Method of determining service flow path and electronic device

Description

1 / 40 METHOD FOR DETERMINING THE SERVICE FLOW PATH AND ELECTRONIC DEVICE CROSS-REFERENCE TO RELATED REQUESTS

[001] The present application claims priority over Chinese patent application No. 202310463598.7 filed with the China Patent Office on April 20, 2023, and entitled “METHOD FOR DETERMINING SERVICE FLOW PATH AND ELECTRONIC DEVICE”, the full content of which is incorporated herein by reference. TECHNICAL FIELD

[002] The modalities of the present application relate to the technical field of data communications, in particular to a method for determining service flow paths and to an electronic device. BACKGROUND

[003] The RFC8655 standard published by the IETF describes the architecture of deterministic networks and defines the Quality of Service (QoS) objectives for deterministic routing: minimum and maximum delay from a source to a destination, and limited delay jitter; an allowed packet loss rate; and an upper limit on the transmission of unordered packets. To achieve these objectives, deterministic networks adopt measures such as resource reservation, explicit routing, and service protection. Resource reservation refers to the occupation, exclusive occupation, or sharing of a certain proportion of resources by service traffic, such as physical links, link bandwidths, and queuing resources.

[004] To meet the deterministic QoS requirements mentioned above, an individual draft (draft-peng-detnetpacket-time slot-mechanism-01) published by the IETF describes a scheduling mechanism for packet time slots. That is, time slots are converted into resources and are opened and reserved for services. However, different flows of Petition 870250094503, dated 10 / 16 / 2025, page 14 / 87 2 / 40 services have different amounts of delay demand and correspond to different traffic information, such as burst volumes, which presents certain difficulties in reserving time slot resources. When a plurality of service flows arrive at an endpoint node simultaneously, network congestion tends to occur, affecting the transfer rate of the entire communication network. SUMMARY

[005] The embodiments of the present application provide a method for determining service flow paths and an electronic device so as to at least solve the problem that network congestion tends to occur when a plurality of service flows arrive at an endpoint simultaneously in the related state of the art.

[006] To solve the above technical problems, this request is implemented as follows.

[007] In a first aspect, an embodiment of the present application provides a method for determining service flow paths, including: acquiring a quantity of delay demand of a target service flow and time slot resource information from each egress port of each node; determining at least one candidate path corresponding to the target service flow based on the time slot resource information from each egress port of each node; reserving a time slot resource for the target service flow on egress ports of candidate nodes on each candidate path, and determining an assessed value of total residence delay of the target service flow on the candidate path;and select a candidate path that satisfies a first condition as a target path for the target service flow, wherein the first condition includes that the time slot resource is successfully reserved for the target service flow on egress ports of all candidate nodes of a selected candidate path, and a sum of an evaluated delay value; Petition 870250094503, dated 10 / 16 / 2025, page 15 / 87 3 / 40 of the total residence time of the selected candidate path and a cumulative link propagation delay of the selected candidate path satisfies the delay demand amount.

[008] In a second aspect, an embodiment of the present application provides an electronic device, including a processor and a memory, wherein the memory stores at least one computer program, and at least one computer program is loaded and executed by the processor to implement the above method.

[009] In a third aspect, an embodiment of the present application provides a computer-readable storage medium storing at least one computer program, wherein the computer program is loaded and executed by a processor to implement the above method.

[010] According to the service flow path determination method provided by the embodiment of the present application, the amount of delay demand of the target service flow and the time slot resource information of each egress port of each node are obtained; at least one candidate path corresponding to the target service flow is determined based on the time slot resource information of each egress port of each node; the time slot resource is reserved for the target service flow at the egress ports of the candidate nodes on each candidate path, and the assessed value of total residence delay of the target service flow on the candidate path is determined; and the candidate path that satisfies the first condition is selected as the target path for the target service flow.

[011] Thus, it can be observed that, when using the method provided in the modality of the present application, when the target service flow arrives at an endpoint node, the corresponding target path is selected according to the amount of delay demand of the target service flow and the information of Petition 870250094503, dated 10 / 16 / 2025, page 16 / 87 4 / 40 time slot resource of each output port of each node, which can reduce the possibility of network congestion. Furthermore, the time slot resource is reserved on the output ports of the candidate nodes in the candidate path. The assessed value of the total residence delay of the target service flow in the candidate path is determined, and then whether the target path meets the amount of delay demand of the target service flow is determined according to the assessed value of the total residence delay. This provides quantitative guidance for determining the target path that meets the specific delay requirements in a network, and is conducive to quickly and accurately selecting the target path corresponding to the target service flow, thus improving the throughput of the entire communication network.

[012] It should be understood that the general description above and the detailed description below are merely illustrative and explanatory, and cannot limit the present application. BRIEF DESCRIPTION OF THE DRAWING(S)

[013] The attached drawings, which are incorporated into and form part of this descriptive report, illustrate embodiments consistent with the present application and, together with the descriptive report, serve to explain the principles of the present application.

[014] Figure 1 shows a schematic flowchart of a method for determining the service flow path provided by the modalities of the present application.

[015] Figure 2 shows a schematic diagram of a time slot resource model for a link provided by modalities of the present application.

[016] Figure 3 shows an extended network protocol structure diagram provided by embodiments of the present application.

[017] Figure 4 shows a schematic diagram of various typical and possible service traffic specifications. Petition 870250094503, dated 10 / 16 / 2025, page 17 / 87 5 / 40 results of time slot resource reservation for service flows provided by the modalities of this request.

[018] Figure 5 shows a schematic structural diagram of a model for a time slot reserved for service provided by modalities of the present application.

[019] Figure 6 shows a schematic structural diagram of a network provided by modalities of the present application.

[020] Figure 7 shows a schematic diagram of a time slot mapping relationship between two adjacent upstream and downstream nodes provided by embodiments of the present application.

[021] Figure 8 shows a schematic diagram of a displacement relationship between burst volumes (or burst subvolumes) of a service and an orchestration period provided by embodiments of the present application.

[022] Figure 9 shows a schematic structural diagram of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION

[023] Exemplary embodiments will be illustrated in detail herein, examples of which are represented in the attached drawings. In the event that the following descriptions refer to the attached drawings, the same number in the different attached drawings represents the same or similar elements, unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with this application. Rather, they are merely examples of an apparatus and method consistent with some aspects of the present application, as detailed in the attached claims.

[024] For an understanding of the modalities of this request, the relevant terms are explained below. Petition 870250094503, dated 10 / 16 / 2025, p. 18 / 87 6 / 40

[025] Time slot: a minimum unit of time-division multiplexing that needs to be designed with a reasonable value, such as 10 us, in order to be able to send at least one complete packet.

[026] Time slot scheduling: a scheduling behavior for sending a packet in a designated time slot.

[027] Input time slot: for a given intermediate node on a specific path, a time slot contained in a packet received by the intermediate node from an upstream node (i.e., an output time slot from the upstream node) is an input time slot.

[028] Output time slot: for a given intermediate node on a specific path, when continuing to send a packet to a downstream node, the intermediate node selects to send the packet in a designated time slot according to resource reservation or certain rules, and this time slot is called the output time slot.

[029] Ongoing sending time slot: For an initial node in a path, the initial node continues to send a packet received from a client side to a downstream node; and a time slot, in which the output port is currently in a sending state at the time the packet arrives at an output port, is the ongoing sending time slot. For an intermediate node in the path, the intermediate node continues to send a packet received from an upstream node to the downstream node; and a time slot, in which the output port is currently in a sending state at the time the end of an input time slot to which the packet belongs arrives at the output port, is the ongoing sending time slot. Note that the ongoing sending time slot is not an output time slot.

