Time-driven or Fractional Lambda Switching for "Green Internet"

Y. Ofek, "Ultra Scalable Optoelectronic Switching Fabric for Streaming Media over IP," ISSPIT 2005 (IEEE International Symposium on Signal Processing and Information Technology)

ABSTRACT

As data traffic on the Internet continues to grow exponentially, there is a real need to solve the switch bottleneck by developing ultra-scalable switching fabrics. Although in recent years there have been a lot of efforts to solve the switching fabric scalability problem, in the optical domain, the proposed solutions have been (very) expensive, (very) complex and (much) larger than electronic switching fabric alternatives. Unfortunately, electrical interconnection of existing (off-the-shelf) electronic switching devices is very difficult due to wave reflections on transmission lines, impedance mismatching, crosstalk and noise. Consequently, electronic switching with electrical interconnections has major scalability limitations. Thus, the question is whether and how optical interconnects can be used to link off-the-shelf electronic switching devices in order to develop optoelectronic switching fabrics that can scale up to 10-100 terabit per second (Tb/s) capacity in a single chassis. In order to further increase the scalability of the proposed novel optoelectronic switching fabric a UTC-based (coordinated universal time) time driven switching or fractional lambda switching (FS) architecture is proposed. This novel low-complexity switching architecture capitalizes on the ubiquitous of UTC from GPS and Galileo. FS architecture is especially suitable to support high capacity streaming media applications over the Internet.

Ofek, Pattavina - IEEE Globecom 2003

ABSTRACT

Ofek, Pattavina - IEEE Globecom 2004

ABSTRACT

Ofek, Pattavina - Journal paper 2004

ABSTRACT

M. Baldi, Y. Ofek, "Fractional Lambda Switching - Principles of Operation and Performance Issues," SIMULATION: Transactions of The Society for Modeling and Simulation International, Vol. 80, No. 10, Oct. 2004, pp. 527-544.

ABSTRACT

This article introduces fractional lambda (�) switching (F�S) and studies its blocking issues. F�S uses a global common time reference (CTR) for implementing pipeline forwarding (PF) inside the network. A global CTR is conveniently realized with the UTC (coordinated universal time) standard. Resource reservation over an F�S network requires a schedule. As in other scheduling cases, a call may not be accepted, even though there is enough capacity, because the schedule is not available�the call is then considered blocked. This work studies the probability of call blocking as a function of link utilization. The results show that (especially if multiple-wavelength division multiplexing channels are deployed on optical links between fractional � switches) high-link utilization can be achieved with negligible call blocking, even when the switching fabric is a Banyan network.

M. Baldi, Y. Ofek, "Comparison of Ring and Tree Embedding for Real-time Group Multicast," IEEE/ACM Transactions on Networking, Vol. 11, No. 3, June 2003, pp. 451-464.

ABSTRACT

In general topology networks, routing from one node to another over a tree embedded in the network is intuitively a good strategy, since it typically results in a route length of O(log n) links, being n the number of nodes in the network. Routing from one node to another over a ring embedded in the network would result in route length of O(n) links. However, in group (many-to-many) multicast, the overall number of links traversed by each packet, i.e., the networks elements on which resources must be possibly reserved, is typically O(N) for both tree and ring embedding, where N is the size of the group. This paper focuses on the tree versus ring embedding for real-time group multicast in which all packets should reach all the nodes in the group with a bounded end-to-end delay. In this paper, real-time properties are guaranteed by the deployment of time-driven priority in network nodes.In order to have a better understanding of the nontrivial problem of ring versus tree embedding, we consider the following group multicast scenarios: 1) static--fixed subset of active nodes, 2) dynamic--fixed number of active nodes (i.e., the identity of active nodes is changing over time, but its size remains constant), and 3) adaptive--the number and identity of active nodes change over time.Tree and ring embedding are compared using the following metrics: 1) end-to-end delay bound, 2) overall bandwidth allocated to the multicast group, and 3) signaling overhead for sharing the resources allocated to the group. The results are interesting and counterintuitive, since, as shown, embedding a tree is not always the best strategy. In particular, dynamic and adaptive multicast on a tree require a protocol for updating state information during operation of the group. Such a protocol is not required on the ring where the circular topology, and implicit token passing mechanisms are sufficient. Moreover, the bandwidth allocation on the ring for the three multicast scenarios is O(N), while on a general tree it is O(N) for the static multicast scenario and O(N²) for the dynamic and adaptive multicast scenarios.

M. Baldi, Y. Ofek, "Fractional Lambda Switching," IEEE International Conference on Communications (ICC2002), Optical Networking Symposium, New York, NY, USA, Apr. 2002, pp. 2692-2696.

ABSTRACT

Fractional Lambda (�) Switching (F�S) adds the necessary efficiency to Wavelength Division Multiplexing (WDM) while preserving the simplicity of whole � switching. Due to its provisioning capability, from a fraction of STS-1 to a full optical channel capacity, F�S� will extend the reach of optical networking all the way to the network edges in the metro and enterprise. Finally, F�S� uniquely enables the implementation of dynamic all-optical switches since its operation does not require (1) optical processing (e.g., packet header processing) and (2) optical buffering.

M. Baldi, Y. Ofek, "Realizing Dynamic Optical Networking," Optical Networks Magazine, Special Issue "Dynamic Optical Networking: around the Corner or Light Years Away?", Vol. 4, No. 5, Sep./Oct. 2003, pp. 100-111.

