Scalable and energy-efficient packet switches based on multi-granular forwarding operations

We propose a coordinated non-sensing medium access control (MAC) protocol in dynamic spectrum access networks namely distributed coordinated dynamic spectrum reservation (DCDSR) protocol which is not based on the cognitive radio technology. All primary users (PUs) and secondary users (SUs) have two transceivers operating in common control channel and licensed data channels to access control and data packets in the corresponding control and data frames, respectively. PUs will inform SUs which channels will be used in next data frame and then SUs reserve unused channels by contending with each other during contention interval (CI) in control frame. The available channels which can be reserved by SUs will depend on the channel usage of PUs and further affect the throughput of SUs. Whenever SU successfully reserve channel in CI, they will start access reserved channel in next data frame after the end of CI. This will cause the wasted idle time for those SUs which have successfully reserved channel whenever the CI doesn't finish. SU which couldn't successfully reserve channel in current control frame should try to reserve again in next control frame. Thus, the duration of CI will affect the delay and wasted idle time of SUs. We analyze and simulate the performance of throughput, delay and wasted idle time in DCDSR. The analytical and simulation results show that the proposed DCDSR performs very well.

The authors examine a growable architecture for broadband packet (asynchronous transfer mode) switching, consisting of a memoryless, self-routing interconnect fabric and modest-size packet switch modules, proposed by K.U. Eng et al. (1989). They focus on the cell loss probability, because the architecture attains the best possible delay-throughput performance if the packet switch modules use output queueing. They compute an upper bound on the cell loss probability for arbitrary patterns of independent cell arrivals, possibly including isochronous circuit connections, and show that both sources of cell loss can be made negligibly small. For example, to guarantee a cell loss probability of less than 10/sup -9/, this growable architecture requires packet switch modules of dimension 47*16, 45*16, 42*16, and 39*16 for 100%, 90%, 80%, and 70% traffic loads, respectively. The analytic techniques used to bound the cell loss probabilities are applicable to other output queueing architectures. >

Chapter 9 - Performance and Architectural Issues

A three-stage packet switch architecture that uses a two stage, self-routing, memoryless, interconnection network to interconnect a third stage of smaller packet switches was proposed by K.Y. Eng. et al. (Proc. IEEE Globecom, p.1159-65, vol.2, 1989). Two new routing algorithms that facilitate the multicastings of packets with this architecture are proposed. The two routing algorithms, one of which duplicates packets at only the first stage and the other which duplicates packets at the first and third stages, require only minimal alterations to the switch. An analysis is made of the switch's multicast performance, finding an overbound on the packet loss probability. The analysis, which identifies three sources of packet loss, shows that good multicast performance is maintained with only a small increase in the architecture's expansion provided that, on average, the load is not severely concentrated onto any particular input section as a result of multicasting. >

Next-generation wireless networks have been designed to transport integrated multimedia services based on a cellular extension of a packet-switched architecture using variants of the Internet protocol (IP). Each call, arriving to or active within the network, carries demand for one or more services in parallel, where each service type has a guaranteed quality of service (QoS). Admission of new calls to the wireless IP network (WIN) from the gateway of a wired network or from a mobile subscriber (MS) is allowed by call admission control (CAC) procedures. MS roaming among the nodes of the WIN is controlled by handoff procedures between base stations (BSs), or BS controllers (BSCs), and the MSs. Both handoff and CAC procedures are typically embedded in the media access control (MAC) and logical link control (LLC) layers of the WIN protocol stack. Performance metrics, such as, probabilities of call blocking and handoff, control procedures within the WIN protocol. Earlier published results by the author present a method for the performance analysis of wireless multimedia networks based on multivariate point-process (MVPP) models and their semi-martingale representations. The models describe the finite-horizon, transient behavior of the packet flows in integrated multimedia traffic. This paper extends the earlier work to focus on the optimization of real-time control procedures embedded in the WIN protocol stack. The point processes corresponding to the packet flow of each service type and to collateral random network events are decomposed into right-continuous, pure jump processes and predictable, integrated random rate processes. Control of the set of point processes describing WIN behavior is implemented through the construction of an absolutely continuous change of a reference probability measure on network events to a controlled measure. With respect to the constructed controlled measure, the integrated random rates in the semi-martingale representations of the network point processes explicitly depend on the parameters of the protocol mechanisms for packet access, routing, switching, handoff, power control, and other resource allocations. The MVPP control approach via constructed probability measure also supports predictive models of packet flows that incorporate measurement-based estimates of the probability distributions for voice, video, data, and other Internet traffic as well as radio path distortions. Optimization of the performance metrics, represented in terms of the MVPPs and the control parameters of candidate WIN protocols for multimedia services, are discussed for both partial and complete observations of network events. The models are further used to develop optimal, recursive stochastic filters of network state, based on partial or incomplete observations of packet flow dynamics.

