Content uploaded by Piet Demeester
Author content
All content in this area was uploaded by Piet Demeester on Sep 13, 2014
Content may be subject to copyright.
Hybrid optical network architectures: Bringing packets and circuits together
Abstract
In recent years hybrid optical network architectures, which employ two or more network technologies simultaneously, were proposed. They aim at improving the overall network design by combining the advantages of different technologies while avoiding their disadvantages. In order to structure this developing research field, we classify such hybrid architectures based on the degree of interaction and integration of the network technologies. Also, we discuss the three classes and their main representatives regarding key characteristics, performance benefits, and realization complexity. Finally, we highlight two hybrid architectures and show their key benefits compared to the respective non-hybrid architectures through a dimensioning case study
... The problem of carrying mixed traffic in backbone networks that utilize hybrid switching has been widely investigated. Three types of network architectures that can support hybrid switching are discussed in [11]. In one of our previous works, we proposed the BLOC framework for backbone networks to characterize the dynamic resource allocation problem in hybrid switching systems [6]. ...
... Existing studies on hybrid switching in backbone networks divide the architecture into three categories based on the degree of integration between network technologies [11]. The first architecture is the client-server type, e.g., the IP-overwavelength division multiplexing (WDM) network, where the underlying service layer provides a virtual topology for the upper client layer. ...
... Equation (10) is based on the definition of packet loss probability in the RED queue [38]. Equation (11) is based on the definition of bandwidth utilization. Equation (12) defines the throughput of packet streams [8]. ...
Access links that connect client networks to public networks have to carry multiple-type and time-varying traffic. Those hybrid traffic flows compete for limited bandwidth, yet at the same time have drastically different performance requirements. As planning the link capacity for peak traffic demand is not economically viable, it is important that bandwidth on the access links is shared between different traffic classes in a way that maximal network utility can be achieved. In this paper, we study the frequency slot allocation problem on IP-over-elastic optical network (EON) access links that carry three types of traffic, namely, packet streams, latency critical circuit connections (e.g., video conferencing), and delay tolerant circuit connections (e.g., bulk data transfers). We define four network operation states; in each of which an access link serves traffic with different levels of fulfillment. We then formulate the allocation problem into a weak-constrained optimization problem and propose a genetic algorithm to solve it in real time. Numerical results show that the relative error of the genetic algorithm is within 3% and the access link keeps maintaining the optimal achievable network operation state. We also show that, by increasing the storage size, the access link can adapt to the increasing traffic load within a certain range without upgrading the expensive access link bandwidth. Our study provides useful insights in managing and operating IP-over-EON access links, and the concept of multiple network operation states can be generalized to networks that serve more than one type of traffic.
... In recent years, the dynamic growth of fiber deployment has brought more transmission capacity in to hand, which is beyond the processing capacity of electronic switches [1]. As a result, fiber switching technologies [2], have been introduced to overcome the electronics processing and switching bottlenecks. ...
... Even though optics offers bulk capacity, the utilization is inefficient due to slow adaptation of wavelength switching for burst traffic. All-optical switching is believed to enhance the efficiency, but it is not practically deployed [1]. Hence, a new adaptive design approach, the hybrid optical architecture [1], was proposed: it combines the best features of both optical circuit and packet switching while diminishing their demerits to provide better performance and cost reduction [1]. ...
... All-optical switching is believed to enhance the efficiency, but it is not practically deployed [1]. Hence, a new adaptive design approach, the hybrid optical architecture [1], was proposed: it combines the best features of both optical circuit and packet switching while diminishing their demerits to provide better performance and cost reduction [1]. Integrated hybrid optical network (IHON) is one type of hybrid network which completely integrates both circuit and packet technologies, and uses the same wavelength to transmit both traffic. ...
