Dynamic bandwidth allocation method for multi-user multi-service passive optical network

Abstract

The invention discloses a multi-user multi-service passive optical network dynamic bandwidth allocation method, and belongs to the field of optical access networks and industrial internet. According to the passive optical network dynamic bandwidth allocation method based on traffic classification, a differential allocation method is provided for different types of traffic of users by improving the architecture and the bandwidth allocation method of a traditional time division multiplexing passive optical network; and bandwidth resource allocation with high bandwidth utilization rate and high framework flexibility is realized in an optical access network scene oriented to the industrial internet. According to the invention, by dividing the two virtual OLT and the two slices which are logically isolated, the flexibility and the expandability are improved; historical traffic data are collected and input into a joint neural network model, accurate traffic classification is carried out, and differential resource allocation is realized; the user data traffic is divided into specific network slices through traffic classification, and different bandwidth allocation methods are adopted in different slices, so that the utilization rate of bandwidth resources of the whole network is improved.

Description

Technical Field

[0001] The invention relates to a dynamic bandwidth allocation method for a multi-user multi-service passive optical network, and belongs to the field of optical access networks and industrial Internet. Background Art

[0002] With the rapid development of industrial Internet and network function virtualization technology, various new network services including mobile fronthaul and industrial Internet of Things will deploy a large number of access points, requiring industrial passive optical networks to develop towards lower transmission latency and higher throughput.

[0003] Network slicing technology is a method of logically dividing network resources so that different applications or services can share resources on the same physical network infrastructure while ensuring isolation and performance requirements. Through network slicing technology, the entire network is divided into multiple independent virtual network slices, each with its own network function, topology, bandwidth and security policy.

[0004] Classify the network traffic generated by users and assign different types of network traffic to different network slices to achieve more targeted resource allocation. Neural networks accurately classify different types of traffic through learning and pattern matching. Long Short-Term Memory (LSTM) neural network is a variant of recurrent neural network, suitable for processing time series data, which can capture the long-term dependencies of time series and extract and capture the periodic characteristics of time series; Gated Recurrent Unit (GRU) neural network is another variant of recurrent neural network, which improves the efficiency of training and reasoning by reducing the number of gated units and simplifying the calculation process.

[0005] In response to the problems of high service coupling, insufficient bandwidth resource utilization and fixed loss in optical access networks, this patent provides a dynamic resource allocation method for multiple services and multiple users to reduce the delay and packet loss rate of optical access networks. Summary of the invention

[0006] In view of the problems of low uplink bandwidth resource utilization and inflexible architecture in optical access networks for the Industrial Internet, the purpose of the present invention is to provide a multi-user multi-service passive optical network dynamic bandwidth allocation method based on traffic classification, which can achieve bandwidth resource allocation with high bandwidth utilization and high architectural flexibility in the optical access network scenario for the Industrial Internet.

[0007] The objective of the present invention is achieved through the following technical solutions:

[0008] A multi-user multi-service passive optical network dynamic bandwidth allocation method of the present invention improves the architecture and bandwidth allocation method of the traditional time division multiplexing passive optical network, provides a differentiated allocation method for different types of user traffic, so as to improve the bandwidth resource utilization of the entire optical access network, and comprises the following steps:

[0009] Step 1: Based on the network function virtualization technology, the physical OLT of TDM-PON is logically divided into two virtual OLTs, and the ONU connected to the OLT is divided into two slices. The SDN controller is used to collect the historical traffic generated by users. During the bandwidth allocation cycle, the user historical traffic is input into the joint neural network model of LSTM and GRU. The neural network model predicts the user's future traffic, including the following sub-steps:

[0010] Step 1.1: Based on the network function virtualization technology, two virtual OLTs are constructed in the physical OLT of the traditional TDM-PON architecture: vOLT1 and vOLT2. vOLT1 is used for bandwidth allocation in periodic traffic slices, and vOLT2 is used for bandwidth allocation in non-periodic traffic slices.

