In this study, we consider wide-sense nonblocking operation of a converting-space-converting switching fabric. We propose a routing strategy based on the functional decomposition of second-stage switches, called the XsVarSWITCH routing strategy. This routing strategy is suitable for all kinds of three-stage switching fabrics operating in a slotted mode, regardless of frequency, time, or other types. We consider in this study a three-stage switching fabric. The first and third stages comprise switches with converting capability, while the middle stage is composed of space switches. The proposed routing strategy can reduce the number of required center-stage switches by almost 94% in some cases, as compared to the number of middle-stage switches under strict-sense nonblocking conditions.

In this paper, we study the wide-sense nonblocking operation for the Converting-Space-Converting switching fabric architecture. The investigated switching fabric is suitable for all kinds of slot-operating switches, no matters the slots are wavelengths, time, or others. One control algorithm for the mentioned architecture is proposed. The control algorithm decomposes the slots in the interstage links into X subsets. The proposed algorithm can reduce the required number of slots in the interstage links by almost 84%, as compared to the number of slots required in the strict sense.

Four variants of the optical data center network architecture (DCN1 - DCN4) based on optical circuit switching and elastic optical switching elements are proposed. For each variant theoretical limits for non-blocking operation have been derived and proved. The proposed networks are also compared in terms of the required number of switching elements. Our results show that variant DCN1 is more appropriate for very small data centers with tens of racks with servers, while variant DCN4 requires much fewer switching elements in data centers with hundreds or thousands of such racks.

In this paper, we address the problem of the strict-sense nonblocking operation of an elastic optical switch. Elastic optical switches are used as nodes in elastic optical networks (EONs), where fine granular frequency slots or flexible grids are used in opposite to the traditional optical networks, where fixed grids are used. The switch architecture considered in this paper consists of waveband converting switches in input and output stages, while in the center stage there is only one wavelength selective space switch without wavelength conversion capabilities. We derive the number of frequency channels k needed in interstage links to ensure the nonblocking operation of the switch when input and output links have n frequency channels and a frequency slot of one connection is not greater than m frequency channels. Both, necessity and sufficient conditions are proved.

Changes in traffic characteristics impose new requirements on interconnection network in large data centers. These requirements force changes to be done in the network architecture and used technology. Traditional tree network topology is being replaced by Clos network, called also Leaf-Spine topology, and optical interconnection networks are added in parallel to conventional electronic switching networks. Circuit switching is also used in such network instead of packet switching. In this paper we analyze the three-stage Clos network in which three multiplexing schemes are used simultaneously on input, output, and interstage links and with circuit switching mode. Strict-sense nonblocking conditions for this network with three multiplexing levels (space, wavelength, and mode) are derived and proved. Up till now such network with at most two multiplexing levels (space and time, or space and wavelength) has been considered.

This paper presents the control algorithm for non-interruptive rearrangements in log2(N, 0, p) switching networks. The proposed algorithm is able to find a plane for any new connection in the rearrangeable switching network with no more than three rearrangements. Moreover, these rearrangements can be realized in the switching network without interrupting transmission for existing connections. Rearrangements are done by setting up a new connecting path before the old connecting path for the rearranged connection is disconnected. We also describe the FPGA implementation of this algorithm for the log2(32, 0, 4) switching network. This FPGA controller has been tested and verified.

In this paper we present rearrangeable log2(N, 0, p) switching fabrics and the control algorithm for the case of an even number of stages. The main topic of this paper is the implementation of a hardware controller for such fabrics. The algorithm is described in VHDL code and realized in ML505 - the demo board for Virtex 5 - FGPA chip from the Xilinx Company. The implementation presented here works very fast, the controller can send out the set of actual signals just 20 nanoseconds after the request has been made.

This paper presents a new way of representing the internal states of the log2((N; 0; p) switching fabrics. Basing on this representation a new algorithm for rearrangements in such fabric was developed and is presented here. Also, the maximal number of rearrangements in the log2((N; 0; p) switching fabrics with odd number of stages is presented.

