Method and system for reallocating maximum tolerable delay in elastic optical network for edge cloud computing
Through the maximum tolerant delay redistribution method of edge cloud computing, the problems of connection request classification and spectrum resource allocation in elastic optical networks are solved, efficient utilization of network resources and reduction of blocking rate, and network performance is improved.
Patent Information
- Application Number
- CN202211346546.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-10-31
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2042-10-31
AI Technical Summary
When existing elastic optical networks process large number of connection requests, they cannot effectively classify and allocate spectrum resources, resulting in high blocking rates and low spectrum utilization, especially when huge connection requests arrive in a short period of time, they cannot meet low latency and high urgency. demand.
Adopting the maximum tolerant delay redistribution method of edge cloud computing, by constructing and initializing the edge cloud computing elastic optical network model, a candidate transmission path is established based on the maximum tolerant delay time classification request of connection request, and the number of required spectrum gaps is calculated, and combined with Edge cloud computing allocates resources, prioritizes and classification types to find free spectrum gaps, and realizes reasonable allocation of resources.
有效提升了网络资源利用率,降低了阻塞率,提高了频谱利用率和传输效率,提供了更可靠、更高效的网络服务质量。
Smart Images

Figure CN115914890B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of elastic optical network technology, and in particular to an elastic optical network maximum tolerable delay redistribution method and system for edge cloud computing. Background Art
[0002] In elastic optical networks, optical fiber spectrum resources consist of continuous spectrum slots with fixed spectrum widths. However, with the emergence of data center applications such as data migration and cloud computing, the original flexible spectrum resource allocation method is obviously insufficient. With the rapid development of today's networks, the demand for bandwidth is increasing, and the number of connection requests is also growing exponentially. When there are a large number of immediate reservation connection requests and advance reservation connection requests in the network in a short period of time, the reasonable allocation of spectrum resources through network resource allocation methods becomes increasingly important.
[0003] Nowadays, the widespread use of smartphones has made mobile data account for a considerable proportion of the total Internet traffic. New applications, such as cloud computing, video on demand, and online games, have led to an explosive growth of online information, which contains a large amount of content and extremely high network traffic. How to effectively classify and process these connection requests to make elastic optical networks more efficient. According to the urgency of different connection requests for computing resource allocation, connection requests can be classified into advance reservation connection requests and immediate reservation connection requests. According to the advantages of edge cloud computing networks, more reasonable network resource allocation can be implemented. In the advance reservation connection request, the size of the maximum tolerable delay time reflects the urgency of the advance reservation connection request for spectrum resource allocation. Therefore, for the advance reservation connection request, by utilizing the higher computing power and storage capacity in the edge cloud computing network, allocating the spectrum resources existing in the network, and allocating them to the advance reservation connection request before the maximum tolerable delay time, the high blocking rate and low spectrum utilization caused by the arrival of connection requests in a short period of time in the network can be greatly improved.
[0004] Routing and spectrum allocation are the core issues of elastic optical networks. In the existing traditional network models, there are already some routing, modulation and spectrum allocation methods for planning and supplying in elastic optical networks. However, most of these routing and spectrum allocation methods, due to the lack of high computing power and node storage capacity in traditional networks, do not consider how to handle a large number of concentrated, tolerable low-latency connection requests. They only consider the situation of handling a large number of immediate reservation connection requests, that is, immediately allocating spectrum resources to them when the connection requests arrive. However, these connection requests do not need to be served immediately upon arrival and occupy spectrum resources, which also causes a waste of spectrum resources. Therefore, the advance reservation class connection requests can be further divided according to the maximum tolerable delay of different connection requests. According to this characteristic of different connection requests, more reasonable network resource allocation can be made in edge cloud computing elastic optical networks.