[030] Scheduling period: a cyclical period of a time-division multiplexing scheduling mechanism implemented by a network node, including a fixed number of slots Petition 870250094503, dated 10 / 16 / 2025, page 19 / 87 7 / 40 of time. For example, a scheduling period contains 100 fixed time slots.

[031] Service burst interval: according to the traffic specification for a deterministic service flow, generally a certain burst volume is generated within a cyclic burst interval (i.e., service burst interval). For example, if a given service flow generates 1,000 bits of traffic every 1 ms, the corresponding service burst interval is 1 ms.

[032] Orchestration period: an orchestration period adopted by a control plan according to the requirements of all deterministic services. Time slot resources are allocated to services within the orchestration period, i.e., which time slots in the orchestration period are occupied by the services and how many bits are occupied within the time slots. The orchestration period is usually an integer multiple of a scheduling period or equal to the scheduling period.

[033] In the individual draft published by the IETF, it is mentioned that, at an entry node of a network, the corresponding egress time slot resources need to be reserved on an egress port for service flows with corresponding arrival times; and, at an intermediate node of the network, the corresponding egress time slot resources need to be reserved on an egress port for service flows with corresponding ingress time slots. However, the draft does not provide a specific scheme for reserving time slot resources, and the variable amount of delay demand from different deterministic service flows and the differences in corresponding traffic information, such as burst volumes, present certain difficulties in reserving time slot resources.

[034] Therefore, an embodiment of the present application provides a method for determining service flow paths, which can Petition 870250094503, dated 10 / 16 / 2025, page 20 / 87 8 / 40 can be applied to a centralized controller or to multiple nodes in a network. Figure 1 shows a schematic flowchart of a service flow path determination method provided by an embodiment of the present application. As shown in the figure, the service flow path determination method 100 includes the following steps.

[035] In S101, a quantity of delay demand from a target service flow and time slot resource information from each output port of each node are acquired.

[036] In S102, at least one candidate path corresponding to the target service flow is determined based on the time slot resource information of each egress port of each node.

[037] In S103, a time slot resource is reserved for the target service flow on the outgoing ports of the candidate nodes on each candidate path, and an evaluated value of total residence delay of the target service flow on the candidate path is determined.

[038] In S104, a candidate path that satisfies a first condition is selected as the target path for the target service flow, wherein the first condition includes that the time slot resource is successfully reserved for the target service flow on the egress ports of all candidate nodes of a selected candidate path, and the sum of an evaluated value of total residence delay of the selected candidate path and an accumulated link propagation delay of the selected candidate path satisfies the amount of delay demand.

[039] Through the steps above, when the target service flow arrives, at least one candidate path corresponding to the target service flow is determined based on the time slot resource information of each output port of each node, and the time slot resource is reserved for the target service flow on the output ports of the candidate nodes on each candidate path, in order to select the target path that meets the Petition 870250094503, dated 10 / 16 / 2025, page 21 / 87 9 / 40 amount of delay demand for the target service flow. This can reduce the possibility of network congestion and, in the time slot resource reservation process, the assessed value of total residence delay of the target service flow on the candidate path is determined, thus providing a basis for calculating the target path that meets the specific delay requirements in the network.

[040] The illustration is made in conjunction with specific examples. Each node in the network can configure time slot resource information for a transmission link. The time slot resource information of a link can be represented as the amount of bits corresponding to all time slots within an orchestration period. Figure 2 shows a schematic diagram of a time slot resource model for a link provided by modalities of the present application. The orchestration period includes N time slots with sequence numbers from 0 to N-1, and the forwarding of resource information contained in each time slot mainly includes the following attributes.

[041] The sequence number of a time slot in the orchestration period: in the orchestration period containing N (note: N>0) time slots, the number of the first time slot is denoted as 0, and the number of the last time slot is denoted as N-1.

[042] A time slot interval timeslot_interval: in units of us (other units such as ns and ms can also be used). Generally, the interval of each time slot contained in the orchestration period is the same.

[043] The bits_quota corresponding to a time slot: in bit units (other units such as byte and word may also be used). This quota may additionally include attributes such as an initial amount (initial_amount), an amount already used by a service (used_amount), and an amount Petition 870250094503, dated 10 / 16 / 2025, page 22 / 87 10 / 40 unused (idle) (unused_amount). Since initial_amount = used_amount + unused_amount, only two of these attributes need to be included. For example, when the initial_amount in the bits_quota attribute is provided and the unused_amount attribute is additionally used to provide an idle value, it can be inferred that an initial idle value is equal to an initial value.

[044] An initial value (initial_amount) of bits_quota is a configurable default value related to the total bandwidth of a link. If the total bandwidth of the link is denoted as C and the time slot interval is denoted as ti, then initial_amount should be set to a value no greater than C*ti, which primarily considers the transmission needs of other non-deterministic service flows in the network. Generally, the initial_amount of each time slot contained in the orchestration period is the same.

[045] Because nodes in the network can notify each other about link time slot resource information, or acquire time slot resource information from multiple links in the network through a centralized controller, when the target service flow arrives, at least one candidate path corresponding to the target service flow is determined based on the time slot resource information of each output port of each node, and the time slot resource is reserved for the target service flow on the output ports of the candidate nodes on each candidate path, so as to select the target path that meets the delay requirements of the target service flow.

[046] In some cases, within a control plane, the centralized controller or an initial path node, collectively referred to as a routing module, calculates an end-to-end target path for a deterministic service flow. Before determining the target path, the routing module may select some suitable candidate paths. Petition 870250094503, dated 10 / 16 / 2025, page 23 / 87 11 / 40 for service flows in the network and then select one of the candidate paths as the target path. Candidate paths for the target service flow can be acquired based on the identification of the target service flow or the amount of demand of the target service flow for link resources, bandwidth, etc. Alternatively, candidate paths for the target service flow can be acquired based on the starting node and ending node information corresponding to the target service flow.

[047] Furthermore, after acquiring the candidate paths, the centralized controller reserves the time slot resource uniformly, or each node on the candidate paths reserves the time slot resource according to a distributed protocol (such as RSVP-TE), collectively referred to as the resource reservation module. The resource reservation module then reserves the time slot resource in the idle bit quota of a corresponding time slot of the link. If the reservation fails, such as when a link on the candidate paths does not include sufficient idle bit quota of time slots in a suitable position within the orchestration period, the selected candidate path does not meet the requirements and other candidate paths need to be tested. If no candidate path that meets the requirements is found throughout the network, the calculation result fails.

[048] When the time slot resource is successfully reserved for the target service flow on the output ports of all candidate nodes on the candidate path and the candidate path becomes the target path, the idle time slot resources of the links along the target path are changed, which will trigger a new notification of the time slot resource information of those links on the network.

[049] The acquisition of time slot resource information for each output port of each node includes: the time slot resource information for one or more time slots are Petition 870250094503, dated 10 / 16 / 2025, page 24 / 87 12 / 40 notifications are sent to each egress port via a network protocol extension, where time slot resource information includes the time slot sequence number, time slot interval, initial resource information, busy resource information, and idle resource information.

[050] In applications, an internal gateway protocol (IGP) can be extended to support notification of link delay resource information on the network. For example, an Intermediate System-to-Intermediate System (ISIS) protocol or an Open Shortest Path First (OSPF) protocol can be extended to notify resource information for one or more time slots within the link orchestration period, including the sequence numbers of the time slots, the time slot interval (time unit), an initial amount of time slot bits, and an amount of idle time slot bits.

[051] A Border Gateway Protocol (BGP) can also be extended to support notification of link delay resource information on the network or between the network and the controller. That is, resource information for one or more time slots within the link orchestration period is notified, including the sequence numbers of the time slots, the time slot interval (time unit), the initial bit count of the time slots, and the idle bit count of the time slots.