Today, deployment of optical fiber has offered large transmission capacity which cannot be efficiently utilized by the electronic switches. Rather, Integrated Hybrid Optical Network (IHON) is a promising approach which combines both packet and circuit switching techniques. As a result, it achieves efficient utilization of the bulk capacity and guarantees absolute Quality of Service (QoS) by optimizing the advantages of the two switching schemes while diminishing their disadvantages. Transpacket has developed a Fusion node implementing IHON principles in Ethernet for the data plane. Hence, this paper investigates and evaluates IHON network for 5G access networks. The simulated results and numerical analysis confirm that the Packet Delay Variation (PDV), Delay and Packet Loss Ratio (PLR) of Guaranteed Service Transport. (GST) traffic in IHON network met the requirements of 5G mobile fronthaul using CPRI. The number of nodes in the network limits the maximum separation distance between Base Band Unit. (BBU) and Remote Radio Head (RRH), link length; for increasing the number of nodes, the link length decreases. In addition to this, we verified how the leftover capacity of fusion node can be used to carry the low priority packets and how the GST traffic can have deterministic characteristics on a single wavelength by delaying it with Fixed Delay Line (FDL). For example, for L<sup>SM</sup><sub>1GE</sub>=0.3 the added Statistical Multiplexing (SM) traffic increases the 10GE wavelength utilization up to 89% without any losses and with SM PLR=1 E <sup>−03</sup> up to 92% utilization.
... With the case of data centers, the two basic premises for the application of traffic overflow are: equalizing loads [8,9] and energy consumption limitations [10,11]. In optical networks, the traffic overflow mechanism is one of the key mechanisms, considered as early as the dimensioning stage for these networks (e.g. in Optical Burst Switching Networks) [12,13]. Traffic overflow is also widely used in access networks, primarily in mobile wireless networks, [14]. ...
... The distribution (11) allows us to determine the blocking probability (12), and in consequence the average value of the intensity of traffic of class i that overflows from j-th PR: ...
... Another assumption is that the thresholds in PR are arranged in such a way [89] that the blocking phenomenon occurs exclusively in the oldest LA: T i,k < n ≤ V j . Then, (12) can be written as follows: ...
This article presents a universal and versatile model of multiservice overflow systems based on Hayward's concept. The model can be used to analyze modern telecommunications and computer networks, mobile networks in particular. The advantage of the proposed approach lies in its ability to analyze overflow systems with elastic and adaptive traffic, systems with distributed resources and systems with non-full-availability in primary and secondary resources.
... M. Głąbowski is with Faculty of Electronics and Telecommunications, Poznan University of Technology, Poland (e-mail: mariusz.glabowski@put.poznan.pl). many applications, which among others include: optical networks (e.g. in Optical Burst Switching Networks) [13], [14], mobile networks 2G, 3G, 4G [15], [16], [17], [18], [19], [20], [1], [21], packet networks [22], [23], [24] as well as commutation fields [25], [26], [27]. In computer networks, this mechanism is used to improve the performance of cloud computing [28], file organization [29], load balancing [30], [31] and even to reduce energy consumption [32], [33]. ...
... Hybrid data center network (HDCN) offers a possible solution to the current state of high demand and rapid growth in network traffic, along with the exponential increase in power consumption and cost of DCNs by integrating multiple switching schemes to make use of their numerous advantages. A typical hybrid DCN constitutes a multiplicity of switching schemes such as the optical packet switching (OPS), optical burst switching (OBS), optical circuit switching (OCS) and electronic packet switching (EPS) [3], [4]. This integration of schemes or switching systems allows for a new and enabling technology capable of transforming current data center networks to an agile and flexible kind. ...
... Maximum E2E latency (L total ): L total = N × (D node + Delay_PDV node ) + D T × Link length (1) where, N = Number of nodes. D node = Latency in a single node, 1.2µsec (obtained from the simulation). ...
Today, deployment of optical fiber has offered large transmission capacity which cannot be efficiently utilized by the electronic switches. Rather, Integrated Hybrid Optical Network (IHON) is a promising approach which combines both packet and circuit switching techniques. As a result, it achieves efficient utilization of the bulk capacity and guarantees absolute Quality of Service (QoS) by optimizing the advantages of the two switching schemes while diminishing their disadvantages. Transpacket has developed a Fusion node implementing IHON principles in Ethernet for the data plane. Hence, this paper investigates and evaluates IHON network for 5G access networks. The simulated results and numerical analysis confirm that the Packet Delay Variation (PDV), Delay and Packet Loss Ratio (PLR) of Guaranteed Service Transport. (GST) traffic in IHON network met the requirements of 5G mobile fronthaul using CPRI. The number of nodes in the network limits the maximum separation distance between Base Band Unit. (BBU) and Remote Radio Head (RRH), link length; for increasing the number of nodes, the link length decreases. In addition to this, we verified how the leftover capacity of fusion node can be used to carry the low priority packets and how the GST traffic can have deterministic characteristics on a single wavelength by delaying it with Fixed Delay Line (FDL). For example, for LSM 1GE=0.3 the added Statistical Multiplexing (SM) traffic increases the 10GE wavelength utilization up to 89% without any losses and with SM PLR=0.001 up to 92% utilization.
