Resource optimization method based on spatial hyper-channel network

Abstract

The invention discloses a resource optimization method based on a spatial hyper-channel network, which comprises the following steps of: firstly, allocating a shortest path for each user service request by adopting a Dijkstra shortest path algorithm when a group of user service requests arrive in an SCN (Service Control Network), then allocating channel resources, then allocating spectrum resources, and finally allocating a resource allocation path for each group of user service requests by adopting a Dijkstra shortest path algorithm; according to the method, the device cost, the network throughput and the spectrum resource utilization rate under different wavelength exchange granularities are calculated in parallel, finally, different performance indexes are normalized, the ratio of the network throughput to the device cost under different wavelength exchange granularities is evaluated, and the wavelength exchange granularities enabling the spatial hyper-channel network performance to be optimal are found. According to the method, the SCN design with different wavelength switching granularities is researched, the network throughput and the spectrum resources are considered at the same time, the wavelength switching granularities in the SCN can obviously influence the cost and performance of devices, the wavelength switching granularities are adjusted to obtain the optimal SCN design, and valuable guidance is provided for actual planning and management of SCN deployment in the future.

Description

Technical Field

[0001] The present invention belongs to the technical field of resource allocation optimization, and particularly relates to a resource optimization method based on a spatial super-channel network. Background Art

[0002] With the continuous growth of the demand for network bandwidth, the need to enhance optical transmission systems has become increasingly prominent. Utilizing a single optical channel (OCh) across the entire C-band eliminates the requirement for wavelength switching, giving rise to the spatial super-channel network (SCN). The SCN aims to provide a cost-effective solution to address the challenges posed by the upcoming era of large-scale space-division multiplexing (SDM).

[0003] The SCN was proposed in 2019 and provides a framework for solving the performance bottlenecks of SDN-based transmission systems. In the SCN, two main technologies - spatial bypass and spectrum grooming - enable optical transmission to adapt to a wide range of traffic demands.

[0004] The spatial bypass technology in the SCN is inspired by the optical bypass concept in WDM systems, which can be traced back to the early 2000s. In the upcoming era dominated by large-scale SDM, requests for transmission using the entire C-band spectrum do not require wavelength switching. Inspired by the optical bypass technology in wavelength-division multiplexing systems, the spatial bypass technology has been developed for the SCN. Through spatial channel cross-connect (SXC), this technology allows end-to-end routing without wavelength switching, providing a more cost-effective and resource-efficient solution.

[0005] Another key technology in the SCN is spectrum grooming, a strategy that has been widely recognized for its effective allocation and management of spectrum resources. As an advanced end-to-end multiplexing technology, spectrum grooming aggregates traffic demands at intermediate and terminal nodes. It has two main features: channel aggregation and dynamic spectrum allocation. Channel aggregation combines multiple low-speed channels into a single high-speed channel, thus minimizing the need for optical transceivers and increasing spectrum utilization. On the other hand, dynamic spectrum allocation allows for flexible allocation of optical channels according to different bandwidth demands, thereby optimizing the utilization of available spectrum resources.

[0006] In the SCN, an innovative reconfigurable optical add / drop multiplexer (ROADM) architecture called hierarchical optical cross-connect (HOXC) is introduced. This architecture utilizes spatial bypass and spectral grooming, bringing two benefits. First, by integrating spatial channel cross-connect (SXC) and an appropriate number of wavelength cross-connects (WXC), HOXC reduces equipment costs. Second, HOXC enables optical signals to bypass the WDM layer through SXC, eliminating the need for optoelectronic conversion and passing through multiple devices. This reduces signal loss, thereby extending the optical reach distance. These advantages further emphasize the crucial role of the HOXC architecture design for the SCN.

[0007] The wavelength exchange granularity is defined by the proportion of spatial channels (SLs) that support wavelength exchange, which is directly related to the number of WXCs deployed. This factor critically shapes the architecture of SXC and, by extension, the overall design of HOXC. In HOXC, the main switch and edge switch, known as SXC and WXC respectively, handle multiplexing and grooming at the spatial channel and wavelength levels. The spatial channels supported in SXC are divided into two categories: SLs affected by spatial bypass (spatial bypass channels) and SLs affected by wavelength exchange (wavelength exchange channels). The proportion of wavelength exchange channels determines the wavelength exchange granularity and plays a crucial role in determining the distribution of these two types of spatial channels. This granularity also affects the number of ports of WXC, and a lower granularity increases cost efficiency at the expense of routing. Summary of the Invention

[0008] To solve the above technical problems, the present invention provides a resource optimization method based on a spatial superchannel network, which solves the resource allocation problem of dynamic routing, spatial channel, and spectrum assignment (RSCSA) considering equipment cost, network throughput, and resource utilization.