[0005] Traditional elastic optical networks have shown limitations in allocating spectrum resources for connection requests when processing a large number of different types of connection requests. Most existing networks cannot guarantee tolerable low-latency services while meeting the needs of processing a large number of connection requests. With the rapid development of edge cloud computing and data centers, it is becoming increasingly important to improve the spectrum utilization and transmission efficiency of elastic optical networks. In order to improve network performance, edge cloud computing can be combined with elastic optical networks to provide more reliable and efficient network service quality. In edge cloud computing networks, connection requests can be offloaded to edge computing networks to reduce the computing pressure of the central network. On the other hand, edge cloud computing networks have high computing power and moderate response time, which can meet the low latency requirements of connection requests. These connection requests with zero initial delay tolerance are called immediate reservation connection requests because they need to be provided with connection services immediately after arrival. If the network does not have enough spectrum resources available at this time, the connection request is immediately blocked. Another type of connection request is the advance reservation connection request. These advance reservation connection requests allow a certain initial delay during the establishment of the connection as long as the spectrum resources are allocated before the maximum tolerable delay. However, when faced with a huge number of connection requests, the existing technology is unable to formulate a reasonable spectrum resource allocation plan based on different connection request types such as urgency, resulting in high congestion and affecting network resource utilization and network performance. Summary of the invention
[0006] To this end, the technical problem to be solved by the present invention is to overcome the deficiencies in the prior art and provide a method and system for maximum tolerant delay redistribution of an elastic optical network for edge cloud computing, which can effectively improve the resource utilization of the network and reduce the network blocking rate when a huge number of connection requests arrive in the network.
[0007] In order to solve the above technical problems, the present invention provides a method for redistributing the maximum tolerable delay of an elastic optical network for edge cloud computing, comprising:
[0008] Build and initialize the edge cloud computing elastic optical network model, generate connection requests, and classify the connection requests by urgency based on the maximum tolerable delay time of the connection requests;
[0009] Establishing a set of candidate transmission paths for the connection request, calculating the number of spectrum slots required for the connection request on each candidate path, and prioritizing the candidate paths;
[0010] In combination with edge cloud computing to allocate network resources, the required number of idle spectrum slots is searched on the candidate path according to the priority order and the classification type of the connection request, and a connection request is established after a successful search.
[0011] In one embodiment of the present invention, the edge cloud computing elastic optical network model is: G(N e ,N c ,N o ,E,D st ,S), where N e 、N c 、N o They represent the set of edge nodes, cloud nodes, and optical switching nodes in the network, respectively. E represents the fiber link, S represents the number of available spectrum slots in each fiber link, and D st By set N e 、N c 、N o It consists of a set of physical distances between each pair of adjacent nodes in .
[0012] In one embodiment of the present invention, each connection request in the connection request CR is CR(s, d, C, t a ,td max ,H t ), where s represents the source node of the connection request, d represents the destination node of the connection request, C represents the capacity of the connection request, and t a is the arrival time of the connection request, td max is the maximum tolerable delay time of the connection request, H t Is the hold time for the connection request.
[0013] In one embodiment of the present invention, the required number of spectrum slots N S The calculation method is:
[0014]
[0015] Among them, C represents the connection request capacity, F is the modulation level determined according to the path, and C slotIt is the bandwidth capacity of a single spectrum slot under the corresponding modulation level, and GB is the number of spectrum slots of the protection bandwidth.
[0016] In one embodiment of the present invention, the priority sorting of the candidate paths is specifically as follows:
[0017] The number of spectrum slots required by the connection request on each candidate path and the number of nodes on each candidate path are multiplied, and the candidate transmission paths are prioritized in ascending order according to the product results.
[0018] In one embodiment of the present invention, the network resources are allocated in combination with edge cloud computing. When searching for the required number of idle spectrum slots on the candidate paths according to the priority order and the classification type of the connection request, if there are no available required number of idle spectrum slots on all candidate paths, the spectrum resources required by other connection requests in the current network are first allocated through the storage and computing capabilities of the edge nodes and cloud nodes, and then a search is made on the candidate transmission paths to determine whether there are other idle spectrum resources released due to the expiration of connection requests within the maximum tolerance time of the cached service.
[0019] In one embodiment of the present invention, the connection requests are classified according to the urgency according to the maximum tolerable delay time of the connection requests, specifically:
[0020] A connection request whose maximum tolerable delay time is 0 is regarded as an immediate reservation connection request, and a connection request whose maximum tolerable delay time is not 0 is regarded as an advance reservation connection request.