ABSTRACT

Electronic asynchronous packet switching has been developed since the early 1960s and now, at the dawn of the 21st century, it is a well-understood technology. On the other hand, all-optical packet switching, a major candidate for the realization of dynamic optical networking, is a new technology, currently undergoing research and development. In this paper we show that the transition from electronic to optical is not as straightforward as one would like to think.
The main thesis of this paper is that the transition to dynamic all-optical switching will require changing the current asynchronous packet switching paradigm. This is shown by focusing the analysis on the simple problem of optical memory, while not considering the more complex problems associated with optical packet header processing and switching as if they had been resolved - which is indeed not the case.
In order to evaluate optical memory, various comparative measures are used. For example, much more glass (Silicon) is needed for optical memory versus electronic memory, in fact a factor of 1 million! If we assume shrinking of Silicon circuitry by a factor of two every 18 months, the transition to optical packet switching will translate into reversing Moore's law by 30 years.
It will be shown that a common time reference (CTR) is needed for realizing the optical random access memory (O-RAM) required for dynamic optical networking. Furthermore, it is shown that by using coordinated universal time (UTC a.k.a. GMT) CTR can be globally distributed, then it is possible to (globally) distribute the O-RAM and, consequently, to realize a new dynamic optical networking architecture called Fractional Lambda Switching (F�S), that additionally, does not require optical header processing.

M. Baldi, Y. Ofek, "UTC for Enabling New Applications and Services in GRID Networks," AICA 2005, Udine (Italy), Oct. 2005.

ABSTRACT

A central challenge in the GRID today is to develop applications and services, specifically, profitable services. such developed has to be supported by proper mapping from the logical/virtual application to the underlying combined/converged networking and computing physical infrastructure. Proper mapping should ensure: (1) well-defined resource allocation, such as: computing, storage, bandwidth - while guaranteeing (2) well-defined delay bounds. Proper use of UTC, ubiquitously and almost free of charge available from GPS and Galileo, enables proper (i.e., cost effective) allocation of computing and networking resources.

V.-T. Nguyen, M. Baldi, R. Lo Cigno, and Y. Ofek, "Wavelength Swapping using Tunable Lasers for Fractional � Switching," 14th IEEE Workshop on Local and Metropolitan Area Networks, Chania, Crete (Greece), Sep. 2005.

ABSTRACT

Fractional Lambda Switching (F�S) is a novel proposal for the management of all-optical networks with sub-wavelength provisioning capability. The unique characteristic of F�S is the utilization of the UTC (coordinated universal time) for alignment and switching. Several central research issues are still open in F�S and need to be formally defined and analyzed. Within the scope of this paper, we introduce three novel switch architecture designs that are based on the use of tunable lasers. As an important goal, we introduce the notion �scheduling feasibility� that measure the number of possible different scheduling between an input and output time-frames.

M. Baldi, Y. Ofek, "Dynamic Optical Switching: The Network is the Memory," 11th IEEE International Conference on Software, Telecommunications and Computer Networks (SoftCOM 2003), Split, Dubrovnik (Croatia), Ancona, Venezia (Italy), Oct. 2003, pp. 631-635.

ABSTRACT

This paper shows that the transition to dynamic all-optical packet switching, that fully exploits the benefits of fast dynamic optical components, will require changing the current asynchronous packet switching paradigm. This is shown by focusing on the relatively simple problem of optical memory, while assuming that the more complex problems associated with optical packet header processing and fast switching in the optical domain have been resolved - which is indeed not in the case. Various comparative measures are used in the optical memory evaluation. For example, much more glass (Silicon) is needed for optical memory than for electronic memory; in fact the ratio is on the order of 1 million. As shown, a common time reference (CTR) is needed for realizing the optical random access memory (O-RAM) required for optical packet switching. Furthermore, a CTR is used for both minimizing the O-RAM requirements and distributing the O-RAM in the network realizing a new optical network architecture called Fractional Lambda Switching (F�S).

M. Baldi, Y. Ofek, "Grooming and Degrooming with Coordinated Universal Time (UTC)," Workshop on Contemporary Communications, 11th IEEE International Conference on Software, Telecommunications and Computer Networks (SoftCOM 2003), Split, Dubrovnik (Croatia), Ancona, Venezia (Italy), Oct. 2003, pp. 63-67.

ABSTRACT

Coordinated Universal Time (UTC) enables the efficient and cost effective implementation of a grooming and degrooming method. The number of times data packet headers are being decoded throughout a global network deploying a UTC is limited to the necessary minimum of one. The simple implementation of grooming and degrooming devices results in high scalability and low cost.

M. Baldi, Y. Ofek, "Global Time for Interactive Applications over Global Packet Networks," 4th International Workshop on Quality of Future Internet Services (QoFIS 2003), Stockholm, Sweden, Oct. 2003, Springer-Verlag Lecture Notes on Computer Science (LNCS) 2811, pp. 52-62.

ABSTRACT

This work is present that global time (a.k.a. time-of-day or coordinated universal time - UTC) is essentital in maximizing user perceived quality of service, while eliminating both switching bottleneck - critical in very high capacity network core - and communications link bottlenecks - at the low speed access (e.g., wireless and DSL). Global time obtained, for example, from GPS (Global Positioning System) or Galileo, is used in the design of all streaming media applications such as toll quality telephony, videotelephony and videoconferencing. The proposed solution, that can be applied to the Internet without changes to any of the existing protocols, provides a guaranteed quality service to each application without requiring nodes to keep state information on microflows.