[0009] The technical solution adopted by the present invention is as follows: A resource optimization method based on a spatial superchannel network, and the specific steps are as follows:

[0010] S1. When a group of user service requests arrives in the SCN, use the Dijkstra shortest path algorithm to assign the shortest path to each user service request, then assign channel resources to each user service request, and then assign spectrum resources;

[0011] Among them, after assigning the shortest path to each user service request, start to assign channel resources to each user service request. Adopt the first-fit channel allocation strategy and preferentially assign spatial bypass channels to user service requests; after assigning channel resources to each user service request, start to assign spectrum resources to each user service request, and adopt the best-fit resource allocation strategy.

[0012] S2. Based on step S1, for each group of user service requests, calculate the device cost, network throughput, and spectrum resource utilization at different wavelength switching granularities in parallel;

[0013] S21. For each group of user service requests, set the values of different wavelength switching granularities, and use the RSCSA algorithm based on the spatial super-channel network SCN to calculate the shortest path, channel resources, and spectrum resources allocated to each service request at different wavelength switching granularities in parallel;

[0014] Among them, the values of the wavelength switching granularity include 9 values, namely 0.1 - 0.9.

[0015] S22. Based on step S21, after the parallel calculation is completed, obtain the device cost, network throughput, and spectrum resource utilization of this group of user service requests at different wavelength switching granularities.

[0016] S3. Based on step S2, normalize different performance metrics, evaluate the ratio of network throughput to device cost at different wavelength switching granularities, and find the wavelength switching granularity that optimizes the performance of the spatial super-channel network;

[0017] Use the min-max scaling method to normalize different performance metrics. The normalization calculation formula is as follows:

[0018]

[0019] Among them, x represents the original feature value, x max represents the maximum value of feature x, x min represents the minimum value of feature x, x nom represents the value after normalization.

[0020] Then introduce the ratio method, that is, use the ratio of device cost to network throughput as a metric to quantify the throughput per unit device cost in different systems. By evaluating the ratio of network throughput to device cost at different wavelength switching granularities, find the wavelength switching granularity that optimizes the performance of the spatial super-channel network.

[0021] Furthermore, in step S21, the RSCSA algorithm based on the spatial super-channel network SCN is specifically as follows:

[0022] A1. For each request r(s r , d r , t r ) arriving at the SCN network, calculate the shortest path from the source node s r to the destination node d r ;

[0023] Among them, t r represents the transmission bandwidth.

[0024] A2. Determine whether the source-destination node pair of request r matches any of the in-use spatial bypass channel sets SLs. If there is an in-use spatial bypass channel set SLs whose node pair matches that of request r, go to step A3; otherwise, jump to step A8.

[0025] A3. Find the in-use spatial bypass channel set SLs that match the current request r, that is, the source-destination node pair is s r and d r of the spatial bypass channel set SLs.

[0026] A4. Based on step A3, take out a spatial bypass channel SL from the in-use spatial bypass channel set SLs with the source-destination node pair being s r and d r of the spatial bypass channel set SLs.

[0027] A5. Based on step A4, determine whether the unused frequency gaps FSs in the taken-out spatial bypass channel SL are greater than the transmission bandwidth t required by request r r , if it is greater, allocate the t of request r r to this spatial bypass channel SL, and at the same time allocate the route and frequency gaps FSs. If it is not greater, go to step A6;

[0028] A6. Fill part of the t of request r r into the spatial bypass channel SL, that is, allocate the unused FSs on the selected spatial bypass channel SL to t r , and update t r ←t r -F current ;

[0029] where F current represents the unused FSs on the selected spatial bypass channel SL.