[0011] Step 1.2, the ONU connected to the physical OLT of TDM-PON is divided into two slices, which are used to receive the periodic traffic and non-periodic traffic of users respectively. The user equipment is connected to the ONU through wired transmission or wireless transmission;

[0012] Step 1.3: Set up an SDN controller on the physical OLT side of TDM-PON, collect the historical traffic of the user in the previous N cycles through the SDN controller, and set up a LSTM and GRU joint neural network inside the SDN controller to predict the user's traffic in the next two cycles based on the user's historical traffic;

[0013] Step 2: Based on the traffic prediction results of step 1, the actual traffic of the user in the next period is compared to classify the user traffic into periodic traffic and non-periodic traffic. The SDN controller will classify different types of network traffic into different ONU slices based on the classification results, and finally upload them to their respective vOLTs to complete differentiated bandwidth allocation, including the following sub-steps:

[0014] Step 2.1: The traffic prediction module of the SDN controller receives the user's historical traffic, sets the parameters of the LSTM neural network, and trains the GRU model using the maximum likelihood estimation method;

[0015] Step 2.2: After the parameters of the LSTM and GRU joint prediction model in the traffic prediction module are set, the user's historical traffic is input into the model as input to obtain the user's traffic status in the next two cycles, as shown in formula (1):

[0016] Predict final =w1*RES LSTM +W2*RES GRU (1)

[0017] Among them, Predict final is the final prediction result of the joint neural network, w1 and w2 are weight coefficients, RES LSTM and RES GRU These are the prediction results of the LSTM model and the GRU model respectively;

[0018] Step 2.3: According to the comparison between the joint prediction result and the actual value of the next period, the incoming traffic is classified into different categories: the traffic that is close to the actual value and meets the expectation is classified as periodic traffic, and the traffic that does not meet the expectation is classified as non-periodic traffic;

[0019] Determine the mean absolute percentage error (MAPE) based on the predicted and actual values Periodic , and MAPE Periodic Traffic that is less than the set threshold η is classified as periodic traffic;

[0020] Step 3: According to the classification of traffic in step 2, the bandwidth of the entire network is first allocated to different ONU slices. In this cycle, the bandwidth of the periodic traffic slice is allocated according to the prediction result of the traffic prediction module, and the non-periodic traffic slice uses the remaining bandwidth of the total bandwidth; in the ONU slice of periodic traffic, the vOLT directly allocates the bandwidth to each ONU according to the prediction result of the SDN controller in a polling manner; in the ONU slice of non-periodic traffic, a fixed bandwidth is allocated to each ONU to avoid the problem that user data in the ONU cannot be transmitted for a long time. In addition, the resources other than the fixed bandwidth allocated to the non-periodic ONU slice are used as public bandwidth and are allocated to each ONU in a polling manner;

[0021] Step 3.1: Based on the traffic classification result of step 2, the SDN controller uploads the classification result to the OLT to perform inter-chip bandwidth allocation for the next cycle. The inter-chip bandwidth allocation is shown in equations (2) and (3):

[0022] B P =B total *n p (2)

[0023] B a =B total *n a (3)

[0024] Among them, B P is the allocated bandwidth of the periodic traffic slice, B ais the allocated bandwidth of the aperiodic traffic slice, B total is the total upstream bandwidth of the PON network, n p and n a is the normalization coefficient of the statistical flow, as shown in formula (4) and formula (5) respectively:

[0025]

[0026] n a =1-n p (5)

[0027] Among them, Predict p is the prediction result of periodic traffic, Predict a It is the prediction result of the fee cycle traffic;

[0028] Step 3.2: Based on the bandwidth allocation between slices in step 3.1, perform bandwidth allocation within the slice;

[0029] In the non-periodic traffic slice, each ONU is allocated guaranteed bandwidth to prevent low-priority traffic from being unable to be transmitted for a long time, resulting in a large number of packet losses and reduced network throughput. The remaining bandwidth is allocated among different ONUs in round-robin fashion according to priority. remaina As shown in formula (6):