Wide-sense nonblocking (WSNB) multiplane banyan-type multirate switching networks with multicast connections are considered in this article. It is shown that the known results constitute in many cases only lower bounds for WSNB operation under the blocking window algorithm. New necessary and sufficient conditions are derived and proved. The obtained results can also be extended to networks with a limit imposed on the multicast group size. © 2010 Wiley Periodicals, Inc. NETWORKS, Vol. 57(4), 333-343 2011

In this article, we introduce a new space-division optical switching fabric architecture based on baseline switching networks. The new structure is built from optical switching elements (OSE) 2 × 2, 3 × 3, 2 × 3, and 3 × 2. The new structure is called the log2N-1 switching network and consists of fewer numbers of active and passive optical elements than traditional baseline switching networks composed of symmetrical OSEs. In the paper, we investigate space-division multiplane strict-sense and rearrangeable log2N-1 nonblocking switching networks and compare these with space-division multiplane baseline switching networks. The new structure has lower cost than other architectures for strict-sense nonblocking (SSNB) conditions and for rearrangeable (RNB) networks with odd number of stages. For RNB networks, when the number of stages is even, the cost of the multiplane log2N-1 optical switching network is equal to or higher than traditional networks.

This paper presents a hardware implementation of a control algorithm for the log2(N, 0, p) switching fabric. This algorithm controls both connections and disconnections in the strict sense of a nonblocking switching fabric. The hardware implementation of this algorithm in Virtex5 circuits is described. The presented implementation has been optimized in order to minimize the time response of the controller. The controller is suitable to work in applications which require very fast (even immediate) decisions. Simulations were performed and the hardware implementation shows that the controller is able to determine a plane for a new connection in one clock cycle. After this clock cycle the controller is also ready for the next connection.

In this paper, we propose the new architecture of the optical switching fabric which is based on the baseline switching network. Traditional baseline switching networks are composed of 2 × 2 (or d × d) switches. The new architecture considered in this paper is constructed from 2 × 2, 3 × 3, 2 × 3 and 3 × 2 switching elements. We show, that the proposed architecture requires less crosspoints than the traditional baseline architecture. We assumed an optical application, where semiconductor optical amplifiers are used as the optical switching elements. The proposed structure requires less number of active elements as well as passive optical splitters and combiners than the traditional baseline switching fabric. It also contains one stage fewer than the banyan network (therefore we called it the log2N - 1 switching fabric), which results in lower signal losses, distortions, and crosstalk.

A new control algorithm for log2(N, 0, p) switching networks composed of 2 x 2 switching elements has been proposed recently. Under this algorithm, log2(N,0,p) switching networks with even number of stages are wide-sense nonblocking (WSNB) if p is the same as for the rearrangeable nonblocking (RNB) one. The considerred algorithm and WSNB conditions did not take into account crosstalk constraint, which is an important factor in photonic switching. This paper enhanced this algorithm to the case when crosstalk in the switching fabric is not allowed. WSNB conditions for this enhanced algorithm are also derived. It is shown, that the number of planes required is less than those derived earlier in other papers for WSNB multiplane banyan- type switching fabrics under crosstalk constraint. Under this algorithm, log2( N,0,p) switching networks with odd number of stages and with zero crosstalk are WSNB if p is the same as for RNB one.

In this paper the new control algorithm for log2(N, 1, p) switching fabrics is proposed. This algorithm is the extension of the algorithm we propose for log2(N, 1, p) switching fabrics. The idea of the algorithm is to choose such a plane for the new connection in which this new connection will block the fewest number of possible future connections. Wide-sense nonblocking conditions are derived and proved when this algorithm is used for connection set up in such kind of switching fabrics with odd or even number of stages. The number of planes needed is the same as for the rearrangeable switching fabrics when the number of stages is even, and greater than for the rearrangeable switching fabric but lower than for the strict-sense nonblocking fabric, when the number of stages is odd.

A new control algorithm for log2(N, 0,p) switching networks composed of 2 × 2 switching elements has been proposed. Under this algorithm, log2(N, 0,p) switching networks with even number of stages are wide-sense nonblocking (WSNB) ifpis the same as for the rearrangeable nonblocking (RNB) one. In this paper we extend this solution to logd(N, 0,p) switching networks. The control algorithm is appropriately modified to switching networks composed ofd×dswitches. Wide-sense nonblocking operation of such switching networks with even number of stages is proved. We showed that number of planes needed for WSNB operation is also the same as in RNB switching networks.

