TRANSMISSION OF DATA

FIELD OF THE INVENTION

The present invention relates to a method and apparatus for allocating slots for transmission of data. In particular, but not exclusively, it relates to allocating slots for transmission of data over an optical network and gathering information as to the availability of slots in such a network.

BACKGROUND TO THE INVENTION

Optical networks, such as for example, distributed Wavelength Switched Optical Networks (WSONs) support end-to-end optical paths, called lightpaths, between nodes requiring connection in the network. Intermediate nodes in this type of network support wavelength switching and may also support wavelength conversion. Establishing routes over such a network is constrained by the availability of wavelengths. This is typically addressed by resorting to the Label Set (LS) Generalized Multi-Protocol Label Switching (GMPLS) object, which collects wavelength availability information during the signalling phase of the lightpath setup process. The recent availability of flexible optical cross-connects prototypes is driving the introduction of flexible grids, i.e. frequency slots of reduced width (e.g., 6.25 GHz or 12.5 GHz) which need to be reserved in a contiguous way according to the number of frequency slots required by the lightpath.

In flexible grid network scenarios, traditional wavelength assignment strategies have been established. One such known strategy is known as "first-fit". This wavelength assignment strategy assigns a first available frequency slot for a data transmission or at least a first available contiguous number of required slots for data transm ission. However, such a strategy may not guarantee efficient network resource utilization because of the different granularity of the requests. Moreover, the evolution of the GMPLS protocol to cope with frequency slots rather than wavelengths is not straightforward since it may determine scalability problems. For instance, regarding Resource Reservation Protocol-Traffic Engineering (RSVP-TE) signalling protocol, while the typical size of LS refers to 40 or 80 wavelengths, in flexible grid scenarios, the number of frequency slots is significantly high, e.g. 320 or 640 with slot widths of 12.5 or 6.25 GHz respectively.

SUMMARY OF INVENTION The present invention seeks to obviate at least some of the disadvantages of the prior art systems and provide an improved method and apparatus for allocating slots for transmission of data.

This is achieved, according to first aspect of the present invention, by a method of allocating slots for transmission of data of a particular transmission type over an optical network. The optical network comprises a plurality of nodes which are interconnected by optical sections. The nodes support a plurality of

transmission types. The transmission of data of a particular transmission type requires a predetermined number n of consecutive slots. A first available slot is selected at an ordinal position corresponding to a multiple of n. This selected first available slot and the next n-1 consecutive slots from the selected first available slot are allocated, if all n-1 consecutive slots are available, for transmission of data. This is also achieved, according to a second aspect of the present invention, by apparatus for allocating slots for transmission of data of a particular

transmission type over an optical network. The optical network comprises a plurality of nodes which are interconnected by optical sections. The nodes support a plurality of transmission types. The transmission of data of a particular transmission type requires a predetermined number n of consecutive slots. The apparatus comprises a processor operable to select a first available slot at a location corresponding to a multiple of n and to allocate the next n-1

consecutive slots from the selected slot, if available, for transmission of data of the particular transmission type.

In this way, an efficient slot assignment strategy during lightpath setup is achieved which depends on the transmission type (value of n) i.e. the adopted transmission type (or modulation format) and bit-rate and, therefore, provides more effective slot selection strategies enabling use of Generalized Multi- Protocol Label Switching (GMPLS) for flexible grids. This is also achieved, according to a third aspect of the present invention, by a method of allocating slots for transmission of data of a particular transmission type over an optical network. The optical network comprises a plurality of nodes interconnected by optical sections. The nodes support a plurality of

transmission types. The transmission of data of the particular transmission type requires a predetermined number n of consecutive slots. The slots are divided into a plurality of grouped slots. Each group of slots comprises the

predetermined number n of consecutive slots. The slots in a group, in which all slots of the group are available, are allocated for transmission of data. This is also achieved, according to a fourth aspect of the present invention, by apparatus for allocating slots for transmission of data of a particular

transmission type over an optical network. The optical network comprises a plurality of nodes interconnected by optical sections. The nodes support a plurality of transmission types. The transmission of data of the particular transmission type requires a predetermined number n of consecutive slots. The slots are divided into a plurality of grouped slots. Each group of slots comprises the predetermined number n of consecutive slots. The apparatus comprises a processor operable to allocate a group of slots for transmission of data in which all slots are available.

In aggregating slot availability information during lightpath setup (indicating availability and allocated slots in terms of groups) overcomes scalability issues by limiting the amount of information carried by the Label Set (LS) without impacting lightpath blocking probability. The methods and apparatus above are suitable for the current GMPLS protocol suite, without requiring any additional extension and, therefore, provides more effective slot selection strategies enabling use of Generalized Multi-Protocol Label Switching (GMPLS) for flexible grids.