[052] Taking the ISIS extension as an example, an existing set of ISIS protocols (referring to RFC5305, RFC5311, RFC5316, and RFC5120) defined the TLVs 22 / 23 / 141 / 222 / 223 to notify an adjacency relationship. According to the modalities of the present application, a Sub-TLV, called the Time Slot Resource Sub-TLV, is added to these TLVs and used to port the Petition 870250094503, dated 10 / 16 / 2025, p. 25 / 87 13 / 40 link time slot resources. The structural diagram of the network protocol is shown in Figure 3:

[053] Type: occupies 1 byte, the value of which must be allocated by the Internet Assigned Numbers Authority (IANA) to represent that a TLV is a Time Slot Resource Sub-TLV.

[054] Length: occupies 1 byte and represents the length of TLV content (excluding the Type and Length fields).

[055] Time Slot Count: occupies 2 bytes and represents the number of time slots included in the orchestration period. These time slots are numbered sequentially from 0 to N-1.

[056] Time Slot Interval: occupies 2 bytes and represents the duration of each time slot, in microseconds (µs).

[057] The following fields are one or more groups<Número de Sequência, Quantidade Inicial, Quantidade Não Utilizada> where each group<Número de Sequência, Quantidade Inicial, Quantidade Não Utilizada> represents the initial bit count and the idle bit count of the time slot with the corresponding sequence number.

[058] Sequence Number: occupies 2 bytes and represents the sequence number of the time slot.

[059] Initial Quantity: occupies 4 bytes and represents the initial bit quantity of the timing slot.

[060] Unused Amount: occupies 4 bytes and represents the amount of idle bits in the time slot.

[061] It should be noted that the above extension is only one specific form and is not necessarily limited to this protocol extension. For example, in other extension methods, the count may be done in byte or nanosecond (ns) units. It is also possible to design other TLVs that can also characterize time slot resource information; for example, a separate sub-TLV is defined for<Número de Sequência, Quantidade Inicial, Quantidade Não Utilizada> Petition 870250094503, dated 10 / 16 / 2025, page 26 / 87 14 / 40 above. OSPF or similar protocols, as well as a BGP-LS protocol or similar protocols, can be extended to collect time slot resource information from each outgoing port of each network node, which will not be repeated here.

[062] A detailed introduction to a time slot resource reservation process for the target service flow on the egress ports of all candidate nodes in the candidate path will be provided below. The time slot resource reservation process can be implemented uniformly by a centralized controller or by multiple nodes included in the candidate path, according to a distributed protocol (such as RSVP-TE). It should be noted that, in the embodiments of this application, it is assumed that the time slots reserved by the nodes for all service flows within the orchestration period can be mapped to corresponding time slots within the scheduling period without conflict.

[063] In some possible implementations, in step S103, the reservation of the time slot resource for the target service flow on the egress ports of the candidate nodes in each candidate path includes the following.

[064] In step 1031a, an orchestration period is determined for the candidate nodes in each candidate path.

[065] In some cases, the orchestration period of the candidate nodes on each candidate path can be determined in the following ways: predefined service burst intervals corresponding to a plurality of service flows are acquired; and the orchestration period for the candidate nodes on each candidate path is determined based on the least common multiple of the service burst intervals corresponding to the plurality of service flows.

[066] Compared to the manual selection of a predefined number of time slots as the orchestration period in the related state of the art, the orchestration period determination method Petition 870250094503, dated 10 / 16 / 2025, page 27 / 87 The 15 / 40 provided in the modality of this application can be applied to time slot resource reservation for multiple target service flows, thereby improving the real-time performance of time slot resource reservation.

[067] In step 1031b, a time slot reservation subtask is determined based on the number of time slots reserved for the target service flow in the orchestration period of the candidate nodes, where the traffic corresponding to each time slot reservation subtask belongs to the target service flow.

[068] Generally, deterministic service flows have specific traffic specifications, such as the periodic generation of a specific size of traffic within a specific service burst interval. This traffic arrives at a network entry node (i.e., an initial node in the path) in a regular manner, and the offset between the traffic arrival time and the network entry node's orchestration period is fixed. Since different service burst intervals correspond to different target service flows, the discrete burst volumes within each service burst interval also vary. Therefore, in the embodiment of the present application, the time slot reservation subtask is determined based on the number of time slots reserved for the target service flow in the orchestration period of the candidate nodes.

[069] In some possible implementations, the number of time slots reserved for the target service flow in the orchestration period is determined based on the first number of service burst intervals of the target service flow in the orchestration period and the second number of discrete burst volumes within each service burst interval; and the time slot reservation subtask is determined based on the number of time slots reserved for the target service flow in the orchestration period. Petition 870250094503, dated 10 / 16 / 2025, page 28 / 87 16 / 40

[070] The illustration is made in conjunction with examples. Figure 4 shows a schematic diagram of several typical service traffic specifications and possible time slot resource reservation results for a service flow provided by modalities of the present request. As shown in the figure, the duration of the service burst intervals of deterministic service flow services 1, 2, and 3 is equal to the orchestration period. Service 1 generates a small single burst volume within its burst interval (note: small here is relative to the time slot bit quota), and a time slot 2 in the orchestration period can be reserved for service 1. Service 2 generates two small discrete burst volumes within its burst interval, and time slots 3 and 4 in the orchestration period can be reserved for service 2.Since the duration of a service burst interval for service 4 is only one-third of the orchestration period, service 4' can be built first. The duration of a service burst interval for service 4' is three times longer than that of service 4, and therefore service 4' is similar to service 2, generating three small discrete burst volumes within its burst interval, and time slots 3, 17, and 27 in the orchestration period can be reserved for service 4.

[071] Determining the number of time slots reserved for the target service flow in the orchestration period based on the first number of service burst intervals of the target service flow in the orchestration period and the second number of discrete burst volumes within each service burst interval includes that: the third number of time slots reserved for each discrete burst volume within the service burst interval in the orchestration period is determined based on the time slot resource information of the output ports of all candidate nodes in the path. Petition 870250094503, dated 10 / 16 / 2025, page 29 / 87 17 / 40 candidate; and the number of time slots reserved for the target service stream in the orchestration period is determined based on the first number of service burst intervals of the target service stream in the orchestration period, the second number of discrete burst volumes within each service burst interval, and the third number of time slots reserved for each discrete burst volume in the orchestration period.

[072] The illustration is made in conjunction with examples. As shown in Figure 4, service 3 generates a large volume of single bursts within its burst interval, and a plurality of time slots, such as time slots 20 and 21, in the orchestration period, can be reserved for service 3. In some cases, the determination of the third number of time slots reserved for each discrete burst volume within the service burst interval in the orchestration period, based on the time slot resource information of the output ports of all candidate nodes included in the candidate path, includes: the initial resource information corresponding to a single time slot of the candidate node in the orchestration period is determined based on the time slot resource information of the output ports of all candidate nodes in the candidate path.In a case where the ratio of discrete burst volumes within the service burst interval to the initial resource information corresponding to the single time slot in the orchestration period is not greater than a predefined value, the discrete burst volume is determined to correspond to a reserved time slot. Or, in a case where the ratio of discrete burst volumes within the service burst interval to the initial resource information corresponding to the single time slot in the orchestration period is greater than the predefined value, the discrete burst volume is divided into a plurality of subvolumes. Petition 870250094503, dated 10 / 16 / 2025, page 30 / 87 18 / 40 burst according to a predefined rule, each burst subvolume corresponding to a reserved time slot, where, according to the predefined rule, each burst subvolume contains at least one complete packet, and the ratio of each burst subvolume to the initial resource information corresponding to the single time slot in the orchestration period is no greater than the predefined value.