A new control algorithm for multi-plane banyantype switching fabrics has been proposed recently. It has been shown that using this algorithm such switching fabrics can be wide-sense nonblocking (WSNB) when the number of planes is lower than for the strictly nonblocking (SNB) one, and for even number of stages the WSNB switching fabric requires the same number of planes as rearrangeable one. The considered algorithm and WSNB conditions did not take into account crosstalk constraint, which is an important factor in photonic switching. This paper enhanced this algorithm to the case when the first order crosstalk in the switching fabric is not allowed. WSNB conditions for this enhanced algorithm are also derived. It is shown that the number of planes required is less than those derived earlier in other papers for WSNB multiplane banyantype switching fabrics under crosstalk constraint.

A new control algorithm for the multiplane baseline switching network is proposed in this paper. This algorithm chooses a plane for a new connection in such a way that this new connection blocks the lowest number of possible future connections in the plane. We have proved, that when using this algorithm connections can be routed without blocking when the number of planes in the switching network is the same as in the rearrangeable one for even number of stages. When the switching network contains the odd number of stages, required number of planes is greater than in the rearrangeable switching network but less than in the strict-sense nonblocking one. Different implementations of the proposed algorithm are also considered. It should be noted, that the algorithm can determine the plane for a new connection in O(N0.5) time or even in O(1) time depending on the implementation. The overall time complexity of the proposed algorithm is from O(N3.5log2N) to O(N log2N) depending on its implementation and number of processors used

Recently, Hwang and Lin derived nonblocking conditions for multicast connections in Log/sub 2/(N,m,p) networks. They improved some conditions obtained earlier by Kabacin/spl acute/ski and Danilewicz. In this letter, we identify some inaccuracy in present results concerning wide-sense nonblocking Log/sub 2/(N,m,p) multicast networks, and give the correct results.

Performance evaluation of the MOQ switch recently proposed by us is discussed in this paper. In the MOQ switch both the switch fabric and buffers can operate at the same speed as input and output ports. This solution does not need any speedup in the switch fabric as well as any matching algorithms between inputs and outputs. The description and the first evaluation were presented earlier. In this paper new performance measures for the proposed MOQ switch are evaluated. The simulation studies have been carried out for switches of capacity 4/spl times/4, 8/spl times/8, 16/spl times/16 and 32/spl times/32, and under different traffic patterns. The simulations results are also compared with OQ switches and VOQ switches of the same sizes and under different scheduling algorithms: iSLIP, PIM, iRRM, MMRRS, HRRM.

In this paper the new control algorithm for multi-log/sub 2/N switching networks is proposed. Wide-sense nonblocking conditions are derived and proved when this algorithm is used for connection set up in such kind of switching networks with even number of stages. It is shown, that under this algorithm and even the number of planes required for wide-sense non-blocking operation is the same as for the rearrangeable switching networks. To our knowledge this is the first switching network which achieves the same WSNB conditions as the rearrangeable one.

In this paper, a new packet switch architecture with multiple output queuing (MOQ) is proposed. In this architecture, the nonblocking switch fabric, which has a capacity of N/spl times/N/sup 2/, and output buffers arranged into N separate queues for each output, are applied. Both switch fabric and buffers can operate at the same speed as input and output ports. This solution does not need any speedup in the switch fabric as well as arbitration logic for taking decisions which packets from inputs are to be transferred to outputs. The mean cell delay and cell loss probability, as performance measures for the proposed switch architecture, are evaluated and compared with the performance of the OQ architecture and VOQ architecture. The hardware complexity of the proposed switch is very similar to the VOQ switch but its performance is comparable to the OQ switch.

Multicast connections are used in broad-band switching networks as well as in parallel processing. We consider wide-sense and strict-sense nonblocking conditions for multi-log/sub 2/ N switching networks with multicast connections. We prove that such networks are wide-sense nonblocking if they are designed by vertically stacking at least t /spl middot/ 2/sup n-t-1/ + 2 /sup n-2t-1/ planes of a log/sub 2/ N networks together, where 1 /spl les/ t /spl les/ [n/2] and t defines the size of a blocking window K = 2/sup t/. For t = [n/2] and n even, and for [n/2] /spl les/ t /spl les/ n the number of planes must be at least t /spl middot/ 2/sup n-t-1/ + 1 and 2/sup t/ + (n - t - 1) /spl middot/ 2/sup n-t-1/ - 2/sup 2t-n-1/ + 1, respectively. In the case of strict-sense nonblocking switching networks, the number of planes is at least N/2. The results obtained in this paper show that in many cases number of planes in wide-sense nonblocking switching networks is less than those for t = [n/2] considered by Tscha and Lee (see ibid., vol.47, p.1425-31, Sept. 1999). The number of planes given in the paper is the minimum number of planes needed for wide-sense nonblocking operation provided that Algorithm 1 is used for setting up connections. The minimum number of planes for such operation in general is still open issue.