[0030] A7. Based on step A6, determine whether there is still an available SL in the spatial bypass channel set SLs. If so, go back to step A4; if not, enter step A8;

[0031] A8. Determine whether the remaining t r is greater than the total frequency gaps FSs F of the C-band max . If it is greater, enter step A9; if not, enter step A10;

[0032] A9. Determine whether there is an idle spatial bypass channel SL. If so, allocate it for t r to find an idle spatial bypass channel for allocation, that is, allocate all the frequency gaps FSs of this SL to t r , and update t r ←t r -Fcurrent , then return to step A8, otherwise block the request r;

[0033] A10. Determine whether the remaining t r is greater than the threshold h×F max , if so, enter step A11, otherwise enter step A12;

[0034] where h represents a pre-set threshold coefficient.

[0035] A11. Determine whether there is an idle space bypass channel SL. If so, allocate an idle space bypass channel for t r , and at the same time allocate routing and frequency slots FSs. Otherwise, block the request r;

[0036] A12. Determine whether there is a suitable wavelength switching channel SL. If there is a suitable wavelength switching channel SL and there are sufficient FSs to satisfy t r , then allocate the FSs of this wavelength switching channel SL to t r , and at the same time allocate routing and frequency slots FSs. Otherwise, block the request r.

[0037] Furthermore, in the step S22, the device cost, network throughput, and spectrum resource utilization rate are specifically as follows:

[0038] (1) The device cost C of the HOXC architecture based on multi-core fiber MCF HOXC :

[0039] C HOXC = C SXC + M·C WXC + C VDGA + C SMUX+SDEMUX

[0040] where C SXC represents the device cost of the space channel cross-connect SXC, C WXC represents the device cost of the wavelength cross-connect WXC, C VDGA represents the device cost of the array variable gain dual-stage amplifier VGDA, C SMUX+SDEMUX represents the device cost of the spatial multiplexer SMUX and the spatial demultiplexer SDEMUX, and M represents the number of WXCs.

[0041] (2) Network throughput NT:

[0042]

[0043] where i represents the index of successfully processing the user service request, tr i represents the size of the i-th request, t i represents the duration of the i-th request, Ts Indicates the total time duration of the group of requests.

[0044] (3) Spectrum resource utilization rate RU:

[0045]

[0046] Where, f i Indicates the spectrum resources consumed for transmitting the i-th request, Indicates the guard band resources consumed for transmitting the i-th request.

[0047] Advantages of the present invention: The method of the present invention first uses the Dijkstra shortest path algorithm to allocate the shortest path for each user service request when a group of user service requests arrives in the SCN, then allocates channel resources, then allocates spectrum resources, and then for each group of user service requests, calculates the device cost, network throughput, and spectrum resource utilization rate under different wavelength switching granularities in parallel. Finally, the different performance metrics are normalized, and the ratio of network throughput to device cost under different wavelength switching granularities is evaluated to find the wavelength switching granularity that optimizes the performance of the space superchannel network. The method of the present invention is applicable to the actual planning and management fields of space superchannel network deployment. By studying the design of SCNs with different wavelength switching granularities, considering both network throughput and spectrum resources at the same time, the wavelength switching granularity in the SCN can significantly affect the cost and performance of the devices. Adjusting the wavelength switching granularity to obtain the best SCN design provides valuable guidance for the actual planning and management of future SCN deployments, and solves the problems of high construction cost and low resource utilization rate in existing space superchannel networks. Brief Description of the Drawings

[0048] Figure 1 Is a flowchart of a resource optimization method based on a space superchannel network of the present invention.

[0049] Figure 2 Is a flowchart of the RSCSA algorithm based on the space superchannel network SCN in an embodiment of the present invention. Detailed Embodiments

[0050] The method of the present invention will be further described below in conjunction with the drawings and embodiments.

[0051] As Figure 1 Shown, a flowchart of a resource optimization method based on a space superchannel network of the present invention, the specific steps are as follows:

[0052] S1. When a group of user service requests arrives in the SCN, use the Dijkstra shortest path algorithm to allocate the shortest path for each user service request, then allocate channel resources for each user service request, and then allocate spectrum resources;

[0053] In a spatial hyperchannel network, when a group of user service requests arrives, the network starts to respond to and process the group of user service requests, and allocates corresponding paths, channels, and spectrum resources for each user service request in the group.