[073] The illustration is made in conjunction with examples. For each service, such as service_i, an initial node H (or a centralized controller) determines the number of reservation subtasks required for the service according to the service burst interval, the service burst volume, the bandwidth of an output port of the initial node (link-a), the duration of the orchestration period, and the time slot interval. Here, a reservation subtask corresponds to a separate time slot resource reservation request. As shown in FIG. 3, the number of reservation subtasks required by service 1 is 1, the number of reservation subtasks required by service 2 is 2, the number of reservation subtasks required by service 3 is 2, and the number of reservation subtasks required by service 4 is 3.

[074] The above method for determining the time slot reservation subtask can be summarized as follows: a) The service burst interval is aligned with the orchestration period to ensure that the durations of the two are equal. For example, if the service burst interval of service 4 is only one-third of the orchestration period, then the service burst interval of service 4 is first tripled to obtain an expanded service burst interval, which is aligned with the orchestration period.

[075] b) The number of discrete burst volumes generated by the service flow within the orchestration period is checked. For each burst volume, the ratio of the burst volume magnitude to an initial amount of bits (initial_amount) is checked. Petition 870250094503, dated 10 / 16 / 2025, p. 31 / 87 If 19 / 40 of a single time slot does not exceed a specific value, then the burst volume corresponds to a reserved subtask; or if the ratio between the magnitude of the burst volume and the initial bit amount (initial_amount) of the single time slot exceeds a specific value, the burst volume is divided into a plurality of burst subvolumes (note that each burst subvolume contains a complete packet), such that the ratio between each burst subvolume and the initial bit amount (initial_amount) of the single time slot does not exceed the specific value, and each burst subvolume corresponds to a reserved subtask.

[076] In step 1031c, for each time slot reservation subtask, the time slot resource for the target service flow is reserved on the egress ports of the candidate nodes.

[077] A path can carry a plurality of service flows, and each service flow requires a separate reserved time slot resource. In particular, for a single service flow, a plurality of time slots may need to be reserved within the orchestration period. Therefore, in the present application's modality, the time slot resource for the target service flow is reserved at the output ports of the candidate nodes for each time slot reservation subtask, respectively. Figure 5 shows a schematic structural diagram of a model for a service reservation time slot provided by the modalities of the present application. As shown in the figure, assuming that a candidate path Path_HE from initial node H to final node E passes through one or more intermediate nodes I (note: for simplicity, only one intermediate node is represented here) to provide supporting services for m service flows (service_1, ..., service_m), and within the orchestration period of each egress port along the path, the service flow service_i (1^i^m) needs to trigger cnt_i times of time slot resource reservation. Petition 870250094503, dated 10 / 16 / 2025, page 32 / 87 20 / 40

[078] In some possible implementations, in step S103, reserving the time slot resource for the target service flow on the egress ports of the candidate nodes on each candidate path, and determining the estimated total residence delay value of the target service flow on the candidate path includes the following.

[079] In step 1032a, a residence delay budget value is determined for each candidate node in the candidate path based on the amount of delay demand and the accumulated link propagation delay of the candidate path.

[080] The illustration is made in conjunction with specific examples. Figure 6 shows a schematic structural diagram of a network provided by modalities of the present application. As shown in the figure, the network includes six nodes: node S, node A, node C, node E, node D, and node B. A link between each two nodes is a bidirectional link. The propagation delay parameters of each link are marked in the figure. For example, a link propagation delay between node S and node A is 20 us. Assuming that an end-to-end delay required by the target service flow is 200 us, which is the same for reservation subtasks corresponding to all burst volumes of the target service flow. The centralized controller or initial node S first determines a candidate path path_SD that passes through SACED. The accumulated link propagation delay of the current candidate path is: 20 us + 10 us + 10 us + 10 us = 50 us.The difference between the amount of delay demand of the target service flow and the accumulated link propagation delay is determined as the total residence delay budget value (total_residence_budget) of each candidate node in the candidate path, which is 200 us - 50 us = 150 us. Assuming that, according to the average budgeting method, each candidate node will share a residence delay budget value of 150 / 5 = 30 us.

[081] It should be noted that the average budget method can be used to divide the residence arrears budget amount. Petition 870250094503, dated 10 / 16 / 2025, page 33 / 87 21 / 40 total (total_residence_budget) multiplied by the number of all candidate nodes in the candidate path (denoted as node_count), to obtain an average residency delay budget value, namely: averaged_node_residence_budget = total_residence_budget / node_count. Then, averaged_node_residence_budget is used as the residency delay budget value for each node (i.e., node_residence_budget). It is also possible to specify the corresponding residency delay budget value for each candidate node according to actual needs. For example, a specified budget list for node residency delay is configured in advance. For example, a specified budget list for node residency delay is configured in advance.When reading the specified node residence delay budget list, it is determined that a specified residence delay budget for the initial node H is node_h_residence_specified_budget, which is used as the residence delay budget value for node H; a specified residence delay budget for an intermediate node I is node_i_residence_specified_budget, which is used as the residence delay budget value for node I; and a specified residence delay budget for the final node E is node_e_residence_specified_budget, which is used as the residence delay budget value for node E. Furthermore, total_residence_budget = node_h_residence_specified_budget + node_i_residence_specified_budget + node_e_residence_specified_budget is satisfied.

[082] In step 1032b, for each candidate node in the candidate path, a target time slot resource is reserved for the target service flow, according to the time slot resource information of an egress port of the candidate node and an estimated residence delay value of the candidate node, and an estimated residence delay value of the candidate node is determined. Petition 870250094503, dated 10 / 16 / 2025, page 34 / 87 22 / 40

[083] The illustration is made in conjunction with specific examples. As shown in Figure 6, in the direction of the candidate path, from the initial node S to the final node D, the time slot resource is reserved for the target service flow at the egress port of each candidate node sequentially. For each candidate node, a target time slot is reserved for the target service flow, according to the time slot resource information of the candidate node's egress port and the estimated residence delay value of the candidate node determined in the previous step, and the estimated residence delay value corresponding to the candidate node is determined.

[084] In the process of reserving the target time slot for the target service flow, an embodiment of the present application provides a method of automatically reserving resources based on a delay budget. In the direction from an initial node to an end node in a candidate path, an egress time slot resource is reserved on an egress port of each node sequentially, according to a residence delay budget adjustment value (node_residence_budget_adjustment) of the current node, such that the rated residence delay value corresponding to the reserved egress time slot is equal to or nearly equal to the residence delay budget adjustment value (node_residence_budget_adjustment) of the current node.That is, in step 1032b, reserve the target time slot resource for the target service flow according to the time slot resource information of the candidate node's egress port and the residency delay budget value of the candidate nodes, and determine the assessed residency delay value of the candidate node includes that: a residency delay offset of the candidate node is determined according to an assessed cumulative residency delay value and a cumulative residency delay budget value of all upstream nodes corresponding to the candidate node; a delay budget adjustment value of. Petition 870250094503, dated 10 / 16 / 2025, page 35 / 87 23 / 40 residency of the candidate node is determined based on the residency delay offset and the residency delay budget value of the candidate nodes; an offset of the target time slot reserved by the candidate node for the target service flow relative to an ongoing dispatch time slot of the egress port is determined based on the time slot resource information of the candidate node's egress port and the residency delay budget adjustment value, where the ongoing dispatch time slot is used to represent a time slot that is currently in a dispatch state at a target time on the candidate node's egress port; in a case where the candidate node is an initial node, the target time is a time when the target service flow arrives at the candidate node's egress port;Or, in a case where the candidate node is an intermediate node, the target time is a time when the end of a reserved target time slot on an output port of an upstream target node reaches the output port of the candidate node, the upstream target node being adjacent to the candidate node; and the evaluated residence delay value of the candidate node is determined based on the offset.