This is also achieved, according to another aspect of the present invention, by a method and apparatus of obtaining slot availability information in an optical network. The optical network comprises a plurality of nodes interconnected by optical sections for transmission of data of a particular transmission type. The nodes support a plurality of transmission types. The transmission of data of the particular transmission type requires a predetermined number n of consecutive slots. The slots for transmissions are divided into a plurality of grouped slots. Each group of slots comprises the predetermined number n of consecutive slots. The group in which all slots are available is indicated as available.

This is also achieved, according to another aspect of the present invention, by an optical network which comprises a plurality of nodes. The nodes support a plurality of transmission types. Transmission of data of a particular transmission type requires a predetermined number n of consecutive slots. At least one node comprises a source node and at least one other node comprises a destination node. The network further comprises at least one optical section connected by the plurality of nodes wherein the destination node sends a reservation message to establish a path for transmission of data of a particular transmission type from the source node to the destination node via the at least one optical section. The destination node comprises a processor operable to select a first available slot at an ordinal position corresponding to a multiple of n; and to allocate the next n-1 consecutive slots from the selected first available slot, if all of the n-1 slots are available, for transmission of data of a particular

transmission type.

This is also achieved, according to yet another aspect of the present invention, by an optical network which comprises a plurality of nodes. The nodes support a plurality of transmission types. Transmission of data of a particular

transmission type requires a predetermined number n of consecutive slots. At least one node comprises a source node and at least one other node comprises a destination node. The network further comprises at least one optical section connected by the plurality of nodes. The destination node sends a reservation message to establish a path for transmission of data of a particular transmission type from the source node to the destination node via the at least one optical section. The destination node comprises a processor operable to divide the slots for transmissions into a plurality of grouped slots, each group of slots comprising the predetermined number n of consecutive slots; and to allocate slots in a group, of slots in which all slots are available, for transmission of data. BRIEF DESCRIPTION OF DRAWINGS

For a more complete understanding of the present invention, reference is now made to the following description taken in conjunction with the accompanying drawings in which:

Figure 1 is a simplified schematic of apparatus for allocating slots according to an embodiment of the invention;

Figure 2(a) is a flowchart of a method of an embodiment of the invention;

Figure 2(b) is a flowchart of a method of an alternative embodiment of the invention; Figures 3(a) and (b) illustrate the label sets of the embodiments of Figures 2(a) and (b), respectively; Figure 4 is a graph representing a comparison of the blocking probability of the prior art and the embodiment of Figures 2(a) and (b).

Figure 5 is a graph representing a comparison of the number of label set elements of the prior art and the embodiment of Figures 2(a) and (b).

DETAILED DESCRIPTION OF EMBODIMENTS OF THE INVENTION

Figure 1 illustrates apparatus 101 for allocating slots for transmission of data across an optical network. The optical network comprises a plurality of nodes interconnected by optical sections i n wh ich data is com m un icated b i- directionally, such as, for example, a Wavelength Switched Optical Network (WSON). The apparatus 101 comprises an input terminal 103, a receiver 107, a processor 109, a monitor unit 1 13, a transmitter 1 1 1 and an output terminal 1 15 and may form a component of the nodes of the optical network.

At least one of the nodes of the optical network comprises a source node in which the input terminal of the node is connected to an external source of data and at least one node comprises a destination node in which the output terminal of the node is connected to an external destination. There may be any number of intermediate nodes connected therebetween such that the input terminal of each intermediate node is connected to the output terminal of a previous node in the transmission path and the output terminal of each intermediate node is connected to the input terminal of a subsequent node in the transmission path. The source node sets up a path to the destination node. Allocation of slots is determ ined at the destination node and intermediate nodes update the availability of slots and inform the destination node of the slots available. The intermediate nodes within the path are "transparent". In an embodiment, the optical sections (or links) between these nodes supports S = 320 slots of bandwidth B = 1 2.5 GHz. The nodes of the embodiment support transmission and reception at a bit-rate of 100, and 400 Gb/s. It can be appreciated that the present invention is equally applicable to networks having differing slot and bandwidth capabilities and differing bit-rates to these specified here with respect to the embodiments of the present invention.

Depending on the bit-rate and transmission type (modulation format), n consecutive frequency slots are required for such a lightpath (for transmission of the data). For example, a bit-rate of 100 Gb/s and a transmission type of dual polarization quadrature phase shift keying (DP-QPSK) and coherent detection requires 28 GHz (covered by 3 slots). In another example, a bit-rate of 400 Gb/s and a transmission type of DP- and quadrature amplitude modulation (QAM), i.e. 64-QAM, 16-QAM, or 4-QAM, and coherent detection requires 37 GHz (3 slots) , 56 G Hz (5 s lots) , and 1 1 2 G Hz (9 slots) , respectively. These requirements are set out in T. Pfau, "Hardware requirements for coherent systems beyond 100G," in ECOC 2009.