[0054] Among them, after allocating the shortest path for each user service request, start to allocate channel resources for each user service request. Adopt the first-fit channel allocation strategy and preferentially allocate spatial bypass channels for user service requests. After allocating channel resources for each user service request, start to allocate spectrum resources for each user service request, and adopt the best-fit resource allocation strategy.

[0055] S2. Based on step S1, for each group of user service requests, calculate the equipment cost, network throughput, and spectrum resource utilization rate under different wavelength switching granularities in parallel;

[0056] S21. For each group of user service requests, set the values of different wavelength switching granularities, and adopt the RSCSA algorithm based on the spatial hyperchannel network SCN to calculate the shortest path, channel resources, and spectrum resources allocated to each service request under different wavelength switching granularities in parallel;

[0057] Among them, the values of the wavelength switching granularity include 9 values, namely 0.1 - 0.9.

[0058] S22. Based on step S21, after the parallel calculation is completed, obtain the equipment cost, network throughput, and spectrum resource utilization rate of this group of user service requests under different wavelength switching granularities.

[0059] S3. Based on step S2, normalize different performance metrics, evaluate the ratio of network throughput to equipment cost under different wavelength switching granularities, and find the wavelength switching granularity that optimizes the performance of the spatial hyperchannel network;

[0060] To ensure the comparability between different metrics, use the min-max scaling method to normalize different performance metrics. The normalization calculation expression is as follows:

[0061]

[0062] Among them, x represents the original feature value, x max represents the maximum value of feature x, x min represents the minimum value of feature x, x nom represents the value after normalization.

[0063] Then, the ratio method is introduced, that is, the ratio of the equipment cost to the network throughput is used as a metric to quantify the throughput per unit of equipment cost in different systems. By evaluating the ratio of the network throughput to the equipment cost under different wavelength switching granularities, the wavelength switching granularity that optimizes the performance of the spatial superchannel network is found.

[0064] As Figure 2 shown, in this embodiment, in the step S21, the RSCSA algorithm based on the spatial superchannel network SCN is specifically as follows:

[0065] A1. For each request r(s r , d r , t r ) arriving at the SCN network, calculate the shortest path from the source node s r to the destination node d r .

[0066] Among them, t r represents the transmission bandwidth.

[0067] A2. Determine whether the source-destination node pair of the request r matches any of the currently used spatial bypass channel sets SLs. If there is a node pair in the currently used spatial bypass channel set SLs that matches the node pair of the request r, go to step A3; otherwise, jump to step A8.

[0068] A3. Find the currently used spatial bypass channel set SLs that match the current request r, that is, the spatial bypass channel set SLs with the source-destination node pair s r and d r .

[0069] A4. Based on step A3, take out a spatial bypass channel SL from the currently used spatial bypass channel set SLs with the source-destination node pair s r and d r .

[0070] A5. Based on step A4, determine whether the unused frequency slots FSs in the taken-out spatial bypass channel SL are greater than the transmission bandwidth t r required by the request r. If it is greater, allocate the t r of the request r to this spatial bypass channel SL, and at the same time allocate the route and frequency slots FSs. If it is not greater, go to step A6.

[0071] A6. Fill a part of the t r of the request r into the spatial bypass channel SL, that is, allocate the unused FSs on the selected spatial bypass channel SL to t r , and update t r ←t r -F current .

[0072] Among them, F current represents the unused FSs on the selected spatial bypass channel SL.

[0073] A7. Based on step A6, determine whether there is still an available SL in the set of spatial bypass channels SLs. If so, return to step A4; if not, proceed to step A8.

[0074] A8. Determine whether the remaining t r is greater than the total frequency gap FSs F of the C-band max . If it is greater, proceed to step A9; if not, proceed to step A10.

[0075] A9. Determine whether there is an idle spatial bypass channel SL. If so, find an idle spatial bypass channel for t r and allocate all the frequency gaps FSs of this SL to t r , and update t r ←t r -F current , then return to step A8; if not, block the request r.

[0076] A10. Determine whether the remaining t r is greater than the threshold h×F max . If it is, proceed to step A11; if not, proceed to step A12.

[0077] Among them, h represents a preset threshold coefficient.

[0078] A11. Determine whether there is an idle spatial bypass channel SL. If so, find an idle spatial bypass channel for t r and allocate the routing and frequency gaps FSs simultaneously; if not, block the request r.