[0021] In one embodiment of the present invention, searching for the number of idle required spectrum slots on the candidate path according to the priority order and the classification type of the connection request is specifically:
[0022] S1: Search in order of priority whether there are the required number of spectrum slots available on the candidate transmission path. If so, allocate resources to the connection request and the connection is established successfully. If not, execute S2.
[0023] S2: Determine the type of the connection request. If the connection request is an immediate reservation connection request, determine the connection request as blocked; if the connection request is an advance reservation connection request, execute S3;
[0024] S3: After the waiting time, check whether there are the required number of spectrum slots available on the candidate transmission path in order of priority. If so, allocate resources to the connection request and the connection is established successfully. If not, execute S4.
[0025] S4: Execute S3 until the total waiting time reaches the maximum tolerable delay time of the connection request, stop searching, and judge the connection request as blocked.
[0026] The present invention also provides an elastic optical network maximum tolerance delay redistribution system for edge cloud computing, including a network initialization module, a connection request generation module, a spectrum resource calculation module and a spectrum resource allocation module.
[0027] The network initialization module obtains network topology information and initializes edge cloud computing elastic optical network parameters;
[0028] The connection request generation module randomly selects a source node and a destination node of a connection request to generate a connection request;
[0029] The spectrum resource calculation module establishes a set of candidate transmission paths for the connection request and calculates the number of spectrum slots required for the connection request on each candidate path;
[0030] The spectrum resource allocation module prioritizes the candidate paths, allocates network resources in combination with edge cloud computing, searches for the required number of idle spectrum slots on the candidate paths according to the priority order and the classification type of the connection request, and establishes a connection request after a successful search.
[0031] In one embodiment of the present invention, it also includes a resource release module and a network status monitoring module.
[0032] After the connection request is successfully transmitted, the resource release module releases the spectrum resources occupied by the working path, releases the computing resources of the edge computing server that processes the connection request, and clears the working path information established by the connection request;
[0033] The network status monitoring module monitors the working status of the network initialization module, the connection request generation module, the spectrum resource calculation module, the spectrum resource allocation module and the resource release module in real time to obtain the real-time status of whether the connection request is successfully established or the network is blocked.
[0034] The above technical solution of the present invention has the following advantages compared with the prior art:
[0035] The present invention classifies connection requests according to their urgency, and combines the number of spectrum slots required for connection requests calculated by establishing candidate transmission paths with the allocation of spectrum slot resources by edge cloud computing, so as to effectively improve the resource utilization of the network and reduce the congestion rate of the network when a huge number of connection requests arrive in the network. BRIEF DESCRIPTION OF THE DRAWINGS
[0036] In order to make the content of the present invention more clearly understood, the present invention is further described in detail below according to specific embodiments of the present invention in conjunction with the accompanying drawings, wherein:
[0037] Figure 1 is a flow chart of the method of the present invention,
[0038] Figure 2 is a specific flow chart of the method in the embodiment of the present invention,
[0039] Figure 3 is a structural block diagram of the system of the present invention,
[0040] Figure 4 is a diagram of the NSFNet network topology in an embodiment of the present invention,
[0041] Figure 5 is a schematic diagram of link spectrum occupancy status when processing an immediate reservation connection request CR1 in an embodiment of the present invention,
[0042] Figure 6 It is a schematic diagram of link spectrum occupation status when processing an advance connection reservation request CR2 in an embodiment of the present invention. DETAILED DESCRIPTION
[0043] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments so that those skilled in the art can better understand the present invention and implement it, but the embodiments are not intended to limit the present invention.
