Network slicing method suitable for LoRaWAN system under mixed QoS application scene

Abstract

The invention discloses a network slicing method suitable for a LoRaWAN system in a mixed QoS application scene, which comprises the following steps of: constructing a channel capacity model of the LoRaWAN system, and according to the requirements of transmission success rates PDR and delay time DL of different priority applications, fusing a PDR transmission model and a DL constraint model to calculate the maximum theoretical capacity of a single channel; and carrying out channel soft isolation dynamic network slicing algorithm design to realize channel slicing division of the multi-priority application. By perfecting the system channel capacity model under the PDR and DL dual QoS requirements, the utilization and capacity of system network resources can be calculated more accurately, so that a reliable theoretical basis is provided for a subsequent allocation algorithm; and meanwhile, dynamically allocating network channel resources to different priority applications through a channel soft isolation dynamic network slicing algorithm to establish network slices, and further allocating high-SF waste capacity of a channel occupied by a high-priority application to a low-priority application.

Description

Technical Field

[0001] The present invention relates to the technical field of network slicing, and in particular to a network slicing method suitable for a LoRaWAN system in a hybrid QoS application scenario. Background Art

[0002] LoRa technology, pioneered by Semtech, uses a unique linear frequency modulation (LFM) spread spectrum modulation scheme to achieve long-range communications of several to tens of kilometers in the sub-GHz frequency band. LoRaWAN is a low-power wide-area network protocol based on LoRa technology, with its core feature being the Adaptive Data Rate (ADR) algorithm. The LoRaWAN system, built on this protocol, is the primary deployment form of LoRa communication technology in practical IoT scenarios and typically comprises network servers, gateway devices, and end-application nodes. Hybrid QoS applications occur when different end devices within the same IoT scenario require multiple levels of communication service simultaneously due to varying mission requirements.

[0003] The ADR mechanism used by the LoRaWAN system was not initially designed to address the complex hybrid QoS requirements found in LoRa communication scenarios, namely those requiring different transmission success rates (PDRs) and latency (DLs). It applies an indiscriminate parameter allocation strategy to all terminal devices, failing to reserve resources for high-QoS devices. This ADR parameter adjustment mechanism can easily trigger resource competition for critical services, especially under high load conditions. Therefore, in hybrid QoS application scenarios, to ensure the PDR and DL requirements of high-priority devices, the system can only free up network resources by limiting the access scale or communication frequency of low-priority devices, creating a conflict between hybrid QoS guarantees and throughput guarantees. Therefore, achieving hybrid QoS and throughput guarantees in the LoRaWAN system remains an urgent issue.

[0004] In recent years, academia has proposed a variety of optimization strategies to address these challenges. The first is a high-priority communication enhancement scheme based on redundancy and multi-hop mechanisms. This scheme can enhance the robustness of high-priority application communications, but it does not optimize LoRa parameters and cannot effectively guarantee system throughput. The second is a LoRa parameter scheduling scheme based on SF orthogonality. This scheme optimizes SF allocation according to PDR / DL target requirements, and has flexible deployment and good compatibility. However, LoRa's SF itself is used to adapt to different communication link qualities. If a priority isolation function is added, it will be difficult to balance them. The third is a network slicing scheduling scheme based on channel orthogonality. This scheme has good channel isolation effects, but the channel isolation is relatively rough, which will result in some loss of system throughput. Summary of the Invention

[0005] In response to the above-mentioned technical problems in the related technologies, the present invention provides a network slicing method suitable for the LoRaWAN system in a hybrid QoS application scenario, which can solve the above-mentioned problems.

[0006] To achieve the above technical objectives, the technical solution of the present invention is implemented as follows:

[0007] A network slicing method for a LoRaWAN system in a hybrid QoS application scenario includes the following steps:

[0008] S1. Construct a LoRaWAN system channel capacity model. Based on the transmission success rate (PDR) and delay time (DL) requirements of different priority applications, the maximum theoretical capacity of a single channel is calculated by integrating the PDR transmission model and the DL constraint model.

[0009] S2. Design a dynamic network slicing algorithm for soft channel isolation to implement channel slicing for multi-priority applications.

[0010] Furthermore, the channel capacity model in step S1 integrates the ALOHA access mechanism under non-independent conditions, the PDR transmission model based on signal attenuation theory, and the DL constraint model established based on the air transmission time ToA with different spreading factors SF.

[0011] Furthermore, the PDR transmission model and the DL constraint model in step S1 respectively take the thresholds of the transmission success rate PDR and the delay time DL in QoS as model inputs, and finally output the channel capacity under the QoS requirements.