[0079] A12. Determine whether there is a suitable wavelength exchange channel SL. If there is a suitable wavelength exchange channel SL and it has sufficient FSs to satisfy t r , then allocate the FSs of this wavelength exchange channel SL to t r , and allocate the routing and frequency gaps FSs simultaneously; if not, block the request r.

[0080] In this embodiment, in step S22, the device cost, network throughput, and spectrum resource utilization rate are specifically as follows:

[0081] (1) The device cost C of the HOXC architecture based on the multi-core fiber MCF HOXC :

[0082] C HOXC = C SXC + M·C WXC + C VDGA+C SMUX+SDEMUX

[0083] Among them, C SXC represents the equipment cost of the space channel cross-connection SXC, C WXC represents the equipment cost of the wavelength cross-connection WXC, C VDGA represents the equipment cost of the array variable gain dual-stage amplifier VGDA, C SMUX+SDEMUX represents the equipment cost of the inter-multiplexer SMUX and the space demultiplexer SDEMUX, and M represents the number of WXCs.

[0084] This embodiment adopts the architecture of HOXC based on a 4-core multi-core fiber MCF with a node degree of 3, and the equipment cost expression is as follows:

[0085] C HOXC = C SXC + M·C WXC + C VDGA + C SMUX+SDEMUX

[0086] = 2D·{213·C(N + 2)+969·(C + N + 1)}+

[0087] M·{2DC·C 1×TWSS + 2TC·C 1×DWSS}+ 5400·DC + 2D·M·(213C + 969)

[0088] Among them, D represents the node degree of the intermediate node, C represents the number of cores of the multi-core fiber, N represents the number of outputs of the CSS, and T represents the number of transceiver supported by the WXC.

[0089] (2) Network throughput NT:

[0090]

[0091] Among them, i represents the index of successfully processing the user service request, tr i represents the size of the i-th request, t i represents the duration of the i-th request, T s represents the total time for which this group of requests lasts.

[0092] (3) Spectrum resource utilization RU:

[0093]

[0094] Among them, f i represents the spectrum resource consumed for transmitting the i-th request, f i sw represents the guard band resource consumed for transmitting the i-th request.

[0095] In summary, the method of the present invention solves the resource allocation problem in the SCN, focuses on determining the optimal wavelength switching granularity, evaluates the impact of different wavelength switching granularity levels on equipment cost, network throughput, and resource utilization, determines an optimal trade-off through analysis, and provides practical insights for network operators in network planning and management. Among them, the wavelength switching granularity of 0.2 shows the effective performance of the method of the present invention in balancing equipment cost and network throughput.

[0096] Those of ordinary skill in the art will realize that the embodiments described herein are for helping readers understand the principles of the present invention, and it should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those of ordinary skill in the art can make various other specific deformations and combinations that do not depart from the essence of the present invention based on these technical revelations disclosed by the present invention, and these deformations and combinations are still within the protection scope of the present invention.

Claims

1. A resource optimization method based on a spatial super channel network, the specific steps are as follows: S1. When a group of user service requests arrive in the SCN, the Dijkstra shortest path algorithm is used to allocate the shortest path for each user service request, and then channel resources are allocated to each user service request, and then spectrum resources are allocated; in, After allocating the shortest path to each user service request, channel resources are allocated to each user service request, and the first-fit channel allocation strategy is adopted to preferentially allocate spatial bypass channels to user service requests; after allocating channel resources to each user service request, spectrum resources are allocated to each user service request, and the best-fit resource allocation strategy is adopted; S2. Based on step S1, for each group of user service requests, the device cost, network throughput and spectrum resource utilization under different wavelength switching granularities are calculated in parallel; S21, for each group of user service requests, set different wavelength switching granularity values, adopt the RSCSA algorithm based on the spatial super channel network (SCN), and parallelly calculate the shortest path, channel resources, and spectrum resources allocated to each service request under different wavelength switching granularities; Among them, the value of wavelength exchange granularity includes 9 values, namely 0.1-0.9; S22, based on step S21, after the parallel calculation is completed, the device cost, network throughput and spectrum resource utilization of the group of user service requests at different wavelength switching granularities are obtained; S3. Based on step S2, normalize different performance indicators, evaluate the ratio of network throughput to equipment cost under different wavelength switching granularities, and find the wavelength switching granularity that optimizes the performance of the spatial super-channel network; The minimum-maximum scaling method is used to normalize different performance indicators. The normalized calculation expression is as follows: Among them, x represents the original eigenvalue, x max represents the maximum value of feature x, x min represents the minimum value of feature x, x nom Represents the normalized value; Then, the ratio method is introduced, that is, the ratio of equipment cost to network throughput is used as a metric to quantify the throughput per unit equipment cost in different systems. By evaluating the ratio of network throughput and equipment cost under different wavelength switching granularities, the wavelength switching granularity that optimizes the performance of the spatial superchannel network is found.