[0030] B remaina =B a -m*B f (6)

[0031] Among them, B f is the guaranteed bandwidth allocated to each aperiodic ONU, and m is the number of aperiodic ONUs;

[0032] The allocated bandwidth B of the ONU in the i-th non-periodic cluster allocai As shown in formula (7):

[0033]

[0034] Among them, p i is the priority of the data packet;

[0035] In periodic traffic slicing, no priority is distinguished. The vOLT polls the ONUs and allocates bandwidth to each ONU according to the bandwidth request of the ONU;

[0036] It also includes step 4: guiding the analysis and design of information transmission methods in public place networks according to the bandwidth allocation method in step 3, which can improve traffic load balancing, fair resource allocation, priority management and elastic adaptation in public networks, and improve the robustness and user experience of public place networks.

[0037] Beneficial Effects

[0038] 1. A multi-user multi-service passive optical network dynamic bandwidth allocation method of the present invention is based on network function virtualization technology and network slicing technology. The physical OLT of the traditional TDM-PON architecture is divided into two logically isolated virtual OLTs, which are responsible for the resource allocation of periodic traffic and non-periodic traffic; the ONU connected to the physical OLT is divided into two logically isolated slices, which are used to receive the user's periodic traffic and non-periodic traffic respectively. The user accesses the OLT through wired or wireless means, thereby improving the flexibility and scalability of the traditional TDM-PON architecture.

[0039] 2. A multi-user multi-service passive optical network dynamic bandwidth allocation method of the present invention collects user historical traffic data and inputs the traffic data as input into the LSTM and GRU joint neural network model to accurately classify traffic and achieve differentiated resource allocation.

[0040] 3. A dynamic bandwidth allocation method for a multi-user multi-service passive optical network of the present invention divides user data traffic into specific network slices by classifying traffic. In uplink transmission, different bandwidth allocation methods are used in different slices to improve the utilization rate of the entire network bandwidth resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 for Figure 1 It is a flow chart of a multi-user multi-service passive optical network dynamic bandwidth allocation method of the present invention;

[0042] Figure 2 for Figure 2 This is a diagram of the improved TDM-PON architecture based on network slicing technology of the present invention;

[0043] Figure 3 This is a performance result diagram of using LSTM and GRU neural network models to predict periodic traffic in the embodiment;

[0044] Figure 4 It is a comparison chart of the performance results of the method of the present invention compared with the traditional method in the embodiment. DETAILED DESCRIPTION

[0045] In order to better illustrate the purpose and advantages of the present invention, the invention is further described below with reference to the accompanying drawings and examples.

[0046] Embodiment 1:

[0047] In the embodiment, the OLT of TDM-PON is connected to 32 ONUs, the length of the optical fiber from the ONU to the OLT is 20 kilometers, the round-trip delay is 150 microseconds, there are 160 user terminals in the IPON network, each ONU is connected to five user terminals, connected to different slices through a wireless or wired network, the size of the random data packet generated by each user terminal is between 64 bytes and 1518 bytes, the generated non-periodic traffic follows the generalized Pareto distribution and has a priority label, the periodic traffic of IIoT is generated by users with fixed periods and sizes, and the dynamic bandwidth allocation period of the network is 2 milliseconds;

[0048] In the above scenario, a multi-user multi-service passive optical network dynamic bandwidth allocation method of the present invention is applied to provide a differentiated allocation method for different types of traffic of users, thereby improving the bandwidth resource utilization of the entire optical access network, including the following steps:

[0049] Step 1: Based on the network function virtualization technology, the physical OLT of TDM-PON is logically divided into two virtual OLTs, and the ONUs connected to the OLT are divided into two slices, each with 16 ONUs. The SDN controller is used to collect the historical traffic generated by users. During the bandwidth allocation cycle, the user historical traffic is input into the joint neural network model of LSTM and GRU. The neural network model predicts the user's future traffic, including the following sub-steps:

[0050] Step 1.1: Based on the network function virtualization technology, two virtual OLTs are constructed in the physical OLT of the traditional TDM-PON architecture: vOLT1 and vOLT2. vOLT1 is used for bandwidth allocation in periodic traffic slices, and vOLT2 is used for bandwidth allocation in non-periodic traffic slices.