[0012] Furthermore, the transmission success rate PDR in step S1 is first modeled under the ALOHA mechanism, and the Poisson distribution is used to calculate the exponential relationship between the given spreading factor SF, the number of application transmission frequencies, and the transmission success rate PDR under collision-free conditions; then the exponential relationship between the given spreading factor SF, the number of application transmission frequencies, and the transmission success rate PDR under a single collision condition of the capture effect is supplemented; finally, the two are added together to obtain the PDR transmission model under the ALOHA mechanism.

[0013] Furthermore, the transmission success rate PDR in step S1 is then modeled under the influence of signal fading, and the cumulative distribution function of the channel gain is calculated using Rayleigh fading. Then, the characteristics of the LoRa signal receiving threshold are used to obtain the PDR transmission model under the influence of signal fading.

[0014] Furthermore, the PDR transmission model under the ALOHA mechanism and the PDR transmission model under the influence of signal fading are subjected to probability fusion calculation under non-independent conditions to obtain a complete PDR transmission model.

[0015] Furthermore, the DL constraint model in step S1 limits the maximum transmission spreading factor SF of the application according to the comparison characteristics of the total transmission time and the delay time DL requirement of sending messages using different spreading factors SF in the LoRaWAN system, so as to meet the DL threshold requirement.

[0016] Furthermore, the channel soft isolation dynamic network slicing algorithm in step S2 includes dynamic allocation based on the theoretical capacity of each channel and the number of each priority application, and establishes slice isolation for applications of different priorities.

[0017] Beneficial effects of the present invention: By improving the system channel capacity model under the dual QoS requirements of PDR and DL, the present invention can more accurately calculate the utilization and capacity of the system network resources, thereby providing a reliable theoretical basis for the subsequent allocation algorithm. At the same time, through the channel soft isolation dynamic network slicing algorithm, network channel resources are dynamically allocated to different priority applications to establish network slices, and the high SF wasted capacity of the channel occupied by high-priority applications is further allocated to low-priority applications. Combined with periodic monitoring, full utilization and dynamic allocation of network resources are achieved, and finally the dual guarantee of hybrid QoS and overall throughput in the LoRaWAN system is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0019] The present invention will be described in further detail below with reference to the accompanying drawings.

[0020] Figure 1 This is a design flow chart of a network slicing method for a LoRaWAN system in a hybrid QoS application scenario, as described in an embodiment of the present invention.

[0021] Figure 2 This is a framework diagram of a channel capacity model according to an embodiment of the present invention;

[0022] Figure 3 This is a flowchart of the implementation details of the channel soft isolation dynamic network slicing algorithm design described in an embodiment of the present invention. DETAILED DESCRIPTION

[0023] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention are within the scope of protection of the present invention.

[0024] like Figure 1 As shown, according to the present invention, a network slicing method for a LoRaWAN system in a hybrid QoS application scenario is disclosed, comprising the following steps: S1, constructing a channel capacity model for the LoRaWAN system, and calculating the maximum theoretical capacity of a single channel based on the transmission success rate PDR and delay time DL requirements of applications with different priorities by integrating the PDR transmission model and the DL constraint model; S2, designing a dynamic network slicing algorithm for channel soft isolation to achieve channel slicing for multi-priority applications. The channel capacity model in step S1 integrates the ALOHA access mechanism under non-independent conditions and the PDR transmission model of the signal attenuation theory, and the DL constraint model established based on the air transmission time ToA with different spreading factors SF. The PDR transmission model and the DL constraint model in step S1 respectively use the thresholds of the transmission success rate PDR and the delay time DL in the QoS as model inputs, and finally output the channel capacity under the QoS requirements. In step S1, the transmission success rate (PDR) is first modeled under the ALOHA mechanism. The Poisson distribution is used to calculate the exponential relationship between the transmission success rate (PDR) and the number of transmission frequencies used under collision-free conditions. The exponential relationship between the transmission success rate (PDR) and the number of transmission frequencies used under a single collision condition, supplemented by the capture effect, is then calculated. Finally, the two are summed to obtain the PDR transmission model under the ALOHA mechanism. The transmission success rate (PDR) in step S1 is then modeled under signal fading. Rayleigh fading is used to calculate the cumulative distribution function of the channel gain. The characteristics of the LoRa signal reception threshold are then used to determine the PDR transmission model under signal fading. The PDR transmission model under the ALOHA mechanism and the PDR transmission model under signal fading are probabilistically fused under dependent conditions to obtain the complete PDR transmission model. The DL constraint model in step S1 limits the maximum transmission spreading factor (SF) of applications based on the comparison of the total transmission time and the DL delay time (DL) requirement for messages sent using different spreading factors in the LoRaWAN system. This limits the maximum transmission spreading factor (SF) to meet the DL threshold requirement. The channel soft isolation dynamic network slicing algorithm in step S2 includes dynamic allocation based on the theoretical capacity of each channel and the number of each priority application, and establishes slice isolation for applications of different priority levels.