In modern communication systems based on Ethernet technology, the use of physical ring structures and ring redundancy protocols has been common for some time in the past. A challenge remains the configuration of such a redundancy protocol on each device with the networks stretching out over large areas, like in offshore wind power stations, where each windmill houses at least one Ethernet switch. The distance between windmills can be several kilometers and configuration of the redundancy protocol on each device is an elaborate process. In this paper, a mechanism is proposed which automatizes the configuration of ring redundancy protocols and eliminates the need to configure each device separately. The actual protocol implementation is intended to be part of the next major software release of Hirschmann Industrial Ethernet Switches.

We propose two almost-all optical packet switch architectures, called the packing switch and the scheduling switch, which when combined with appropriate wait-for-reservation or tell-and-go connection and flow control protocols provide lossless communication for traffic that satisfies certain smoothness properties. Both switch architectures preserve the order of packets that use a given input-output pair, and are consistent with virtual circuit switching. We find a lower bound on the number of elementary (2-state) switches required by any switch architecture to meet the objectives we have set. The number of 2-state switches used in the scheduling switch is of the optimal order jointly with respect to the number of inputs and with respect to the burstiness of the traffic streams (as measured by an appropriate parameter). The packing switch requires very little processing of the packet header, and uses a number of 2-state switches that is of the optimal order in terms of the burstiness parameter. We also examine the suitability of the proposed switches for the design of circuit switched networks. We find that the scheduling switch combines low hardware cost with little processing requirements at the nodes, and is an attractive architecture for both packet switched and circuit switched high-speed networks.

The NASA Lewis Research Center is developing a multichannel communication signal processing satellite (MCSPS) system which will provide low data rate, direct to user, commercial communications services. The focus of current space segment developments is a flexible, high-throughput, fault tolerant onboard information switching processor. This information switching processor (ISP) is a destination-directed packet switch which performs both space and time switching to route user information among numerous user ground terminals. Through both industry study contracts and in-house investigations, several packet switching architectures were examined. A contention-free approach, the shared memory per beam architecture, was selected for implementation. The shared memory per beam architecture, fault tolerance insertion, implementation, and demonstration plans are described.

There is a widespread interest in switching architectures that can scale in capacity with increasing interface transmission rates and higher port counts. Furthermore, packet switches that provide Quality of Service (QoS), such as bandwidth and delay guarantees, to the served user traffic are also highly desired. This report addresses the issue of constructing a high-capacity QoS-capable, multi-module switching node. Output queued switches provide the best performance in terms of throughput as well as QoS but do not scale. Input queued switches, on the other hand, require complex arbitration procedures to achieve the same level of performance. We enumerate the design constraints in the construction of a packet switch and present several approaches to build a system composed of lower-capacity memory and space elements, and analyze their performance. Towards this goal, we establish a new taxonomy for a class of switches, which we call Buffered Clos switches, and present a formal framework for optimal packet switching performance, in terms of both throughput and QoS. Within the taxonomy, we augment the existing combined input-output queueing (CIOQ) systems with Aggregation and Pipelining techniques. Furthermore, we present the design and analysis of a novel parallel packet switch architecture. For the items in the taxonomy, we present algorithms that provide optimal throughput and QoS, in accordance with the above performance framework. While some of the presented ideas are still in the investigative stage, we believe that the current state of the work, especially the formal treatment of switching, will be beneficial to the ongoing research in the field.

We describe the dimensioning and benchmarking of packet transport networks with aggregated link capacities up to the Tbit/s range, using a circuit-switching (cross-connect) infrastructure below packet-switching nodes. The different network solutions to be evaluated are designed to carry large numbers of high-definition video and other data streams, with upper bounds on packet latency and packet loss rate for all flows. Trends of evolution are analyzed, based on a performance and cost comparison of several node architectures with different associated transport technologies (packet and circuit layers). Simulations of packet-switch nodes under pseudo-self-similar traffic, extraction of analytical models for latency and loss, end-to-end best path selection, and constrained optimization of crossconnect and packet-switch resources are combined to derive the dimensioning of all the network elements. As a result of this study, photonic cross-connects associated with carrier-grade Ethernet switches appear as the most efficient architectures, thanks to cost savings for transit traffic. Single-layer solutions involving for example only packet-switch routers turn out to be more expensive.