[0051] Step 1.2, the ONU connected to the physical OLT of TDM-PON is divided into two slices, which are used to receive the periodic traffic and non-periodic traffic of users respectively. The user equipment is connected to the ONU through wired transmission or wireless transmission;

[0052] Step 1.3: Set up an SDN controller on the physical OLT side of TDM-PON. The improved TDM-PON network architecture is as follows: Figure 2 As shown in the figure, the SDN controller collects the historical traffic of the user in the previous N cycles. The SDN controller is equipped with a LSTM and GRU joint neural network to predict the user's traffic in the next two cycles based on the user's historical traffic.

[0053] Step 2: Based on the traffic prediction results of step 1, the actual traffic of the user in the next period is compared to classify the user traffic into periodic traffic and non-periodic traffic. The SDN controller will classify different types of network traffic into different ONU slices based on the classification results, and finally upload them to their respective vOLTs to complete differentiated bandwidth allocation, including the following sub-steps:

[0054] Step 2.1, the traffic prediction module of the SDN controller receives the user's historical traffic and sets the parameters of the LSTM neural network. The batch size of the LSTM neural network is 1000, the learning rate is 0.01, the training cycle is 100, and the GRU model is trained using the maximum likelihood estimation method. The seasonal cycle is 1 millisecond and the lag order is 1000.

[0055] Step 2.2: After the parameters of the LSTM and GRU joint prediction model in the traffic prediction module are set, the user's historical traffic is input into the model as input to obtain the user's traffic status in the next two cycles, as shown in formula (1):

[0056] Predict final =w1*RES LSTM +W2*RES GRU (1)

[0057] Among them, Predict final is the final prediction result of the joint neural network, w1 and w2 are weight coefficients, RES LSTM and RES GRU These are the prediction results of the LSTM model and the GRU model respectively;

[0058] Step 2.3: According to the comparison between the joint prediction result and the actual value of the next period, the incoming traffic is classified into different categories: the traffic that is close to the actual value and meets the expectation is classified as periodic traffic, and the traffic that does not meet the expectation is classified as non-periodic traffic;

[0059] Determine the mean absolute percentage error (MAPE) based on the predicted and actual values Periodic , and MAPE Periodic Traffic that is less than the set threshold η is classified as periodic traffic;

[0060] In the embodiment, the result of using LSTM and GRU combined neural network to predict periodic traffic is as follows: Figure 3 As shown, the MAPE result of periodic traffic is 0.08%, which can accurately classify traffic categories;

[0061] Step 3: According to the classification of traffic in step 2, the bandwidth of the entire network is first allocated to different ONU slices. In this cycle, the bandwidth of the periodic traffic slice is allocated according to the prediction result of the traffic prediction module, and the non-periodic traffic slice uses the remaining bandwidth of the total bandwidth; in the ONU slice of periodic traffic, the vOLT directly allocates the bandwidth to each ONU according to the prediction result of the SDN controller in a polling manner; in the ONU slice of non-periodic traffic, a fixed bandwidth is allocated to each ONU to avoid the problem that user data in the ONU cannot be transmitted for a long time. In addition, the resources other than the fixed bandwidth allocated to the non-periodic ONU slice are used as public bandwidth and are allocated to each ONU in a polling manner;

[0062] Step 3.1: Based on the traffic classification result of step 2, the SDN controller uploads the classification result to the OLT to perform inter-chip bandwidth allocation for the next cycle. The inter-chip bandwidth allocation is shown in equations (2) and (3):

[0063] B P =B total *n p (2)

[0064] B a =B total *n a (3)

[0065] Among them, B P is the allocated bandwidth of the periodic traffic slice, B a is the allocated bandwidth of the aperiodic traffic slice, B total is the total upstream bandwidth of the PON network, n p and n a is the normalization coefficient of the statistical flow, as shown in formula (4) and formula (5) respectively:

[0066]

[0067] n a =1-n p (5)

[0068] Among them, Predict p is the prediction result of periodic traffic, Predict a It is the prediction result of the fee cycle traffic;

[0069] Step 3.2: Based on the bandwidth allocation between slices in step 3.1, perform bandwidth allocation within the slice;

[0070] In the non-periodic traffic slice, each ONU is allocated guaranteed bandwidth to prevent low-priority traffic from being unable to be transmitted for a long time, resulting in a large number of packet losses and reduced network throughput. The remaining bandwidth is allocated among different ONUs in round-robin fashion according to priority. remaina As shown in formula (6):

[0071] B remaina =B a -m*B f (6)

[0072] Among them, B f is the guaranteed bandwidth allocated to each aperiodic ONU, and m is the number of aperiodic ONUs;

[0073] The allocated bandwidth B of the ONU in the i-th non-periodic cluster allocai As shown in formula (7):

[0074]

[0075] Among them, p i is the priority of the data packet;

[0076] In periodic traffic slicing, no priority is distinguished. The vOLT polls the ONUs and allocates bandwidth to each ONU according to the bandwidth request of the ONU;

[0077] In the embodiment, the method of the present invention is simulated in the improved TDM-PON architecture described in step 1, and the simulation results are as follows: Figure 4 As shown, under the condition of fixed load, the method of the present invention is superior to the traditional IPACT algorithm in terms of network delay;

[0078] It also includes step 4: guiding the analysis and design of information transmission methods in public place networks according to the bandwidth allocation method in step 3, which can improve traffic load balancing, fair resource allocation, priority management and elastic adaptation in public networks, and improve the robustness and user experience of public place networks.

[0079] The specific description above further illustrates the purpose, technical solutions and beneficial effects of the invention in detail. It should be understood that the above is only a specific embodiment of the present invention and is not intended to limit the scope of protection of 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 method for dynamic bandwidth allocation in a multi-user multi-service passive optical network, characterized in that: By improving the architecture and bandwidth allocation method of the traditional time-division multiplexing passive optical network, a differentiated allocation method is provided for different types of user traffic to improve the bandwidth resource utilization of the entire optical access network, including the following steps: Step 1: Based on the network function virtualization technology, the physical OLT of TDM-PON is logically divided into two virtual OLTs, and the ONU connected to the OLT is divided into two slices. The SDN controller is used to collect the historical traffic generated by users. During the bandwidth allocation cycle, the user historical traffic is input into the joint neural network model of LSTM and GRU. The neural network model predicts the user's future traffic; Step 2: Based on the traffic prediction results of step 1, the actual traffic of the user in the next period is compared to classify the user traffic into periodic traffic and non-periodic traffic. The SDN controller will classify different types of network traffic into different ONU slices based on the classification results, and finally upload them to their respective vOLTs to complete differentiated bandwidth allocation; Step 3: According to the classification of traffic in step 2, the bandwidth of the entire network is first allocated to different ONU slices. In this cycle, the bandwidth of the periodic traffic slice is allocated according to the prediction result of the traffic prediction module, and the non-periodic traffic slice uses the remaining bandwidth of the total bandwidth; in the ONU slice of periodic traffic, the vOLT directly allocates the bandwidth to each ONU according to the prediction result of the SDN controller in a polling manner; in the ONU slice of non-periodic traffic, a fixed bandwidth is allocated to each ONU to avoid the problem that user data in the ONU cannot be transmitted for a long time. In addition, the resources other than the fixed bandwidth allocated to the non-periodic ONU slice are used as public bandwidth and are allocated to each ONU in a polling manner.

2. A multi-user multi-service passive optical network dynamic bandwidth allocation method as claimed in claim 1, characterized in that: It also includes step 4: guiding the analysis and design of information transmission methods in public place networks according to the bandwidth allocation method in step 3, which can improve traffic load balancing, fair resource allocation, priority management and elastic adaptation in public networks, and improve the robustness and user experience of public place networks.