[0025] In a specific embodiment of the present application, the channel capacity model framework is as follows: Figure 2As shown, k represents the Class k application node in the multi-priority application in the system. This channel capacity model is based on the physical layer propagation characteristics, comprehensively considers the influence of channel fading and access mechanism, and gradually derives the theoretical capacity n for hybrid QoS applications. k .

[0026] First, from the perspective of signal propagation, the channel capacity model combines the node transmit power (TP) and the received signal-to-noise ratio (SNR), introduces Rayleigh fading and path loss (PL) under signal attenuation conditions, and derives the PDR transmission model PDR under signal attenuation. H SFi , as shown in formula (1).

[0027]

[0028] In formula (1), P means H is less than H th The probability of, where H is the transmission channel gain under Rayleigh fading, H th is the threshold gain for successful signal reception under Rayleigh fading, H th As shown in formula (2).

[0029] H th =N·SNR SFi PL(d) / P t (2)

[0030] In formula (2), N represents the additive white Gaussian noise in signal transmission, P t That is, the node's transmit power TP. Formula (1) shows that the PDR of the LoRa signal without transmission collision is H_SFi Depends on the node transmit power TP, path loss PL(d) and signal-to-noise ratio receiving threshold SNR SFi .

[0031] Subsequently, based on the ALOHA access mechanism, combined with the node data packet sending frequency (λ), spreading factor (SF) and capture effect, the PDR transmission model PDR under the ALOHA mechanism is obtained through Poisson distribution modeling. ALOHA_SFi , as shown in formula (3), where γ = 6dB represents the capture effect threshold.

[0032]

[0033] In formula (3), SFi represents the i-th LoRa spreading factor SF (i∈{7, 8, 9, 10, 11, 12}); ToA SFi Indicates the LoRa air transmission time when using SFi communication. This formula shows the PDR of the LoRa signal under the ALOHA mechanism ALOHA_SFi Depends on the number of nodes n using SFi in the channel SFi, node packet sending frequency λ and ToA SFi .

[0034] In order to reflect the combined effects of channel characteristics and access conflicts in real networks, the two PDR transmission models are coupled non-independently through conditional probability, and finally the transmission success rate PDR of a k-category node sending a signal under SFi can be obtained. k_SFi , as shown in formula (4).

[0035]

[0036] In formula (4), PDR(X=0) is shown in formula (5), which is the probability that the application device does not encounter a collision within the double time window.

[0037]

[0038] PDR(X=1) in formula (4) is shown in formula (6), which is the probability that an application node transmitting on a certain channel and SFi encounters a collision within twice the collision window time and successfully transmits through the capture effect.

[0039]

[0040] On this basis, in order to meet the delay time (DL) requirements, according to the LoRaWAN system communication process, combined with the data packet air transmission time (ToA), the DL constraints are quantified and the corresponding relationship between them and the spreading factor (SF) is established. k_SFi Sum, and then use the step function ε(DL k -DL i ) to indicate whether the DL requirements are met, where DL k is the transmission delay threshold requirement for Classk LoRa nodes, DL i The actual transmission delay when the LoRa node uses SFi communication. If DL k >DL i Then the step function ε is 1, otherwise it is 0. Finally, the two QoS indicators of PDR and DL are integrated to derive the theoretical maximum node capacity n of Class k nodes in a single channel under hybrid QoS constraints. k , that is, the channel capacity model, as shown in formula (7).

[0041]

[0042] The model is based on the hybrid QoS of Class k nodes (transmission success rate requirement PDR k , delay requirement DL k ) is required to be input, and the output corresponds to the Class k node that meets the PDRk With DL k The theoretical capacity n in a single channel under the requirement k The subsequent channel network slicing algorithm will be based on n k and the number of Class k nodes N k Channel division and dynamic channel adjustment are performed; the LoRa physical layer parameter optimization algorithm is based on the capacity model n k Analyze the theoretical node capacity n of different SFs k_SFi , to achieve balanced allocation and dynamic optimization of SF and TP.