[085] In one implementation, when the time slot resource is reserved on the outbound port of the initial node H, an accumulated node residence delay deviation (accumulated_node_residence_deviation) is an initial value of 0. Therefore, the node residence delay budget adjustment value (node_residence_budget_adjustment) of the initial node H is equal to the node residence delay budget value (node_residence_budget).

[086] When the time slot resource is reserved on the egress port of any node, if the assessed residence delay value corresponding to the reserved egress time slot is equal to the intra-node residence delay budget adjustment value (node_residence_budget_adjustment), the accumulated deviation Petition 870250094503, dated 10 / 16 / 2025, p. 36 / 87 The 24 / 40 intra-node residence delay (accumulated_node_residence_deviation) for the next node in the path will be 0, and the residence delay budget adjustment value (node_residence_budget_adjustment) of the next node will be equal to the residence delay budget value (node_residence_budget) of its own node. Under normal circumstances, if the candidate node can reserve the expected time slot according to its residence delay budget value, there will be no difference between the assessed residence delay value and the residence delay budget value of the candidate node. However, in some cases, due to insufficient resources in the expected time slot, it is necessary to reserve resources in a specific time slot close to the expected time slot, which may lead to a difference between the assessed residence delay value and the residence delay budget value of the candidate node.

[087] In one implementation, for a candidate node I, there is a fixed phase difference between a time slot of candidate node I and a time slot of an adjacent upstream node. The embodiment of the present application adopts an existing method for determining a mapping relationship from any output time slot of the upstream node to an ongoing output time slot of node I, especially when the time slot intervals between the two adjacent nodes are not the same. Assuming that candidate node I has established a mapping relationship as shown in FIG.7, as shown in the figure, the upstream node U (such as the initial node H or other candidate nodes) of candidate node I sends a packet in any output time slot X from an output port (link a), after a link propagation delay period (d_p) and an intranode forwarding delay (d_f), the end of time slot X arrives at an output port (link b) of candidate node I no later than time slot Y. That is, the output time slot X of the output port (link a) of... Petition 870250094503, dated 10 / 16 / 2025, page 37 / 87 The 25 / 40 upstream node is mapped to the ongoing transmission time slot Y of the output port (link b) of node I, and the remaining time before the end of time slot Y is T_ui.

[088] Thus, assuming that a given egress time slot is reserved for a given reservation subtask on the egress port of the upstream node, the ongoing send time slot of the egress port of node I to which the egress time slot is mapped is numbered j, the time remaining before the end of the ongoing send time slot is t_remaining, and the reserved egress time slot is offset by x (^1) time slots after time slot j, i.e., the sequence number of the egress time slot is (j+x)%N, where N is the count of time slots contained in the orchestration period and % is a remainder operator. Then, the evaluated residence delay value of the candidate node is determined based on the offset x.The assessed residence delay value is determined as follows: a first time slot interval corresponding to the target upstream node and a second time slot interval corresponding to the candidate node are acquired, the target upstream node being a node adjacent to the candidate node among the upstream nodes; an assessed queuing delay value at the candidate node's output port is determined based on the offset, the first time slot interval, and the second time slot interval; a remaining time from the target moment until the end of the candidate node's ongoing sending time slot is determined; and a sum of the assessed queuing delay values, an intra-node forwarding delay of the candidate node, and the remaining time is determined as the assessed residence delay value of the candidate nodes.

[089] The rated queuing delay value comprises a best rated queuing delay value, a worst rated queuing delay value, and an average rated queuing delay value, and the determination of the value Petition 870250094503, dated 10 / 16 / 2025, page 38 / 87 The 26 / 40 rated queuing delay at the candidate node's output port based on offset, in the first time slot interval and the second time slot interval, includes selecting one of the best rated queuing delay values, the worst rated queuing delay values, and the average rated queuing delay value as the rated queuing delay value at the candidate node's output port according to a predefined strategy.

[090] In some cases, the best rated queuing delay value Tbest can be determined according to the following formula: Tbest=(O-1) *Tv.

[091] Or, the worst-rated Tworst queuing delay value is determined according to the following formula: Tworst = Tu + O*Tv .

[092] Or, the average rated queuing delay value Tave is determined according to the following formula: Tave=(Tu+(2*O-1) *Tv) / 2.

[093] Where, O is the offset, Tu is the first time slot interval corresponding to the target upstream node, and Tv is the second time slot interval corresponding to the candidate nodes. It should be noted that, for the initial node, the first time slot interval corresponding to the target upstream node adjacent to the initial node is 0.

[094] The best-rated queuing delay value, or the worst-rated queuing delay value, or the average-rated queuing delay value mentioned above is selected to be used to determine the rated residence delay value of the candidate node. The rated residence delay value is compared with the intra-node residence delay budget adjustment value (node_residence_budget_adjustment), so that when an appropriate x is selected, the two are equal or nearly equal, and Petition 870250094503, dated 10 / 16 / 2025, page 39 / 87 27 / 40 the idle bit quantity resource of the corresponding output time slot (j+x)%N meets the burst volume (or burst subvolume) requirements corresponding to the reservation subtask. Finally, the sequence number of the output time slot reserved on the output port of intermediate node I for the backup subtask and the number of bits occupied by the backup subtask within the time slot are maintained.

[095] In some possible implementations, after determining the assessed residence delay value of the candidate node based on the offset, the method additionally includes that: a target time slot reserved by the candidate node for the target service flow in the orchestration periods is determined based on the offset; and in a case where an idle time slot resource of the determined target time slot is not less than the time slot resource required by the target service flow, and a difference between the residence delay budget value and the assessed residence delay value that is determined according to the offset is less than a first threshold, it is determined that the time slot resource is successfully reserved for the target service flow on the candidate node's egress port.

[096] In some possible implementations, after determining the evaluated residence delay value of the candidate node based on the offset, the sequence number of the reserved output time slot on the output port of candidate node I for the reservation subtask and the amount of bits occupied by the reservation subtask within the time slot are maintained. In some cases, a first mapping relation between a reserved target time slot on the output port of the target upstream node and a reserved target time slot on the output port of the candidate node is generated; and the target service flow sent by the target upstream node is received, and the reserved target time slot for the flow of Petition 870250094503, dated 10 / 16 / 2025, page 40 / 87 The 28 / 40 target service at the egress port is determined based on the initial mapping relationship.

[097] In some cases, in a case where the candidate node is the initial node, the time slot reservation subtask corresponding to the target service flow is received; the ongoing sending time slot of the candidate nodes is determined based on a position mapping relationship of an arrival time of the time slot reservation subtask within the orchestration period of the initial node and an intra-node forwarding delay corresponding to the candidate node; and based on the time slot resource information of the initial node's egress port and the residence delay budget value of the initial node, as well as the ongoing sending time slot, the offset of the target time slot reserved by the candidate node for the target service flow relative to the ongoing sending time slot of the egress port is determined, where the target time slot corresponds to the target time slot resource.

[098] In one implementation, at the initial node H, there is a fixed offset relationship between the orchestration period and the arrival time of traffic from a burst volume (or burst subvolume) corresponding to a specific reservation subtask. That is, the arrival time of the traffic will be within a fixed ongoing sending time slot. Figure 8 shows a schematic diagram of an offset relationship between a burst volume (or burst subvolume) of a service and an orchestration period provided by the present request modes. As shown in the figure, the first burst volume (burst_1) of service_i arrives at the initial node H, and arrives at the egress port link-a after an intra-node forwarding delay period (denoted as d_f), the ongoing sending time slot at that moment is time slot 1, and there is still remaining_1 remaining before the end of time slot 1. The Petition 870250094503, dated 10 / 16 / 2025, p. 41 / 87 29 / 40 last burst volume (burst_cnt_i) from service_i arrives at the initial node H, and arrives at the outgoing link-a after an intra-node forwarding delay period (denoted as d_f), the ongoing sending time slot at that moment is time slot N2, and there is still remaining_cnt_i remaining before the end of time slot N-2.