... using collaboration algorithms such as genetic algorithm [2]. In addition, it is hard to upgrade and repair because of the all-in-one BS architecture. ...
p>Increasing mobile data traffic due to the rise of both smartphones and tablets has led to high-capacity demand of mobile data network. To meet the ever-growing capacity demand and reduce the cost of mobile network components, Cloud Radio Access Network (C-RAN) has emerged as a promising solution. In such network, the mobile operator’s Remote Radio Head (RRH) and Base Band Unit (BBU) are often separated and the connection between them has very tight timing and latency requirements imposed by Common Public Radio Interface (CPRI) and 3rd Generation Partnership Project (3GPP). This fronthaul connection is not yet provided by packet based network. To employ packet-based network for C-RAN fronthaul, the carried fronthaul traffic are needed to achieve the requirements of fronthaul streams. For this reason, the aim of this paper is focused on investigating and evaluating the feasibility of Ethernet networks for mobile fronthaul. The fronthaul requirements used to evaluate and investigate this network are maximum End to End (E2E) latency, Packet Loss Ratio (PLR) and Packet Delay Variation (PDV). The simulated results and numerical analysis confirm that the PDV and PLR of High Priority (HP) traffic in Ethernet network meet the requirements of mobile fronthaul using CPRI. However, the PDV of HP traffic meets the fronthaul network when the number of nodes in the Ethernet network is at most four. For Ethernet network, the number of nodes in the network limits the maximum separation distance between BBU and RRH (link length); for increasing the number of nodes, the link length decreases.
Consequently, Radio over Ethernet (RoE) traffic should receive the priority and Quality of Service (QoS) HP can provide. On the other hand, Low Priority (LP) classes are not sensitive to QoS metrics and should be used for transporting time insensitive applications and services.</p
In this article, we present an analysis of the accuracy level of methods for modeling the multi-service overflow systems that service Erlang, Engset, and Pascal traffic. In systems with traffic overflow, new calls that cannot be serviced by the primary resources are overflown (directed) to other available resources that can service a given call, that is, to the secondary resources (alternative resources). In the article, we focus on studying the influence of methods for determining the parameters of traffic that overflows to the secondary resources on the accuracy of determining the traffic characteristics of overflow systems. Our analysis revealed that the main source of the inaccuracy of the existing methods is their approach to determining both the average value and the variance of multi-service Pascal traffic streams offered to the secondary resources. Therefore, we proposed a new method for determining the parameters of Pascal overflow traffic. The method is based on the decomposition of multi-service primary resources into single-service resources and the subsequent conversion of Engset and Pascal streams into equivalents of Erlang traffic. The results of the analytical calculations obtained on the basis of the new method are then compared with the results of simulation experiments for a number of selected structures of overflow systems that service Erlang, Engset, and Pascal traffic. The results of the study indicate that the proposed theoretical model has a significantly higher accuracy than the models proposed in the literature. The method can be used in the analysis, dimensioning, and optimization of multi-service telecommunication systems composed of separated resources, for example, mobile cellular systems.