[0044] Reference Figure 1-2 As shown, the present invention discloses a method for redistributing the maximum tolerable delay of an elastic optical network for edge cloud computing. For advance reservation connection requests, different maximum tolerable delay times also indicate the priority of the connection request in obtaining network resource allocation. First, a connection request with different maximum tolerable delays is generated. A connection request with a maximum tolerable delay of 0 is an immediate reservation connection request, and a connection request with a maximum tolerable delay of not 0 is an advance reservation connection request. Using K shortest path algorithms, a working path is established for each connection request, the available frequency slot range on each link is found, and spectrum resources are allocated according to the arrival time. If the connection request is blocked, the advance reservation connection request with the maximum tolerable delay is secondary allocated. Specifically, the following steps are included:
[0045] Step 1: Construct and initialize an edge cloud computing elastic optical network model; the edge cloud computing elastic optical network model is G(N e ,N c ,N o ,E,D st ,S), where N e 、N c 、N o They represent the set of edge nodes, cloud nodes, and optical switching nodes in the network, E represents a set of optical fiber links, S represents the number of available spectrum slots in each optical fiber link, and D st By set N e 、N c 、N oIt consists of a set of physical distances between each pair of adjacent nodes in ; for example, D st (i,j) is the actual distance between node i and node j in the network.
[0046] Step 2: Generate a set of connection requests; when tolerable delay time is not considered, specify the network to operate according to the arrival time size. Classify the connection requests by urgency according to the maximum tolerable delay time of the connection requests.
[0047] Step 2-1: Each connection request in the connection request CR is CR(s,d,C,t a ,td max ,H t ), where s represents the source node of the connection request, d represents the destination node of the connection request, C represents the capacity of the connection request, and t a is the arrival time of the connection request, td max is the maximum tolerable delay time of the connection request, H t is the holding time of the connection request. The connection request arrival time follows the Poisson process, and the connection request holding time follows the negative exponential distribution. For each immediate reservation connection request, their maximum tolerable delay time is 0, and for each advance reservation connection request, the maximum tolerable delay time is uniformly generated within a reasonable range of time.
[0048] Step 2-2: The connection request with a maximum tolerable delay time of 0, that is, the connection request that needs to be processed as soon as it arrives, is regarded as an immediate reservation connection request, and the connection request with a maximum tolerable delay time of not 0, that is, the connection request that does not need to be processed immediately upon arrival, is regarded as an advance reservation connection request. By sensing the maximum tolerable delay time of different connection requests, the urgency of the connection request for spectrum resource allocation is judged, and more reasonable network resource allocation is performed, thereby achieving better network performance improvement in terms of blocking rate and spectrum occupancy rate.
[0049] Step 3: Establish a candidate transmission path set for the connection request. In this embodiment, when a connection request CR (s, d, C, t a ,td max ,H t )∈CR arrives, the K shortest paths algorithm is used to select K available candidate paths from the source node s to the destination node d.
[0050] Step 4: Calculate the number of spectrum slots required for the connection request on each candidate path in the candidate transmission path set, and prioritize the candidate paths.
[0051] Step 4-1: Determine whether the candidate transmission path set is empty. If the candidate transmission path set is empty, it means that the path establishment has failed, and the connection request is considered blocked. If the candidate transmission path set is not empty, it means that the path establishment has succeeded. Calculate the number of spectrum slots N required for the connection request on each candidate path in the candidate transmission path set. S for:
[0052]
[0053] Among them, C represents the connection request capacity, F is the modulation level determined according to the path, and C slot is the bandwidth capacity of a single spectrum slot at the corresponding modulation level, and GB is the number of spectrum slots in the protection bandwidth. The bandwidth capacity of a single spectrum slot can be expressed as F·C slot Gb / s, for BPSK, QPSK, 8-QAM and 16-QAM, F is 1, 2, 3 and 4 respectively. For the same connection request, different routing paths are selected, and the number of spectrum slots required is N S Therefore, in the routing and spectrum allocation problem, considering the physical distance between adjacent nodes, the appropriate modulation format is selected for the arriving connection request according to the selection criteria in Table 1; N is calculated according to the modulation format through formula (1) S .
[0054] Table 1 Physical distance between candidate paths and selection criteria between modulation formats
[0055] Physical distance Selection criteria Transmission distance 3000KM-6000KM Select BPSK (M = 1) Transmission distance 1500KM-3000KM Select QPSK (M = 2) Transmission distance 750KM-1500KM Select 8-QAM (M=3) Transmission distance 0KM-750KM Select 16-QAM (M=4)
[0056] Step 4-2: Multiply the number of spectrum slots required by the connection request on each candidate path by the number of nodes on each candidate path, and prioritize the candidate transmission paths from small to large according to the product results.