In addition to the specification of the different number of slots required for different transmission types (modulation formats), the several modulation formats also require a different minimum Optical Signal Noise Ratio (OSNR) to have a target Bit Error Rate (BER). For instance, to have a BER lower than 10 ^"3 , 64-QAM, 16-QAM, and 4-QAM require at least 24.3 dB, 20.1 dB, 16.3 dB of OSNR, respectively, as described by T. Pfau, "Hardware requirements for coherent systems beyond 100G," in ECOC 2009. The node performs path computation (e.g., source node or Path Computation Element) and is aware of physical layer information. Thus, the node is able to assess if a specific path has acceptable quality of transm ission (QoT) given the bit-rate and the modulation format. Upon receipt of a lightpath and bit-rate request, a path connecting the source node and the destination node is computed. The number of required frequency slots to serve the requested bit-rate with the best performance (in terms of bandwidth) available modulation format guaranteeing the QoT (e.g. , assuming that both 64-QAM and 16-QAM guarantees the QoT along the transmission path for a 400 Gb/s request, 64-QAM will outperform 16-QAM because the former requires 3 slots and the latter 5).

According to the signalling based on Resource Reservation Protocol-Traffic Engineering (RSVP-TE), slot availability information can be gathered within the Label Set (LS) object. If slots are treated similarly as wavelengths, LS object carries an identifier of each slot available in each link of the path. The LS object is maintained within the monitor unit 1 13 of the apparatus 101 of each node along the computed path. In the example above, the number of elements in the LS object is up to 320 identifiers (for each of the 320 slots) if the grid spacing is 12.5 GHz (bandwidth). The apparatus 101 of each intermediate node receives the LS object from the previous node along the path of that node on the input terminal 103. Each intermediate node updates the LS object within the monitor unit 1 13 of the apparatus 101 by removing the identifier of the slots which are unavailable in the outgoing link for that node. This is described in more detail below.

In an embodiment as shown in Figs. 2(a) and 3(a), a path is computed via intermediate node 305. The intermediate node 305 has an incoming link 307 and an outgoing link 309 to form a part of the computed path as shown in Fig. 3(a). The computed path may comprise any number of intermediate nodes as required to transmit data from a source node to a destination node. Each of the intermediate nodes, source node and destination node of the computed path comprise the apparatus 101 as shown in Fig. 1 and described above. The apparatus 101 of each intermediate node 305 receives at the input terminal 103, and hence the receiver 107, data for transmission along the computed path. It also receives the corresponding LS object 301 . The LS object is forwarded to the processor 109 which assesses the number (n) of slots required for data transmission and the slots available for the outgoing link 309 from the availability of slots indicated by the received LS object 301 (slots 31 1_0 to 31 1_8 as shown by the slots 31 1 ). The received LS object 301 is updated by the monitor unit 1 13 according to the slots allocated by the processor 109 and transmitted with the data, via the output terminal 1 15.

In the specific example illustrated in Fig. 3(a), a transmission type is 64-QAM with a bit-rate of 400 Gb/s giving, n=3. According to the received LS object 301 at the intermediate node 305, slots 413_0, 413_1 and 413_7 of the slots 41 1 are indicated as busy. The particular allocation of slots illustrated in Fig. 4(a) is for illustrative purposes only. At the destination node, n=3 consecutive slots are required by the specific modulation format and bit-rate of this example. These are selected as follows:

The first available slot at an ordinal position which is a multiple of n (the multiple including 0) indicated by the received LS object is determined by the processor 109 at the destination. In the example, shown in fig. 3(a), the first available slot is determined at the second ordinal position (mn, 1x3), i.e. the fourth slot 313_3 since the slot at the first ordinal position (mn, 0x3), i.e. first slot 313_0 is unavailable on the outgoing l ink 309, LS object 303. Therefore, the first available slot is considered to be the slot at the lowest ordinal position which is available. This slot is selected, steps 201 . Next it is established if the next n-1 (3-1 =2) slots 313_4, 313_5 are available, step 205. In the example shown in fig. 3(a) these are available and these n slots 313_3, 313_4, 313_5 are reserved, i.e. allocated, step 207. This updated allocation is provided to the monitor unit 1 13 which updates the LS object 303. The data is transmitted in slots 313_3, 313_4, 313_5 of slots 313 via the transmitter 1 1 1 and output terminal 1 15 onto the outgoing link 309. In respect of the prior art "first-fit" (FF) strategy, the slots 41 1_2, 41 1_3, 41 1_4 would have been allocated.

The method above operates on detailed slot availability information within the LS object. An equivalent slot selection output can be obtained if slot availability information is divided into a plurality of groups of slots (i.e. the slot availability information is grouped) depending on the number of slots required by the lightpath. This is illustrated in the alternative embodiment of Figures 2(b) and 3(b).