2. The resource optimization method based on spatial super channel network according to claim 1, characterized in that: In step S21, the RSCSA algorithm based on the spatial super channel network SCN is specifically as follows: A1. For each request r(s r ,d r ,t r ), calculate from the source node s r To the target node d r The shortest path; Among them, t r Indicates the transmission bandwidth; A2. Determine whether the source-destination node pair of the request r matches any of the spatial bypass channel sets SLs in use. If there is a node pair of the spatial bypass channel set SLs in use that matches the node pair of the request r, go to step A3; otherwise, jump to step A8. A3. Find the set of spatial bypass channels SLs in use that matches the current request r, that is, the source-destination node pair is s r and d r The spatial bypass channel set SLs; A4. Based on step A3, the source and destination nodes currently in use are s r and d r Taking out a spatial bypass channel SL from the spatial bypass channel set SLs; A5. Based on step A4, determine the transmission bandwidth t required by the unused frequency slot FSs request r in the extracted spatial bypass channel SL. r , if it is greater than t, r will be requested r Assign to the spatial bypass channel SL, and assign the route and frequency slot FSs at the same time. If it is not greater than, go to step A6; A6. Request part t of r r Fill the space bypass channel SL, that is, allocate the unused FSs on the selected space bypass channel SL to t r , and update t r ←t r -F current ; Among them, F current Indicates the unused FSs on the selected space bypass channel SL; A7. Based on step A6, determine whether there are any usable SLs in the spatial bypass channel set SLs. If yes, go back to step A4. If no, go to step A8. A8. Determine the remaining t r Is it greater than the total frequency gap FSsF of the C band? max If it is greater than, go to step A9; if it is not greater than, go to step A10; A9. Determine whether there is an empty space bypass channel SL. If so, then t r Find the free space bypass channel allocation, that is, allocate all the frequency slots FSs of the SL to t r , and update t r ←t r -F current , then return to step A8, if otherwise, block request r; A10. Determine the remaining t r Is it greater than the threshold h×F? max If yes, go to step A11, if no, go to step A12; Wherein, h represents a preset threshold coefficient; A11. Determine whether there is an empty space bypass channel SL. If so, then t r Find the free space bypass channel allocation, allocate the route and frequency slot FSs at the same time, if not, block the request r; A12. Determine whether there is a suitable wavelength exchange channel SL. If there is a suitable wavelength exchange channel SL and it has enough FSs to satisfy t r , then allocate the FSs of the wavelength switching channel SL to t r , assign routes and frequency slots FSs at the same time, otherwise block request r.

3. The resource optimization method based on spatial super channel network according to claim 1, characterized in that: In step S22, the equipment cost, network throughput and spectrum resource utilization are specifically as follows: (1) Architecture equipment cost C of HOXC based on multi-core fiber MCF HOXC : C HOXC =C SXC +M·C WXC +C VDGA +C SMUX+SDEMUX Among them, C SXC represents the equipment cost of space channel cross-connect SXC, C WXC represents the equipment cost of wavelength cross-connect WXC, C VDGA represents the equipment cost of the array variable gain dual-stage amplifier VGDA, C SMUX+SDEMUX represents the equipment cost of the space multiplexer SMUX and the space demultiplexer SDEMUX, M represents the number of WXCs; (2) Network throughput NT: Among them, i represents the index of the successfully processed user service request, tr i represents the size of the i-th request, t i represents the duration of the i-th request, T s Indicates the total duration of this group of requests; (3) Spectrum resource utilization RU: Among them, f i represents the spectrum resources consumed by transmitting the i-th request, Represents the guard band resources consumed by transmitting the i-th request.

Images