[0057] Step 5: Combine edge cloud computing to allocate network resources, search for the required number of idle spectrum slots on the candidate path according to the priority order and the classification type of the connection request, and establish a connection request after successful search.
[0058] When searching for the required number of idle spectrum slots on the candidate paths, if there are no available idle spectrum slots on all candidate paths, the spectrum resources required by other connection requests in the current network are first allocated through the storage and computing capabilities of edge nodes and cloud nodes, and then the idle spectrum resources released by other connection requests due to expiration are searched on the candidate transmission paths within the maximum tolerance time of the cached service. The idle spectrum slots on the candidate paths need to meet both spectrum consistency and spectrum continuity to be available.
[0059] Step 5-1: Check whether there are the required number of idling spectrum slots on the candidate transmission path in order of priority. If so, allocate resources to the connection request and the connection is established successfully. If not, execute step 5-2.
[0060] Step 5-2: Determine the type of the connection request. If the connection request is an immediate reservation connection request, determine the connection request as blocked; if the connection request is an advance reservation connection request, execute step 5-3.
[0061] Step 5-3: After the waiting time has passed, check whether there are the required number of spectrum slots available on the candidate transmission path in order of priority. If so, allocate resources to the connection request and the connection is established successfully. If not, execute step 5-4. The value of the waiting time ts is set according to the actual situation and the duration is less than the maximum tolerable delay time td of the connection request. max .
[0062] Step 5-4: Execute step 5-3 until the total waiting time reaches the maximum tolerable delay time of the connection request and stop searching, and judge the connection request as blocked.
[0063] The method of the present invention can effectively reduce the high congestion caused by massive connection requests by processing data at the edge of the network and combining it with network resource allocation of edge cloud computing, and improve the problem of low spectrum utilization of traditional optical networks.
[0064] like Figure 3 As shown, the present invention also discloses an elastic optical network maximum tolerance delay redistribution system for edge cloud computing, including the following modules:
[0065] (1) Network initialization module: obtain network topology information and initialize edge cloud computing elastic optical network parameters; initialize the optical network link connection status, the number of node computing resources, the number of network switching nodes, the number of optical fiber links, and the bandwidth size of each spectrum slot.
[0066] (2) Connection request generation module: randomly selects the source node and destination node of the connection request and generates the connection request; the arrival time of the connection request follows the Poisson process, and the connection request holding time follows the negative exponential distribution. For each immediate reservation connection request, their maximum tolerable delay time is 0, and for each advance reservation connection request, the maximum tolerable delay time is uniformly generated within a reasonable range of time.
[0067] (3) Spectrum resource calculation module: The calculation module establishes a set of candidate transmission paths for the connection request and calculates the number of spectrum slots required for the connection request on each candidate path; the spectrum resource calculation module includes a routing calculation and modulation format selection module, which uses a K shortest path algorithm to calculate K candidate paths and stores the corresponding K candidate paths, and successively establishes routing paths for the connection request. According to the physical distance between adjacent nodes, the corresponding appropriate modulation format is selected for each routing path, and the spectrum resource calculation module calculates the number of spectrum slots N required for the connection request on each candidate path according to the modulation format. S .
[0068] (4) Spectrum resource allocation module: prioritizes candidate paths, allocates network resources in combination with edge cloud computing, searches for the required number of idle spectrum slots on candidate paths based on the priority order and the classification type of the connection request, and establishes a connection request after the search is successful. a ,td max ,H t ) Calculate the required number of spectrum gaps N S , when searching for spectrum resources required to meet the connection request in the selected working path, if the dual constraints of spectrum continuity and spectrum consistency are met at the same time, the connection request is successfully established; if the dual constraints of spectrum continuity and spectrum consistency cannot be met at the same time, the connection request fails to be established. For an immediate reservation connection request, when there are no available spectrum resources, the connection request is reported to be blocked. When processing an advance reservation connection request, if the current network does not have enough available spectrum resources to use, the connection request is cached and allocated within the maximum tolerable delay time to find out whether there are available resources to allocate to the advance reservation connection request. If it is found that sufficient network spectrum resources are generated due to the release of the connection request within the maximum tolerable delay time, spectrum resources are allocated for the advance reservation connection request.