3. A multi-user multi-service passive optical network dynamic bandwidth allocation method as claimed in claim 1, characterized in that: Step 1 includes the following sub-steps: Step 1.1: Based on the network function virtualization technology, two virtual OLTs are constructed in the physical OLT of the traditional TDM-PON architecture: vOLT1 and vOLT2. vOLT1 is used for bandwidth allocation in periodic traffic slices, and vOLT2 is used for bandwidth allocation in non-periodic traffic slices. Step 1.2, the ONU connected to the physical OLT of TDM-PON is divided into two slices, which are used to receive the periodic traffic and non-periodic traffic of users respectively. The user equipment is connected to the ONU through wired transmission or wireless transmission; Step 1.3: Set up an SDN controller on the physical OLT side of TDM-PON, and use the SDN controller to collect the user's historical traffic in the previous N cycles. The SDN controller is equipped with a LSTM and GRU joint neural network to predict the user's traffic in the next two cycles based on the user's historical traffic.

4. A multi-user multi-service passive optical network dynamic bandwidth allocation method as claimed in claim 1, characterized in that: Step 2 includes the following sub-steps: Step 2.1: The traffic prediction module of the SDN controller receives the user's historical traffic, sets the parameters of the LSTM neural network, and trains the GRU model using the maximum likelihood estimation method; Step 2.2: After the parameters of the LSTM and GRU joint prediction model in the traffic prediction module are set, the user's historical traffic is input into the model as input to obtain the user's traffic status in the next two cycles, as shown in formula (1): Predict final =w1*RES LSTM +W2*RES GRU (1) Among them, Predict final is the final prediction result of the joint neural network, w1 and w2 are weight coefficients, RES LSTM and RES GRU These are the prediction results of the LSTM model and the GRU model respectively; Step 2.3: According to the comparison between the joint prediction result and the actual value of the next period, the incoming traffic is classified into different categories: the traffic that is close to the actual value and meets the expectation is classified as periodic traffic, and the traffic that does not meet the expectation is classified as non-periodic traffic; Determine the mean absolute percentage error (MAPE) based on the predicted and actual values Periodic , and MAPE Periodic Traffic flows that are less than the set threshold η are classified as periodic traffic flows.

5. A multi-user multi-service passive optical network dynamic bandwidth allocation method as claimed in claim 1, characterized in that: Step 3 includes the following sub-steps: Step 3.1: Based on the traffic classification result of step 2, the SDN controller uploads the classification result to the OLT to perform inter-chip bandwidth allocation for the next cycle. The inter-chip bandwidth allocation is shown in equations (2) and (3): B P =B total *n p (2) B a =B total *n a (3) Among them, B P is the allocated bandwidth of the periodic traffic slice, Ba is the allocated bandwidth of the aperiodic traffic slice, and B total is the total upstream bandwidth of the PON network, n p and n a is the normalization coefficient of the statistical flow, as shown in formula (4) and formula (5) respectively: n a =1-n p (5) Among them, Predict p is the prediction result of periodic traffic, Predict a It is the prediction result of the fee cycle traffic; Step 3.2: Based on the bandwidth allocation between slices in step 3.1, perform bandwidth allocation within the slice; In the non-periodic traffic slice, each ONU is allocated guaranteed bandwidth to prevent low-priority traffic from being unable to be transmitted for a long time, resulting in a large number of packet losses and reduced network throughput. The remaining bandwidth is allocated among different ONUs in round-robin fashion according to priority. remaina As shown in formula (6): B remaina =B a -m*B f (6) Among them, B f is the guaranteed bandwidth allocated to each aperiodic ONU, and m is the number of aperiodic ONUs; The allocated bandwidth B of the ONU in the i-th non-periodic cluster allocai As shown in formula (7): Among them, p i is the priority of the data packet; In periodic traffic slicing, no priority is distinguished. The vOLT polls the ONUs and allocates bandwidth to each ONU according to the bandwidth request of the ONU.

Images