Our work is motivated by the desire to build an evolvable architecture of Internet routers in which extremely high line-rates and IP-lookup rates are supported simply. In this paper, we consider building a packet-switch net from multi-stage, lower speed components called forwarding and switching node (FSN) performing IP-lookups and switching independently. In particular, we propose a pipelined and scalable packet switch (PSPS) architecture in which the IP lookups and packet switching for multiple packets are overlapped by accessing different FSN (in different stages) for different packets at the same time. Since this is a position paper, we do not present results but rather present the concrete approaches adopted in a PSPS and suggest interesting directions for PSPS research.

Large-scale quantum networks with thousands of nodes require scalable network protocols and physical hardware to realize. In this paper, we introduce packet switching as a paradigm for quantum data transmission in both future and near-term quantum networks. We propose a classical-quantum data frame structure and explore methods of frame generation and processing. Further, we present conceptual designs for a quantum reconfigurable optical add-drop multiplexer to realize the proposed transmission scheme. Packet switching allows for a universal design for a next-generation Internet where classical and quantum data share the same network protocols and infrastructure. In this quantum networking paradigm, entanglement distribution, as with quantum key distribution, is an application built on top of the quantum network rather than as a network designed especially for those purposes. For analysis of the network model, we simulate the feasibility of quantum packet switching for some preliminary models of quantum key and entanglement distribution. Finally, we discuss how our model can be integrated with other network models toward a realization of the quantum Internet.

Among all proposed structures for optical networks, the optical packet switching (OPS) scheme, due to its practical implementation of IPs in an optical configuration and the consequent advantages, is a prizeworthy candidate for being employed in metropolitan area network and local area network communication levels. One of the few problems frequently met using the OPS structure in the fiber-optics realm is the lack of optical buffers, thus deteriorating the system's flexibility and quality of service. For example, optical label switching networks that have been developed recently based on the generalized multiprotocol label switching protocol, profoundly suffer from this setback which is considered as a great hurdle in their evolution. In this paper, we first introduce the input and output buffer switching models while deducing their blocking probability formulations. Then, by utilizing codes in the OPS structure, we closely examine the potential of code and/or wavelength switching in packet switching networks and also determine their blocking probability. Due to close similarities between different scenarios, we present the prospect of virtual optical buffers using codes, which have close performance to input and output buffers in sight of block probability. The most significant distinction in using virtual buffers is the fact that due to their substantial nature, they happen to exhibit some error probability. However, on account of the advantageous features of codes, the simulated and formulated error probability for both coherent and noncoherent optical code division multiple access (OCDMA), such as Spectrally phase encoded OCDMA and optical orthogonal codes, using virtual buffers, tends to be negligible. Two main code and wavelength switching schemes are also brought to attention in this work, the intelligent and random methods. These scenarios demonstrate an even greater behavioral performance to that of the mere code switching scenario, leading into a more reliable adaptation of virtual buffers. In addition, simulations results happen to provide clear verifications to our analytical approach.

The so-called media convergence to the Internet is foreseen. As a consequence of this convergence, MANs will face new demands not only in terms of bandwidth, but also in terms of services. To meet these new demands, new MAN architectures are required. WDM-based MAN architectures that tackle the first problem are available not only in the literature, but also on the market. In this paper we deal with the second problem. Specifically, we describe a novel WDM-based MAN architecture that supports group communication, a service that is expected to increase considerably as new applications converge to the Internet. Based on packet switching, the architecture supports both point-to-point and point-to-multipoint communication in the optical domain.

A data frame can piggyback a control frame to increase the channel efficiency in a wireless communication such as IEEE 802.11 WLAN. However, the piggyback scheme may cause the decrease of the channel efficiency and the increase of the frame transmission delay for other stations if the station has the low transmission rate and the control frame presents the global control information such as the channel reservation time. It is similar to the anomaly phenomenon in the network which supports the multi-rate transmission. In this paper, we define this phenomenon as "the piggyback problem at low physical transmission rate" and evaluate the effect of this problem with respect to the physical transmission rate and the normalized traffic load. Then, we propose the delay-based piggyback algorithm which decides to use the piggyback scheme for a control frame based on the delay efficiency in IEEE 802.11 WLAN. In the simulation results, the proposed algorithm reduces the average frame transmission delay and the channel utilization about 24% and 25%, respectively even if there is one station which has low physical transmission rate.