Science Department, Drexel University Admela Jukan INRS, Montreal |
|
|
Recently, a number of new research initiatives, most notably UCLPv2 and GENI, have promoted the dynamic partition of physical network infrastructure as the means to operate the network, and to implement new protocols and services. This has led to a number of open issues such as resource discovery, implementation of resource partitioning, and the aggregation of resources to create arbitrary network topologies. To us, the key issue is the design of a mechanism to reserve and control network resources, given a choice of providers of physical resources (infrastructure providers). We present an architecture that allows physical resources to be reserved and traded, while granting users controlled access to the acquired resources via a policy enforcement mechanism. In addition, it allows resource provider domains to be linked via configurable, provider-neutral resource exchange points that are the physical resource equivalents of the pooling point, or Internet Exchange Point (IXP). We demonstrate how our resource reservation system will operate by presenting a use case in which a network topology is constructed using resources from multiple providers. The use case also shows how a dynamic reconfiguration can be effected by the customer though the use of simple access control policies, without involving the provider. |
|
|
Designing Robust Telecom Networks Biswanath Mukherjee Department of Computer Science University of California, Davis |
|
|
This talk will focus on the design of robust network architectures, which can recover quickly and efficiently from overloads, failures, and attacks. Some specific applications include provisioning of restorable bandwidth-guaranteed tunnels in MPLS networks, IP flows, and lightpaths in optical networks. Topics to be covered include the benefits of multi-path routing, re-provisioning, etc. |
|
|
Past and Present Optimization Problems in Optical Networks: (NP, IP, LP and many other "P"s)
Chunming Qiao Computer Science and Engineering Department University at Buffalo |
|
|
In this talk, I will first review some of optimization problems in designing WDM ring and mesh networks we have worked on in the past, and then discuss some of the present optimization problems in designing next generation Optical Internet including Polymorphous OBS and Optical Grids. |
|
|
Delay Analysis of Ethernet Passive Optical Networks Abdallah Shami and Xiaofeng Bai Department of Electrical and Computer Engineering, The University of Western Ontario
|
|
|
This talk will present a new analytical framework to evaluate the performance of Ethernet Passive Optical Networks (EPONs). This new analytical framework explores the performance characteristic of EPONs with quasi-leaved polling operation and gated resource allocation policy. Using classical queuing theory, we will present closed form expression for the average packet queuing delay and average queue length of the EPON model. Simulation experiments show that the derived analytical expression can precisely evaluate the network performance for Memoryless traffic inputs, as well as to closely estimate the performance of lightly loaded network for bursty input traffic profile. |
|
|
Low Loss Reconfigurable Optical Add / Drop Modules
Alex Vukovic CRC, Broadband and Optical Networks |
|
|
The next generation of
reconfigurable optical add drop multiplexers (ROADMs) looks to be a key
enabler for dynamic, easily provisioned, remotely activated networks.
ROADMs will allow a fully automated optical layer with single-wavelength
granularity by performing flexible add/drops with colorless ports,
dynamic management of power, power equalization, optical monitoring of
all channels, all of this within a single unit. |
|
|
Photonic techniques of millimeter wave over fiber systems for sensing and communication networks
J.X. Zhang Department of Electrical and Computer Engineering, Concordia University J. Yao University of Ottawa, K. Wu and R. Kashyap Département de Génie Electrique, Ecole Polytechnique de Montreal
|
|
|
The presentation will first give an outline of current state of the art of techniques in photonic millimeter-wave generation, reception, and distribution over fiber. Then, our contributions from each research group are presented, including novel techniques of subcarrier modulations, cost-effective solutions of photonic millimeter wave generation, reception and distribution combined with DWDM technique, broadband optoelectronic mixers and substrate integrated circuits as well as novel lasers. Also, the impact of nonlinear distortion on millimeter-wave over fiber systems is discussed. |
|