Bandwidth and computation-intensive Big Data applications in disciplines like social media, bio- and nano-informatics, Internet-of-Things (IoT), and real-time analytics, are pushing existing access and core (backbone) networks as well as Data Center Networks (DCNs) to their limits. Next generation DCNs must support continuously increasing network traffic while satisfying minimum performance requirements of latency, reliability, flexibility and scalability. Therefore, a larger number of cables (i.e., copper-cables and fiber optics) may be required in conventional wired DCNs. In addition to limiting the possible topologies, large number of cables may result into design and development problems related to wire ducting and maintenance, heat dissipation, and power consumption. To address the cabling complexity in wired DCNs, we propose OWCells, a class of optical wireless cellular data center network architectures in which fixed line of sight (LOS) optical wireless communication (OWC) links are used to connect the racks arranged in regular polygonal topologies. We present the OWCell DCN architecture, develop its theoretical underpinnings, and investigate routing protocols and OWC transceiver design. To realize a fully wireless DCN, servers in racks must also be connected using OWC links. There is, however, a difficulty of connecting multiple adjacent network components, such as servers in a rack, using point-to-point LOS links. To overcome this problem, we propose and validate the feasibility of an FSO-Bus to connect multiple adjacent network components using NLOS point-to-point OWC links. Finally, to complete the design of the OWC transceiver, we develop a new class of strictly and rearrangeably non-blocking multicast optical switches in which multicast is performed efficiently at the physical optical (lower) layer rather than upper layers (e.g., application layer). Advisors: Jitender S. Deogun and Dennis R. Alexander
Bandwidth and computation-intensive Big Data applications in disciplines like social media, bio- and nano-informatics, Internet-of-Things (IoT), and real-time analytics, are pushing existing access and core (backbone) networks as well as Data Center Networks (DCNs) to their limits. Next generation DCNs must support continuously increasing network traffic while satisfying minimum performance requirements of latency, reliability, flexibility and scalability. Therefore, a larger number of cables (i.e., copper-cables and fiber optics) may be required in conventional wired DCNs. In addition to limiting the possible topologies, large number of cables may result into design and development problems related to wire ducting and maintenance, heat dissipation, and power consumption.
To address the cabling complexity in wired DCNs, we propose OWCells, a class of optical wireless cellular data center network architectures in which fixed line of sight (LOS) optical wireless communication (OWC) links are used to connect the racks arranged in regular polygonal topologies. We present the OWCell DCN architecture, develop its theoretical underpinnings, and investigate routing protocols and OWC transceiver design. To realize a fully wireless DCN, servers in racks must also be connected using OWC links. There is, however, a difficulty of connecting multiple adjacent network components, such as servers in a rack, using point-to-point LOS links. To overcome this problem, we propose and validate the feasibility of an FSO-Bus to connect multiple adjacent network components using NLOS point-to-point OWC links.
To complete the design of the OWC transceiver, we develop a new class of strictly and rearrangeably non-blocking multicast optical switches in which multicast is performed efficiently at the physical optical (lower) layer rather than upper layers (e.g., application layer).
This paper proposes a novel polymorphic framework for optical networking and a seamless evolution path from optical circuit-switched towards optical packet-switched networks. We show that by simultaneously supporting several optical switching paradigms in a single physical topology, efficient and flexible optical networks can be built. The supported paradigms are associated with different Classes of Service (CoS) in order to provide service differentiation at the optical layer. Two polymorphic architectures are presented, one based on optical circuit switching paradigms with different grades of dynamism, and a second one based on optical labeled burst-switched networks with the added capability of dynamic lightpath provisioning. These architectures provide a seamless evolution path towards an efficient IP-over-WDM approach with service differentiation. Moreover, the proposed polymorphic architectures are fully compatible with the GMPLS unified control plane. We present in a detailed form the proposed polymorphic framework, including the selection of switching paradigms, its support for CoS, the network and control architecture, and a possible seamless evolution towards optical packet-switched networks. Possible implementation examples of optical network nodes that support the proposed polymorphic architectures are also presented.
Optical burst switches (OBSes) have been proposed to improve the utilization of a network of optical cross connect (OXCs). Current studies on OBS assume a network consisting of OBSes alone. While this is a reasonable assumption for evaluating a new technology, the question of how a network of OXCs can be evolved to a network of OBSes has not been studied. In this paper, we propose a hybrid architecture consisting of OBSes at the network edge and OXCs in the network core. This architecture allows carriers to gradually migrate from an OXC-based network to an OBS-based network with an improved network utilization. In addition, we use queueing analysis to study the performance of this new architecture.
Optical circuit switching (OCS) is a sophisticated technology widely deployed in current optical networks, and has many advantages in the transport of stable and long-duration traffic flows. However, it is not suitable for bursty data traffic. On the other hand, an alternative technology, optical burst switching (OBS), well addresses bursty IP traffic transport, but is not suitable for stable and large flows. To transport both types of traffic effectively, a hybrid optical switching approach is proposed which combines OCS and OBS to exploit the merits of both technologies. The performance has been evaluated in terms of throughput and blocking probability.
We propose a new optical hybrid switching system that takes advantage of both optical burst switching (OBS) and optical circuit switching (OCS) technologies. This system classifies incoming IP traffic flows into short-lived and long-lived flows. We model the system as a single server queue in a Markovian environment. The burst generation process is assumed to follow a two-state Markov modulated Poisson process (MMPP), and the service rate fluctuates based on the number of concurrent OCS sessions. Results for the mean delay and queue size are derived.