In this embodiment, the slots 325, 326 are divided into groups 333, 335, 337, of a predetermined, equal number of slots. The number (n) of slots within each group corresponds to the number of slots required for a particular transmission type and bit-rate. Therefore, for the example above, n=3, the slots are divided into groups of 3 consecutive slots. In particular, the slots are divided into consecutive groups of 3 consecutive slots. As a result there are no slots available in between any one group.

The slots are then allocated to available groups, steps 21 1 , 213. The LS object 321 , 323 then carries the group-identifier only, instead of the identifier for each slot. The apparatus 1 01 of each intermediate node, receives the LS object, which provides identifiers of the groups which contains available slots in the outgoing link. For example, in Fig. 3(b), the computed path includes an intermediate node 327 with an incoming link 329 and an outgoing link 331 . The LS object 321 received by the intermediate node 327 indicates that the groups 333, 335, 337 of the slots are available. The LS object 321 is provided to the monitor unit 1 13 which updates the LS object. The LS object is updated by removing groups 333 and 337 which are unavailable in link 331 (see slots 326). At the destination, the processor performs the slot assignment by selecting the group of n slots with a first-fit policy (i.e., the lowest-indexed available group) or, alternatively, as illustrated in Figure 2(b), the first group is selected and if not available the next is select until a first available group is found, steps 21 1 , 213. In the specific example of Fig. 3(b), the group 335 is selected and allocated for transmission.

The embodiment of Figs. 2(a) and 3(a) will be referred to as the "Multi-n" slot allocation strategy and the embodiment of Figs. 2(b) and 3(b) will be referred to as the "AggrFF" slot allocation strategy.

Fig. 4 shows the blocking probability vs. the offered traffic load of different slot allocation strategies, namely, the prior art "first-fit" strategy (FF), the strategy of the embodiments of Figs. 2(a) and 3(a) "Multi-n" strategy (Multi-n) and the strategy of the embodiment of Figs, 2(b) and 3(b) "AggrFF" strategy (AggrFF). The blocking probability is the probability that a data transmission will be blocked due to unavailability of slots. As generally indicated by Fig. 4, blocking increases with traffic load for all the strategies since resources are more occupied and LS objects do not reach destination, blocking the path. However, it is clear from Fig. 4 that both Multi-n and AggrFF outperforms FF, since with FF a large number of free adjacent slots cannot be selected because these are less than n. Indeed, by analyzing simulations, it was observed that groups of up to 8 consecutive slots are often free with FF, thus enabling the setup of 100 Gb/s and 16- and 64-QAM 400 Gb/s, but preventing the setup of 4-QAM 400 Gb/s. On the contrary, Multi-n and AggrFF obtain a better performance than FF since nine adjacent available slots are more likely to be found. As expected, by observing AggrFF and Multi-n, they obtain the same blocking probability given that they provide the same slot selection output.

Fig. 5 shows the average number of elements in the LS object received at destination node vs. the offered traffic load. Generally, the number of elements in LS object decreases while the traffic load increases, since a smaller amount of slots will be available in the links and, consequently, a higher amount of elements are removed from LS object. AggrFF reduces by up to 80% (at traffic loads of 200 and 700) the number of elements in LS object with respect to Multi- n and F F. Thus, AggrFF reduces the amount of control plane information carried during lightpath setup keeping a lightweight control plane. In particular, with AggrFF, instead of carrying hundreds of elements, the LS object carries an amount of elements which is comparable with the amount of elements carried in currently deployed Wavelength Switched Optical Networks (WSONs). Further AggrFF obtains low blocking probability as shown in Figure 4.

The comparisons of Figs. 4 and 5 were achieved by evaluation through a custom C++ event-driven simulator on the Spanish network with number of nodes, E=30, number of optical sections, V=56, number of slots, S=320. Lightpath QoT is considered to be acceptable (BER < 10 ^"3 before forward error correction) as described in T. Pfau, "Hardware requirements for coherent systems beyond 1 00G," in ECOC 2009 for 400 and 1 00 Gb/s, respectively. Lightpath requests are uniformly distributed among s-d pairs and bit-rates. Inter- arrival and holding times are exponentially distributed with an average of 1 /λ and 1/μ = 500s, respectively. Traffic load is expressed as λ/μ.

Although the embodiments above illustrate a particular transmission type of the 64-QAM and a transmission rate of 400 Gb/s, n=3, it can be appreciated that the embodiments are suitable for different transmission rates (bit-rates) and transmission types (modulation formats) in flex-grid WSONs with path computation either centralized (PCE) or distributed (performed by the source node).

Although embodiments of the present invention have been illustrated in the accompanying drawings and described in the foregoing detailed description, it will be understood that the invention is not limited to the embodiments disclosed, but is capable of numerous modifications without departing from the scope of the invention as set out in the following claims.