[0043] In a specific embodiment of the present application, Figure 3 The following is a flowchart showing the implementation details of the channel network slicing algorithm design. The specific process is as follows:

[0044] First, assume a mixed QoS scenario with three priority classes: Class 1, Class 2, and Class 3, with N1, N2, and N3, respectively. A LoRaWAN system typically has eight receive channels, so assume each class is allocated M1, M2, and M3 channels. n1, n2, and n3 are the theoretical single-channel capacities for Class 1, Class 2, and Class 3 application nodes, respectively, as determined by the channel capacity model.

[0045] S1. First, channel constraints need to be set for Class 1, Class 2, and Class 3 nodes. First, ensure that M1, M2, and M3 are not less than 1; second, limit the maximum number of channels for each priority level node to not exceed the maximum number of channels in the system.

[0046] S2. Establish overlapping channel allocation. First, Class 1 and Class 2 nodes do not share channels with each other; second, Class 3 nodes can fill idle SF resources of Class 1 or Class 2 channels based on the DL restrictions of higher-priority nodes.

[0047] S3. To adapt to environmental changes and fluctuations in communication conditions, a periodic dynamic adjustment mechanism is designed.

[0048] (1) Dynamically adjust the channel theoretical capacity n based on the RSSI, SNR and TP of the transmitted data k and the number of channel assignments M k .

[0049] (2) Statistics are applied to upload data to determine whether the transmission success rate (PDR) of high priority nodes (Class 1 and Class 2) is met.

[0050] (3) If the actual PDR of the high-priority node is lower than its threshold requirement, the access rights of the Class 3 node in the overlapping channel are restricted.

[0051] (4) If the PDR of the high-priority node is significantly higher than the threshold requirement, the communication capacity of the Class 3 node in the overlapping channel is released to improve the overall throughput.

[0052] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A network slicing method for a LoRaWAN system in a hybrid QoS application scenario, characterized in that: The steps include: S1. Construct a LoRaWAN system channel capacity model. Based on the transmission success rate (PDR) and delay time (DL) requirements of different priority applications, the maximum theoretical capacity of a single channel is calculated by integrating the PDR transmission model and the DL constraint model. S2. Design a dynamic network slicing algorithm for soft channel isolation to implement channel slicing for multi-priority applications.

2. A network slicing method for a LoRaWAN system in a hybrid QoS application scenario according to claim 1, characterized in that: The channel capacity model in step S1 integrates the ALOHA access mechanism under non-independent conditions, the PDR transmission model based on signal attenuation theory, and the DL constraint model established based on the air transmission time ToA with different spreading factors SF.

3. A network slicing method for a LoRaWAN system in a hybrid QoS application scenario according to claim 1, characterized in that: The PDR transmission model and DL constraint model in step S1 respectively take the thresholds of the transmission success rate PDR and the delay time DL in QoS as model inputs, and finally output the channel capacity under the QoS requirements.

4. A network slicing method for a LoRaWAN system in a hybrid QoS application scenario according to claim 2, characterized in that: The transmission success rate PDR in step S1 is first modeled under the ALOHA mechanism, and the Poisson distribution is used to calculate the exponential relationship between the given spreading factor SF, the number of applied transmission frequencies, and the transmission success rate PDR under collision-free conditions. Then, the exponential relationship between the given spreading factor SF, the number of applied transmission frequencies, and the transmission success rate PDR under a single collision condition with a capture effect is supplemented. Finally, the two are added together to obtain the PDR transmission model under the ALOHA mechanism.

5. A network slicing method for a LoRaWAN system in a hybrid QoS application scenario according to claim 4, characterized in that: The transmission success rate PDR in step S1 is then modeled under the influence of signal fading. The cumulative distribution function of the channel gain is calculated using Rayleigh fading, and the characteristics of the LoRa signal receiving threshold are used to obtain the PDR transmission model under the influence of signal fading.

6. A network slicing method for a LoRaWAN system in a hybrid QoS application scenario according to claim 5, characterized in that: The PDR transmission model under the ALOHA mechanism and the PDR transmission model under the influence of signal fading are subjected to probability fusion calculation under non-independent conditions to obtain a complete PDR transmission model.

7. The network slicing method for a LoRaWAN system in a hybrid QoS application scenario according to claim 1, characterized in that: The DL constraint model in step S1 limits the maximum transmission spreading factor SF of the application based on the comparison characteristics of the total transmission time and the delay time DL requirement of sending messages using different spreading factors SF in the LoRaWAN system, so as to meet the DL threshold requirement.

8. The network slicing method for a LoRaWAN system in a hybrid QoS application scenario according to claim 1, characterized in that: The channel soft isolation dynamic network slicing algorithm in step S2 includes dynamic allocation based on the theoretical capacity of each channel and the number of each priority application, and establishes slice isolation for applications of different priorities.

Images