[099] Generally, for a given reserved subtask, assuming that the ongoing send time slot to which the burst volume (or burst subvolume) corresponding to the given reserved subtask is mapped is numbered j, the time remaining before the end of the ongoing send time slot is t_remaining, and the reserved output time slot is shifted by x (^1) time slots after time slot j, i.e., the sequence number of the output time slot is (j+x)%N, where N is the count of time slots contained in the orchestration period and % is a remainder operator.

[100] The best-rated queuing delay value, or the worst-rated queuing delay value, or the average-rated queuing delay value mentioned above is selected to be used to determine the rated residence delay value of the initial node. The rated residence delay value is compared with the intra-node residence delay budget adjustment value (node_residence_budget_adjustment), so that when an appropriate offset x is selected, the two are equal or nearly equal, and the corresponding j+x output time slot meets the burst volume (or burst subvolume) requirements corresponding to the reserved subtask. Finally, the sequence number of the reserved output time slot at the output port of the initial node H of the reserved subtask and the amount of bits occupied by the reserved subtask within the time slot are maintained. Petition 870250094503, dated 10 / 16 / 2025, page 42 / 87 30 / 40

[101] In some possible implementations, after selecting the candidate path that satisfies the first condition as the target path for the target service flow, the method additionally includes that: a second mapping relationship between the time slot reservation subtask and the target path, a third mapping relationship between the time slot reservation subtask and a target time slot of the initial node's egress port, and a fourth mapping relationship between the time slot reservation subtask and a list of time slots composed of target time slots of egress ports of all candidate nodes in the target path are generated;Traffic corresponding to the time slot reservation subtask of the target service flow is received, and a target time slot for the traffic corresponding to the time slot reservation subtask at the egress port of each candidate node in the target path is determined based on the second mapping relationship, the third mapping relationship, and the fourth mapping relationship.

[102] In one implementation, the outbound time slot resources (i.e., target time slots) reserved on the outbound ports of all nodes in the target path can be reported to the initial node H during the reservation process, so that the initial node H can generate a list containing all the outbound time slots above, and the list of time slots will be logged in a status of the maintained reservation subtask.

[103] In step 1032c, the estimated total residence delay value of the target service flow on the candidate path is determined based on the estimated residence delay value corresponding to each candidate node on the candidate path.

[104] The sum of the estimated residence delay values ​​corresponding to the various candidate nodes in the candidate path can be determined as the total estimated residence delay value in the candidate path. Petition 870250094503, dated 10 / 16 / 2025, page 43 / 87 31 / 40

[105] Under normal circumstances, for a path_HE carrying a deterministic service flow, the service flow needs to be routed from the end-node E output port to a service side, which also faces the queuing and waiting problem. However, this service-side-facing output port does not belong to the path_HE portion. That is, for path_HE, the time slot resource cannot be reserved on the output port. Therefore, if the service-side-facing output port needs to additionally support time slot-based packet scheduling, the time slot resource must be reserved on a higher-level service path (overlay path) using a method similar to the method provided in the modalities of this application. At this point, path_HE will serve as a virtual link to the overlay path to provide deterministic delay indicators.Therefore, for path_HE, the residence delay at the end node E is contributed only by the intra-node forwarding delay, but does not include a queuing delay facing the service side. That is, reserving the target time slot resource for the target service flow according to the time slot resource information of the candidate node's egress port and the candidate node's residence delay budget value, and determining the evaluated residence delay value of the candidate node includes that: in a case where the candidate node is an end node, the intra-node forwarding delay corresponding to the end node is determined as the evaluated residence delay value of the end node without reserving a time slot resource for the target service flow on an egress port of the end node.

[106] In some possible implementations, in S104, the sum of the evaluated value of total residence delay and accumulated link propagation delay of the selected candidate path that satisfies the amount of delay demand includes that: the sum of the evaluated value of total residence delay of the path Petition 870250094503, dated 10 / 16 / 2025, page 44 / 87 32 / 40 selected candidate and accumulated link propagation delay of the selected candidate path is equal to the amount of delay demand; or a difference between the amount of delay demand and the sum of the estimated value of total residence delay of the selected candidate path and accumulated link propagation delay is less than a second threshold.

[107] Through the steps above, the assessed value of total residence delay of the target service flow on the candidate path is determined, and the sum of the assessed value of total residence delay and accumulated link propagation delay is compared with the amount of delay demand. When the sum is equal to or close to the amount of delay demand of the target service flow, and the time slot resource is successfully reserved for the target service flow on the egress ports of all candidate nodes, the candidate path meets the delay demand of the target service flow and is determined as the target path of the target service flow.

[108] In practical applications, firstly, when the initial node H receives a data packet from a specific service flow, the state of the reservation subtask maintained by the service flow is queried according to characteristic words and arrival time of the service flow, the packet is matched to the corresponding reservation subtask of the corresponding service flow, and the output time slot saved in the reservation subtask state is encapsulated for the packet (according to local strategies, it can choose to encapsulate a single reserved output time slot on the output port of node H, or choose to encapsulate a list of output time slots corresponding to the entire path).

[109] Secondly, when intermediate node I receives a data packet encapsulated with time slot information, an entry in the upstream node's output time slot mapping table is maintained mapped to the output time slot of Petition 870250094503, dated 10 / 16 / 2025, page 45 / 87 33 / 40 The current node is queried and compared for packets encapsulated with a single time slot. The mapped output time slot is obtained from the entry in the table and updated to the time slot encapsulated in the packet. Then, the corresponding time slot resources of the output port are accessed and the packet is forwarded to the downstream node. For packets encapsulated with a time slot list, the next output time slot is obtained directly from the list, then the corresponding time slot resources of the output port are accessed, and the packet is forwarded to the downstream nodes.

[110] Together with the network structure in Figure 6, the steps above are explained below.

[111] Assuming that all nodes in the network are the same in the time slot interval, i.e., 10 nodes, and that the orchestration period is equivalent to a scheduling period, each orchestration period containing 100 time slots. Furthermore, it is assumed that the fixed forwarding delay within each node is 10 us.

[112] It is assumed that the arrival time of the burst volume is mapped to the ongoing sending time slot 11 of the initial node S, and that t_11 = 6 us remains before the end of time slot 11. Assuming that the rated average queuing delay value is used to determine the rated residence delay value corresponding to the initial node S, the rated residence delay value is compared with the intra-node residence delay budget adjustment value (30+0=30 us), an egress time slot 13 will be reserved at the egress port of node S and therefore the offset O of the time slot is equal to 13-11=2, and the time slot interval of the initial node is 10. According to the calculation formula above for the rated average queuing delay value, the rated average queuing delay value is obtained by Tave = (0+(2*2-1)*10) / 2=15. In addition, the sum of intra-node forwarding delay, of the time Petition 870250094503, dated 10 / 16 / 2025, page 46 / 87 The remaining 34 / 40 and the average assessed queuing delay value, all corresponding to the initial node, is determined as the assessed residence delay value corresponding to the initial node, which is 10 + 6 + 15 = 31 us. At this point, the accumulated residence delay deviation is 30 - 31 = -1 us.

[113] Assuming that the end of the outgoing time slot 13 of node S is mapped to the ongoing sending time slot 70 of node A, and that t_70 = 3 us remains before the end of time slot 70. Assuming that the average rated value of queuing delay is used to determine the rated value of residence delay of node A, the rated value of residence delay is compared with the intra-node residence delay budget adjustment value (30-1 = 29 us), an outgoing time slot 71 will be reserved on the outgoing port of node A, and the obtained rated value of residence delay of node A is 10 + 3 + 10 = 23 us. At this point, the accumulated residence delay deviation is 29-23 = 6 us.