We propose a circuit-switched high-speed end-to-end transport architecture (CHEETAH) as a networking solution to provide high-speed end-to-end circuit connectivity to end hosts on a dynamic call-by-call basis. Not only is it envisioned as a complementary service to the basic connectionless service provided by today's Internet; it also relies on and leverages the presence of this service. Noting the dominance of Ethernet in LANs and SONET/SDH in WANs, CHEETAH circuits will consist of Ethernet segments at the ends and Ethernet-over-SONET segments in the wide area. In this article we explain the CHEETAH concept and describe a wide-area experimental network testbed we have deployed based on this concept. The network testbed currently extends between Raleigh, North Carolina, Atlanta, Georgia, and Oak Ridge, Tennessee, and uses off-the-shelf switches. We have created CHEETAH software to run on end hosts to enable automated use of this network by applications. Our first users of this network testbed and software will be the Terascale Supernova Initiative (TSI) project researchers, who plan to use this network for large file transfers and remote visualizations.
Hybrid hierarchical optical cross-connects enhance the performance/cost ratio of optical networks by providing transparent (optical) switching of sets of wavelengths (wavebands) in addition to opaque (electrical) switching of individual wavelengths. As network bandwidth gets cheaper, and the performance bottleneck moves to switching nodes, these systems provide an attractive scalable solution for next-generation optical networks. We describe key technological components (including flexible nonuniform wavebands) of hybrid hierarchical optical cross-connects and discuss their performance/cost implications.
The emerging wavelength switched networks reduce the strain on packet forwarding. Unfortunately, that solution is not really efficient on a bandwidth level, and is not ideally suited for bursty traffic. Packet switched solutions, whether electronic or optical, can use statistical multiplexing to cope with bursty traffic and yield better bandwidth efficiency. We present a novel network concept that can combine these two worlds, withholding their advantages. We introduce this Overspill Routing In Optical Networks (ORION), and discuss several aspects of it: the overall architecture and network concept, node design and implementation, and evaluation at network level as well as node level.
This paper presents novel network architecture, called optical burst transport network (OBTN), which can efficiently transport burst data over a virtual topology and reduce the port count of optical burst nodes. In OBTN nodes, bursts from attached optical feeder networks, e.g., metro networks, are multiplexed and transported to their egress OBTN node in one of two ways: preferably on direct end-to-end lightpaths set up between OBTN nodes or alternately on a relatively small number of hop-by-hop overflow burst wavelengths assigned to specific links which are shared among traffic flows. In contrast to hybrid network architectures which partition network resources completely and classify traffic at the edge for burst or circuit transport to isolate traffic classes, our OBTN approach integrates both resource types, i.e., circuit and burst wavelengths. It minimizes transit traffic in intermediate nodes and reduces node sizes while providing an overall very low burst-loss probability due to optimized contention resolution. First, we outline an emerging optical network scenario. Then, we introduce the OBTN network and node architectures, and show the impact of key design parameters. Finally, we compare OBTN with optical burst switching (OBS) and an approach using optical edge traffic aggregation switches, called burst over circuit switching (BoCS), regarding performance, required network resources, and node complexity.
One promising switching technology for wavelength-division multiplexing optical networks is optical burst switching (OBS). However, there are major deficiencies of OBS. (1) The delay offset between a control message and its corresponding data burst is based on the diameter of a network. This affects network efficiency, quality-of-service, and network scalability.( 2) OBS adopts one-way resource reservation scheme, which causes frequent burst collision and, thus, burst loss. We address the above two important issues in OBS. In particular, we study how to improve the performance of delay and loss in OBS. To reduce the end-to-end delay, we propose a hybrid switching scheme. The hybrid switching is a combination of lightpath switching and OBS switching. A virtual topology design algorithm based on simulated annealing to minimize the longest shortest path through the virtual topology is presented. To minimize burst collision and loss, we propose a new routing algorithm, namely, p-routing, for OBS network. The p-routing is based on the wavelength available probability. A path that has higher available probability is less likely to drop bursts due to collision. The probability-based p-routing can reduce the volatility, randomness, and uncertainty of one-way resource reservation. Our studies show that hybrid switching and p-routing are complementary and both can dramatically improve the performance of OBS networks.