[0069] (5) Resource release module: After the connection request is successfully transmitted, the spectrum resources occupied by the working path are released, the computing resources of the edge computing server that processes the connection request are released, and the working path information established by the connection request is cleared.
[0070] (6) Network status monitoring module: monitors the working status of the network initialization module, connection request generation module, spectrum resource calculation module, spectrum resource allocation module and resource release module in real time, and obtains the real-time status of whether the connection request is successfully established or the network is blocked.
[0071] (7) Judgment and warning module: performs coordination functions between various modules, as well as judgment and warning functions on whether each module has been successfully established, completes the purpose of routing, spectrum allocation, and spectrum resource allocation in the edge cloud computing elastic optical network, and improves the transmission performance of the network.
[0072] (8) Network performance evaluation module: After all connection requests are processed, the network performance parameters are evaluated based on the amount of spectrum resources used in the entire network, the status of successful connection requests, or network congestion.
[0073] The present invention can handle the network resource allocation problem when there are a large number of immediate reservation connection requests and advance reservation connection requests in a short period of time in the network. By defining the maximum tolerable delay time of the connection request, the system can perceive the urgency of different connection requests for network spectrum resource allocation, so as to make more reasonable spectrum allocation choices. Combining edge cloud computing with traditional elastic optical networks can provide more reliable and efficient network services.
[0074] The present invention can effectively reduce network congestion caused by massive connection requests by processing data at the network edge and allocating network resources in combination with edge cloud computing.
[0075] The present invention makes full use of the computing power and node storage capacity of edge nodes and cloud nodes in the elastic optical network to process a large number of concentrated, tolerable low-latency connection requests. It takes into account the processing of a large number of immediate reservations and advance reservation connection requests, avoiding some connection requests that do not need to be served immediately upon arrival and occupy spectrum resources, thereby reducing the waste of spectrum resources.
[0076] The present invention further considers that for reserving a group of connection requests in advance, the maximum tolerable delay time of the connection requests can be different according to the different service requirements of the specific connection requests. Through this feature of the connection requests, more reasonable network resource allocation can be continued to be made to further improve the performance of the network.
[0077] In order to further illustrate the beneficial effects of the present invention, in this embodiment, Figure 4 The experiment was carried out in the network topology shown in the figure. Figure 4 The network topology of NSFNet (a wide area network with a three-level hierarchical structure) is Figure 4 The unit of link distance is km. There are 14 nodes and 21 bidirectional links. The value on the fiber link indicates the length of the link. The bandwidth of each spectrum slot is set to 12.5 GHz, and the capacity of the fiber link is 80 spectrum slots. The capacity range C for generating connection requests is 12.5-200 Gb / s, using different modulation formats.
[0078] For the connection request CR(s,d,C,t a ,td max ,H t ), suppose a connection request CR1(2,11,100,5.5696,0,62.9133) is generated, K=3, this connection request s=2,d=11,C=100Gb / s,t a =5.5696,td max =0,H t =62.9133, the working path 2-5-12-11 has been selected, the modulation level F=2, the number of spectrum slots of the protection bandwidth is GB=1, and the number of spectrum slots used for calculation is This connection request is an immediate reservation connection request. When the connection request CR1 (2, 11, 100, 5.5696, 0, 62.9133) arrives, the spectrum occupancy status in the selected path is as follows: Figure 5 As shown, the gray indicates that the spectrum slot is occupied, and the remaining blank space represents the idle spectrum slot. When the connection request arrives, there is not enough available spectrum resources in the network, and the immediate reservation connection request CR1 (2,11,100, 5.5696, 0, 62.9133) is immediately blocked. For the advance reservation connection request CR2 (2,11,100, 6.44382, 2.41556, 35.9462), it can be seen that for CR2 (2,11,100, 6.44382, 2.41556, 35.9462), except for the arrival time t a , Maximum tolerable delay time td max , and hold time H t , the rest of the parameters are consistent with CR1 (2, 11, 100, 5.5696, 0, 62.9133). Compared with the timely reserved connection request CR1 (2, 11, 100, 5.5696, 0, 62.9133), the current link does not have available spectrum resources to meet the continuity and consistency of the connection request, which will cause network congestion. However, for the advance reserved connection request CR2 (2, 11, 100, 6.44382, 2.41556, 