[114] Assuming that the end of the outgoing time slot 71 of node A is mapped to the ongoing sending time slot 40 of node C, and that t_40 = 5 us remains before the end of time slot 40. Assuming that the average rated value of queuing delay is used to determine the rated value of residence delay of node C, the rated value of residence delay is compared with the intra-node residence delay budget adjustment value (30 + 6 = 36 us), an outgoing time slot 42 will be reserved on the outgoing port of node C, and the obtained rated value of residence delay of node C is 10 + 5 + 20 = 35 us. At this point, the accumulated residence delay deviation is 36-35 = 1 us.

[115] It is assumed that the end of the output time slot 42 of node C is mapped to the ongoing send time slot 90 of node E, and that there is t_90 = 1 us remaining before the end of time slot 90. Assuming the average value evaluated Petition 870250094503, dated 10 / 16 / 2025, page 47 / 87 35 / 40 of the queuing delay is used to determine the assessed residence delay value of node E. The assessed residence delay value is compared with the intra-node residence delay budget adjustment value (30 + 1 = 31 us). An egress time slot of 92 will be reserved on the egress port of node E, and the resulting assessed residence delay value of node E is 10 + 1 + 20 = 31 us. At this point, the accumulated residence delay deviation is 31 - 31 = 0.

[116] At the final node D, the assessed residence delay value is the intra-node forwarding delay of 10 us, and is compared with the intra-node residence delay budget adjustment value (30 + 0 = 30 us) to obtain the cumulative residence delay deviation of 30 - 10 = 20 us. That is, after allocating the time slot resource for the reservation subtask, the cumulative assessed residence delay value is 20 us less than the total residence delay budget. Assuming the difference does not exceed the predefined threshold, path_SD will take effect as the target path for reservation subtask 1.

[117] After the target path comes into effect, a list of time slots <13, 71, 42, 92> can be maintained for reservation subtask 1 on the initial node S. On each node, a mapping relationship table entry from the upstream node's output time slot to the current node's output time slot can be created. For example, 13->71 is created on node A, 71->42 is created on node C, and 42->92 is created on node E. When the initial node S receives a service flow that matches reservation subtask 1, according to the local strategy, it can choose to encapsulate the above list of time slots <13, 71, 42, 92> for the packet or choose to encapsulate the single output time slot 13.

[118] By means of this method, the possibility of network congestion can be reduced and a quantitative guideline can be provided to determine the target path. Petition 870250094503, dated 10 / 16 / 2025, page 48 / 87 36 / 40 meets specific delay requirements in the network, which is beneficial for improving the transfer rate of the entire communication network.

[119] Figure 9 shows a schematic structural diagram of the hardware of an electronic device provided by an embodiment of the present application. Referring to the figure, at the hardware level, the electronic device includes a processor, optionally an internal bus, a network interface, and memory. The memory may include internal memory, such as high-speed random access memory (RAM), or non-volatile memory, such as at least disk memory. Certainly, the computer device may also include other hardware necessary for the services.

[120] The processor, network interface, and memory can be interconnected by the internal bus, which can be an industry standard architecture bus (ISA), a peripheral component interconnect bus (PCI), or an extended industry standard architecture bus (EISA), among others. The bus can be divided into an address bus, a data bus, a control bus, and the like. For ease of illustration, only a bidirectional arrow is shown in the figure, but it does not indicate just one bus or one type of bus.

[121] Memory is configured to store a program. Specifically, the program may include program code, and the program code includes computer operating instructions. Memory may include internal memory and non-volatile memory, and it provides instructions and data to the processor.

[122] The processor reads a corresponding computer program from non-volatile memory into internal memory and then executes the computer program to form a device to locate a target user at a logical level. The processor executes the program stored in memory, Petition 870250094503, dated 10 / 16 / 2025, p. 49 / 87 Specifically, 37 / 40 performs the method disclosed in the embodiments shown in FIGS. 1 to 8 and achieves the beneficial functions and effects of several methods described in the previous method embodiments, which will not be repeated here.

[123] The method described above in the embodiments shown in FIGS. 1 to 8 of the present application may be applied to or implemented by the processor. The processor may be an integrated circuit chip with signal processing capability. During an implementation process, all steps of the above method may be completed by a hardware integrated logic circuit or an instruction in the form of software in the processor. The above processor may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), and the like. The processor may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, and discrete hardware components.The various methods, steps, and logical block diagrams described in the embodiments of this application can be implemented or realized. The general-purpose processor can be a microprocessor, or the processor can also be any conventional processor, etc. The steps of the method disclosed in conjunction with the embodiments of this application can be directly incorporated to be executed and completed by a hardware decoding processor, or they can be executed and completed through a combination of hardware and software modules in the decoding processor. The software module can be located in a mature storage medium in the state of the art, such as random access memory, flash memory, read-only memory, read-only programmable memory, or erasable programmable memory. Petition 870250094503, dated 10 / 16 / 2025, page 50 / 87 38 / 40 electrically, and a register. The storage medium is located in memory. The processor reads the information from memory and completes the steps of the above method in combination with its hardware.

[124] The computational device can also perform the methods described in the previous method embodiments, and achieves the beneficial functions and effects of the methods described in the previous method embodiments, and will not be repeated here.

[125] In addition to software implementations, the electronic device of the present application does not exclude other implementations, such as logical devices or combinations of software and hardware, i.e., a body of execution of the following process flows is not limited to each logical unit, but may be hardware or logical devices.

[126] One embodiment of the present application further proposes a computer-readable storage medium. The computer-readable storage medium stores one or more programs. The one or more programs, when executed by an electronic device including a plurality of applications, cause the electronic device to perform the methods disclosed in the embodiments shown in FIGS. 1 to 8 and achieve the functions and beneficial effects of the methods described in the previous method embodiments, which will not be repeated herein.

[127] Computer-readable storage media include: read-only memory (ROM), random-access memory (RAM), a magnetic disk or an optical disk, etc.

[128] In addition, an embodiment of the present application further provides a computer program product. The computer program product includes a computer program stored on a non-transient, computer-readable storage medium, the computer program includes program instructions, and the program instructions, when executed by a Petition 870250094503, dated 10 / 16 / 2025, page 51 / 87 39 / 40 computer, implement the following flow: the methods disclosed in the embodiments shown in FIGS. 1 to 8, and achieve the functions and beneficial effects of the methods described in the previous method embodiments, which will not be repeated here.

[129] In total, the above arrangements are merely preferred arrangements of this application and are not intended to limit the scope of protection of this application. Any modification, equivalent replacement, improvement and the like made within the spirit and principle of this application shall all be contained within the scope of protection of this application.

[130] The system, apparatus, module or unit illustrated in the above embodiments may be specifically embodied by a computer chip or entity, or by a product with a particular function. A typical implementing device is a computer. In particular, the computer may be, for example, a personal computer, a laptop, a mobile phone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet, a wearable device, or a combination of any of these devices.

[131] Computer-readable media includes permanent and non-permanent, removable and non-removable media, and can store information by any method or technology. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of storage media for a computer include, but are not limited to, phase-change memory (PRAM), static random-access memory (SRAM), dynamic random-access memory (DRAM), other types of random-access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, and a disk. Petition 870250094503, dated 10 / 16 / 2025, page 52 / 87 A 40 / 40 compact read-only disc (CD-ROM), a digital versatile disc (DVD) or other optical storage, a magnetic tape cartridge, a magnetic disk storage device or other magnetic storage devices, or any other non-transmissible medium that can store information accessible by the computing device. According to the definition presented here, computer-readable media does not include transient media such as signals and modulated data carriers.