This paper studies the non-blocking switching operation of generalized three-stage Clos (1953) switching networks in the multirate environment. The analysis presented determines the minimum number of the second-stage switches required for strictly non-blocking operation of such networks at call setup. Both the discrete and the continuous bandwidth cases are considered. For the discrete bandwidth case, sufficient and necessary conditions are derived. For the continuous bandwidth case, only sufficient conditions are given, which, in some cases, also constitute necessary conditions. The results given are, in some cases, generalizations of existing results, but they also include new results.

Non-blocking operation of log/sub 2/(N.m.p.) multicast switching networks is considered. A modified algorithm proposed Tscha and Lee (1992) and by Danilewicz and Kabacinski (see IEEE International Conference on Communications ICC 2000, New Orleans, LA, USA, p.1440-4, 2000) is used for setting up connections. Conditions under which log/sub 2/(N.m.p.) switching networks are non-blocking in the wide-sense are derived and proved. It is shown that wide-sense non-blocking log/sub 2/(N.m.p.) multicast switching networks require less vertically stacked copies of log/sub 2/(N.m.l.) networks and less 2/spl middot/2 switches than multi-log/sub 2/N networks considered by Danilewicz et al.

Non-blocking operation of multi-log/sub 2/N multicast switching network is considered. An algorithm used for setting up multicast connections is proposed and conditions under which multi-log/sub 2/N switching networks are non-blocking in the wide-sense are derived and proved. Both sufficiency and necessity are considered. It is shown that non-blocking multi-log/sub 2/N multicast switching networks require more vertically stacked copies of log/sub 2/N network than point-to-point networks. It is also shown that non-blocking-conditions given by Tscha and Lee (1992) constitute only a lower bound of non-blocking operation of such switching networks.

This paper studies the nonblocking switching operation of generalised 3-stage Clos (1953) switching networks in the multirate environment. The analysis presented in this paper determines the number of the second stage switches required for strictly nonblocking operation of such networks in the continuous bandwidth case. The results given are in some cases generalisation of known results, but they also include new results.

This paper studies the non-blocking switching operation of one-sided two-stage switching networks in the multirate environment. The analysis done in this paper determines the minimum number of the second stage switches required for the strictly non-blocking operation of such networks. Both sufficient and necessary conditions are proved. We concentrate on the continuous bandwidth case, but results obtained may be extended for the discrete bandwidth case. Using non-blocking two-stage switching networks, the non-blocking networks of greater number of stages can be constructed.

A modified dilated Benes (1965) network composed of directional couplers is proposed. This structure is introduced to improve the signal-to-noise ratio (SNR) characteristics of dilated Benes networks. A new estimation of the SNR for dilated Benes networks is derived, and it is shown that this SNR is much worse than that previously known. The SNR for modified dilated Benes networks is estimated and compared to dilated Benes and other network architectures. Some other properties including the number of switching elements required, number of crossovers, and system attenuation are also derived and analyzed. Most of the characteristics are shown to be similar to dilated Benes networks and better than those of other well-known networks fabricated in Ti:LiNbO/sub 3/.

A growable ATM or fast packet switching fabric architecture is proposed. The fabric has a two-sided structure and allows for the modular growth of its size from small to very large number of ports. The fabric is expanded by connecting additional basic fabrics to the existing structure by some interconnecting modules. The expansion process does not disturb the existing cabling. The architecture makes it also possible to adjust the system throughput. The hardware complexity of the new fabric is moderate and is discussed in the paper. Some important implementation issues such as the physical design and reliability are also presented in the paper.>

This paper deals with switching networks for multichannel connections. The conditions under which the three-stage switching network is nonblocking for s-channel connections are given. Switching networks that are nonblocking in the strict sense as well as the switching networks nonblocking in the wide sense are considered. The conditions for two-sided and for one-sided switching networks are derived.>