35.9462), when there is a certain tolerable delay time, secondary allocation can be used to complete the allocation of network resources to improve the performance of network processing services. Figure 6As shown, the advance reservation connection request CR2 (2,11,100,6.44382,2.41556,35.9462) requires 5 consecutive frequency slots. Due to insufficient bandwidth resources, the connection request cannot be provided immediately. Then the connection request is not immediately reported as blocked. Through network allocation, it is found that before the maximum tolerance delay time of CR2 (2,11,100,6.44382,2.41556,35.9462), the connection request currently occupying the spectrum resources will be released before the maximum tolerance delay time, and sufficient available spectrum resources will be generated for CR2 (2,11,100,6.44382,2.41556,35.9462) to use. By observing the spectrum occupancy status of the routing path, it is found that t = t s At time t = t1, the spectrum resources reserved for CR2 (2, 11, 100, 6.44382, 2.41556, 35.9462) in link 2-5 are released. s At time t = t + t2, part of the spectrum resources of link 5-12 and link 12-11 are also released. s +td max At time t1, the available spectrum resources in the link are sufficient to be allocated to the advance reservation connection request CR2(2,11,100,6.44382,2.41556,35.9462), where t1<t2<td max , the connection request CR2 (2,11,100,6.44382,2.41556,35.9462) is reserved in advance to complete the reallocation. If it is found that there are still insufficient spectrum resources available after the allocation, CR2 (2,11,100,6.44382,2.41556,35.9462) is reported as blocked.
[0079] The present invention takes into account the allocation of network resources and defines the blocking probability BP of service provision as follows:
[0080]
[0081] Among them, CR r is the number of blocked connection requests caused by insufficient spectrum resources on the optical fiber link in the first stage, CR s is the number of successful connection requests for reallocating network resources in the second phase, and CR is the total number of all connection requests. The blocking probability BP obtained in the experiment is compared with the blocking probability BP obtained by the traditional method without business type perception and secondary allocation. The present invention obtains a lower blocking rate and a higher spectrum occupancy rate, which proves the beneficial effect of the present invention.
[0082] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.
[0083] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0084] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0085] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0086] Obviously, the above embodiments are merely examples for clear explanation and are not intended to limit the implementation methods. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived from these are still within the protection scope of the invention.
Claims
1. A method for redistributing maximum tolerable delay in elastic optical network for edge cloud computing, characterized in that: include: Build and initialize the edge cloud computing elastic optical network model, generate connection requests, and classify the connection requests by urgency based on the maximum tolerable delay time of the connection requests; Establishing a set of candidate transmission paths for the connection request, calculating the number of spectrum slots required for the connection request on each candidate path, and prioritizing the candidate paths; Combined with edge cloud computing to allocate network resources, the number of idle spectrum slots required on the candidate path is searched according to the priority order and the classification type of the connection request, and a connection request is established after the search is successful; Each connection request in the connection request CR is CR(s, d, C, t a ,td max ,H t ), where s represents the source node of the connection request, d represents the destination node of the connection request, C represents the capacity of the connection request, and t a is the arrival time of the connection request, td max is the maximum tolerable delay time of the connection request, H t is the holding time of the connection request; The required number of spectrum slots N S The calculation method is: Among them, C represents the connection request capacity, F is the modulation level determined according to the path, and C slot is the bandwidth capacity of a single spectrum slot at the corresponding modulation level, GB is the number of spectrum slots in the protection bandwidth; The priority sorting of the candidate paths is specifically: multiplying the number of spectrum slots required by the connection request on each candidate path and the number of nodes on each candidate path, and prioritizing the candidate transmission paths in ascending order of the product result.
2. The method for redistributing the maximum tolerable delay of an elastic optical network for edge cloud computing according to claim 1 is characterized in that: The edge cloud computing elastic optical network model is: G(N e ,N c ,N o ,E,D st ,S), where N e 、N c 、N o They represent the set of edge nodes, cloud nodes, and optical switching nodes in the network, respectively. E represents the fiber link, S represents the number of available spectrum slots in each fiber link, and D st By set N e 、N c 、N o It consists of a set of physical distances between each pair of adjacent nodes in .