[132] It should be further noted that the terms include, comprise or any other variants thereof are intended to encompass non-exclusive inclusion, so that a process, method, article or apparatus that includes a number of elements not only includes those elements, but also other elements not clearly listed, or also includes elements inherent to that process, method, article or apparatus. Without further restrictions, the elements defined by the phrase including a... do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[133] The various modalities in this descriptive report are described progressively, and the same or similar parts among the various modalities may be referenced to one another. Each modality focuses on different content than the other modalities. In particular, as for the system modalities, since they are basically similar to the method modality, the description is relatively simple. As for related parts, see part of the description in the method modalities. Petition 870250094503, dated 10 / 16 / 2025, p. 53 / 87

Claims

1 / 6 CLAIMS 1. A method for determining a service flow path, characterized in that it comprises: acquiring a quantity of delay demand from a target service flow and time slot resource information from each egress port of each node; determining, based on the time slot resource information from each egress port of each node, at least one candidate path corresponding to the target service flow; reserving a time slot resource for the target service flow at the egress ports of the candidate nodes on each candidate path, and determining an estimated value of total residence delay of the target service flow on the candidate path;Select a candidate path that satisfies a first condition as the target path for the target service flow, wherein the first condition comprises that the time slot resource is successfully reserved for the target service flow on the egress ports of all candidate nodes of a selected candidate path, and the sum of an evaluated value of total residence delay of the selected candidate path and an accumulated link propagation delay of the selected candidate path satisfies the amount of delay demand.

2. Method, according to claim 1, characterized in that reserving the time slot resource for the target service flow on the egress ports of the candidate nodes in each candidate path comprises: determining an orchestration period for the candidate nodes in each candidate path; determining a time slot reservation subtask based on a number of time slots reserved for the target service flow in the orchestration period of the candidate nodes, wherein traffic corresponding to each time slot reservation subtask belongs to the target service flow; and reserving, for each time slot reservation subtask, the time slot resource for the target service flow on the egress ports of the candidate nodes.

3. A method according to claim 2, characterized in that determining the orchestration period for candidate nodes in each candidate path comprises: acquiring predefined service burst intervals corresponding to a plurality of service flows; and determining, based on a least common multiple of the service burst intervals corresponding to the plurality of service flows, the orchestration period for candidate nodes in each candidate path.

4. A method according to claim 2, characterized in that determining the time slot reservation subtask based on the number of timelot intervals reserved for the target service flow in the orchestration period of the candidate nodes comprises: determining the number of time slot intervals reserved for the target service flow in the orchestration period based on a first number of service burst intervals of the target service flow in the orchestration period and a second number of discrete burst volumes within each service burst interval; and determining the time slot reservation subtask based on the number of timelot intervals reserved for the target service flow in the orchestration period.

5. Method, according to claim 4, characterized in that determining the number of time slot intervals reserved for the target service flow in the orchestration period based on the first number of service burst intervals of the target service flow in the orchestration period and the second number of discrete burst volumes within each service burst interval comprises: Petition 870250094503, dated 10 / 16 / 2025, p.73 / 87 3 / 6 determine, based on the time slot resource information of the output ports of all candidate nodes in the candidate path, a third number of time slots reserved for each discrete burst volume within the service burst interval in the orchestration period; and determine the number of time slots reserved for the target service flow in the orchestration period based on the first number of service burst intervals of the target service flow in the orchestration period, the second number of discrete burst volumes within each service burst interval, and the third number of time slots reserved for each discrete burst volume in the orchestration period.

6. A method according to claim 5, characterized in that determining, based on the time slot resource information of the output ports of all candidate nodes in the candidate path, the third number of time slots reserved for each discrete burst volume within the service burst interval in the orchestration period comprises: determining, based on the time slot resource information of the output ports of all candidate nodes in the candidate path, the initial resource information corresponding to a single time slot of the candidate node in the orchestration period; and in a case where a ratio of a discrete burst volume within the service burst interval to the initial resource information corresponding to the single time slot in the orchestration period is not greater than a predefined value, determining that the discrete burst volume corresponds to a reserved time slot;or in a case where a proportion of a discrete burst volume within the service burst interval relative to the initial resource information corresponding to the single time slot in the orchestration period is greater than the predefined value, divide the discrete burst volume into a plurality of burst subvolumes according to a predefined rule, each burst subvolume corresponding to a reserved time slot, wherein, according to the predefined rule, each burst subvolume comprises at least one complete packet, and a proportion of each burst subvolume relative to the initial resource information corresponding to the single time slot in the orchestration period is not greater than the predefined value.

7. A method according to claim 2, characterized in that reserving the time slot resource for the target service flow at the egress ports of the candidate nodes in each candidate path, and determining the estimated total residence delay value of the target service flow in the candidate path comprises: determining, based on the amount of delay demand and an accumulated link propagation delay of the candidate path, a residence delay budget value for each candidate node in the candidate path; reserving, for each candidate node of the candidate path, a target time slot resource for the target service flow, according to the time slot resource information of an egress port of the candidate node and an estimated residence delay value of the candidate node, and determining an estimated residence delay value of the candidate node;and determine the estimated total residence delay value of the target service flow on the candidate path based on the estimated residence delay values ​​of all candidate nodes on the candidate path.

8. Method, according to claim 7, characterized in that reserving the target time slot for the target service flow according to the slot resource information of Petition 870250094503, dated 10 / 16 / 2025, p. 75 / 87 5 / 6, the candidate node's egress port time and the candidate node's residence delay budget value, and determining the assessed residence delay value of the candidate node comprises: determining a residence delay deviation of the candidate node according to an assessed cumulative residence delay value of all upstream nodes corresponding to the candidate node and a cumulative residence delay budget value of all upstream nodes; determining a residence delay budget adjustment value of the candidate node based on the residence delay deviation of the candidate node and the residence delay budget value of the candidate node;Based on the time slot resource information of the candidate node's egress port and the candidate node's residence delay budget adjustment value, determine an offset of the target time slot reserved by the candidate node for the target service flow relative to an ongoing sending time slot on the candidate node's egress port, wherein the ongoing sending time slot is used to represent a time slot that is in a sending state at a target time on the candidate node's egress port; in a case where the candidate node is an initial node, the target time is a time when the target service flow arrives at the candidate node's egress port; or, in a case where the candidate node is an intermediate node, the target time is a time when the end of a reserved target time slot on an egress port of a target upstream node arrives at the candidate node's egress port, the target upstream node being adjacent to the candidate node;and determine the estimated residence delay value of the candidate node based on the offset.

9. Method according to claim 8, characterized in that determining the assessed residence delay value of the candidate node based on the offset comprises: Petition 870250094503, 10 / 16 / 2025, p. 76 / 87 6 / 6 acquiring a first time slot corresponding to the upstream target node and a second time slot corresponding to the candidate node, wherein the upstream target node is a node adjacent to the candidate node among the upstream nodes; determining an assessed queuing delay value at the candidate node's output port based on the offset, in the first time slot interval and the second time slot interval; determining the remaining time from the target moment to the end of the candidate node's ongoing sending time slot; and determining a sum of the assessed queuing delay value, an intra-node forwarding delay of the candidate node, and the remaining time as the assessed residence delay value of the candidate node.

10. A method according to claim 8, characterized in that after determining the assessed residence delay value of the candidate node based on the offset, the method further comprises: determining, based on the offset, a target time slot reserved by the candidate node for the target service flow during the orchestration period; and in a case where an idle time slot resource of the determined target time slot is not less than the time slot resource required by the target service flow, and a difference between the budgeted residence delay value and the assessed residence delay value that is determined according to the offset is less than a first threshold, determining that the time slot resource is successfully reserved for the target service flow at the candidate node's egress port. Petition 870250094503, October 16, 2025, pp. 77-87

Images