3. The method for redistributing the maximum tolerable delay of an elastic optical network for edge cloud computing according to claim 1 is characterized in that: The described method of allocating network resources in combination with edge cloud computing searches for the required number of idle spectrum slots on candidate paths according to the priority order and the classification type of the connection request. If the required number of idle spectrum slots is not available on all candidate paths, the spectrum resources required by other connection requests in the current network are first allocated through the storage and computing capabilities of the edge nodes and cloud nodes, and then searches for idle spectrum resources released due to the expiration of other connection requests on the candidate transmission paths within the maximum tolerance time of the cached service.
4. The method for redistributing the maximum tolerable delay of an elastic optical network for edge cloud computing according to any one of claims 1 to 3, characterized in that: The connection requests are classified according to the urgency according to the maximum tolerable delay time of the connection requests, specifically: A connection request whose maximum tolerable delay time is 0 is regarded as an immediate reservation connection request, and a connection request whose maximum tolerable delay time is not 0 is regarded as an advance reservation connection request.
5. The method for redistributing the maximum tolerable delay of an elastic optical network for edge cloud computing according to claim 4 is characterized in that: The method of searching for the number of idle required spectrum slots on the candidate path according to the priority order and the classification type of the connection request is specifically: S1: Search in order of priority whether there are the required number of spectrum slots available on the candidate transmission path. If so, allocate resources to the connection request and the connection is established successfully. If not, execute S2. S2: Determine the type of the connection request. If the connection request is an immediate reservation connection request, determine the connection request as blocked; if the connection request is an advance reservation connection request, execute S3; S3: After the waiting time, check whether there are the required number of spectrum slots available on the candidate transmission path in order of priority. If so, allocate resources to the connection request and the connection is established successfully. If not, execute S4. S4: Execute S3 until the total waiting time reaches the maximum tolerable delay time of the connection request, stop searching, and judge the connection request as blocked.
6. An elastic optical network maximum tolerable delay redistribution system for edge cloud computing, characterized by: It includes a network initialization module, a connection request generation module, a spectrum resource calculation module and a spectrum resource allocation module. The network initialization module obtains network topology information and initializes edge cloud computing elastic optical network parameters; The connection request generation module randomly selects a source node and a destination node of a connection request to generate a connection request; The spectrum resource calculation module establishes a set of candidate transmission paths for the connection request and calculates the number of spectrum slots required for the connection request on each candidate path; The spectrum resource allocation module prioritizes the candidate paths, allocates network resources in combination with edge cloud computing, searches for the number of idle required spectrum slots on the candidate paths according to the priority order and the classification type of the connection request, and establishes a connection request after the search is successful; Each connection request in the connection request CR is CR(s, d, C, t a ,td max ,H t ), where s represents the source node of the connection request, d represents the destination node of the connection request, C represents the capacity of the connection request, and t a is the arrival time of the connection request, td max is the maximum tolerable delay time of the connection request, H t is the holding time of the connection request; The required number of spectrum slots N S The calculation method is: Among them, C represents the connection request capacity, F is the modulation level determined according to the path, and C slot is the bandwidth capacity of a single spectrum slot at the corresponding modulation level, GB is the number of spectrum slots in the protection bandwidth; The priority sorting of the candidate paths is specifically: multiplying the number of spectrum slots required by the connection request on each candidate path and the number of nodes on each candidate path, and prioritizing the candidate transmission paths in ascending order of the product result.
7. The elastic optical network maximum tolerable delay redistribution system for edge cloud computing according to claim 6 is characterized by: It also includes a resource release module and a network status monitoring module. After the connection request is successfully transmitted, the resource release module releases the spectrum resources occupied by the working path, releases the computing resources of the edge computing server that processes the connection request, and clears the working path information established by the connection request; The network status monitoring module monitors the working status of the network initialization module, the connection request generation module, the spectrum resource calculation module, the spectrum resource allocation module and the resource release module in real time to obtain the real-time status of whether the connection request is successfully established or the network is blocked.
