CN100334837C - A method for assigning path bandwidth in bearing control layer - Google Patents

Abstract

The present invention discloses a method of assigning path bandwidth in a bearing control layer. The method comprises that: a: after each hop load network resource manager receives a request of resource connecting application, paths are selected; bandwidth value is obtained according to the stack level MTD of a label with a maximum path in a bearing net, the length of the total message heading of user data, the bandwidth which is requested by a user and is in the request of resource connecting application and the message length MPPL of the peak value of users' services; the bandwidth value is assigned to each selected hop path; b: after the connection of all paths is established, each hop bearing net resource manager uses the bandwidth value obtained by the MPPL for replacing the bandwidth assigned by prior each hop path according to the stack level RTD of a relative path label of each hop path, the length of the total message heading of user data, requested bandwidth of users and bandwidth value obtained by the MPPL. The method of the present invention is adopted for assigning bandwidth in a simple mode; therefore, unnecessary wasting of resources is reduced, and the cost is reduced.

Description

A Method for Allocating Path Bandwidth in Bearer Control Layer

technical field

The invention relates to a differentiated service model (Diff-serv, Differentiated Service) technology with an independent bearer layer, in particular to a method for allocating path bandwidth in the bearer control layer of the differentiated service model with an independent bearer layer.

Background technique

With the increasing scale of the Internet (Internet), various Quality of Service (QoS, Quality of Service) technologies have emerged as the times require. Therefore, the Internet Engineering Task Force (IETF, Internet Engineering Task Force) has suggested many service models and mechanisms to meet the requirements of QoS. At present, the industry generally recognizes that the integrated service model (Int-Serv, Integrated Service) is used at the access and edge of the network, and the differentiated service model (Diff-serv, Differentiated Service) is used at the core of the network. The DiffServ model only sets priority levels to ensure QoS measures. Although this QoS measure has the characteristics of high line utilization, the specific effect is difficult to predict. Therefore, in order to further improve the QoS technology, the industry has begun to introduce an independent bearer control layer for the DiffServ model of the backbone network, and establish a set of QoS signaling mechanisms dedicated to the DiffServ model. This DiffServ model is called the DiffServ model with independent bearer control layer.

Figure 1 is a diagram of a differentiated service model with an independent bearer control layer. As shown in FIG. 1 , in this model, the bearer control layer 102 is placed between the bearer network 103 and the service control layer 101 . The call agent (CA, Call Agent) in the service control layer 101 is a service server, such as softswitch, video on demand (VOD) control server, route gatekeeper (GK, Gate Keeper) etc., CA receives the call request of user equipment, Proxy user equipment to complete the call request and exchange; in the bearer control layer 102, the bearer network resource manager configures management rules and network topology, and allocates resources for the customer's service bandwidth application. Only three bearer network resources are drawn in this figure Managers, that is, bearer network resource manager 1, bearer network resource manager 2, and bearer network resource manager 3, but the number of bearer network resource managers is not fixed, and each bearer network resource manager communicates with each other through signaling Transfer the customer's service bandwidth application request and result, as well as the routing path information allocated for the service application; in the bearer network 103, each bearer network resource manager manages a specific bearer network area, and this specific bearer network area is called is the management domain of the corresponding bearer network resource manager, and in this figure is the management domain 107 of the bearer network resource manager 1, the management domain 108 of the bearer network resource manager 2, and the management domain 109 of the bearer network resource manager 3, The management domain 107 includes an edge router (ER, EdgeRouter) 110, a core router 111, and a border router (BR, Border Router) 112, wherein the ER can connect the call service flow of the user equipment to the bearer network or export the bearer network, manage Domain 108 and management domain 109 also include core routers and border routers.

In the differentiated service model with an independent bearer control layer, the bearer network resource manager applies for a communication path for the user's service connection and allocates bandwidth for the applied path. There is a method of allocating bandwidth in many differentiated service models with independent bearer control layers, such as the bandwidth proxy model of the service backbone experimental network (Qbone, Quality-of-Service backbone). Figure 2 shows the bandwidth proxy model of Qbone As shown in Figure 2, bandwidth agent 1, bandwidth agent 2, and bandwidth agent 3 realize the function of bearer network resource manager. In this model, bandwidth agent is responsible for Or the bandwidth application request of network maintenance personnel, according to the bandwidth request and a large number of information parameters recorded in the bandwidth manager, use the statistical algorithm of traffic engineering to obtain the allocated bandwidth. These information parameters include various configuration information, topology information of the physical network, and routers. A large amount of static or dynamic information such as configuration information and policy information, current resource reservation information, network occupation status information, etc.

The disadvantages of the above-mentioned Qbone bandwidth agent bandwidth allocation scheme are: a large number of parameters are required for calculation, and the calculation program is complicated, the calculation workload is very large, and the consumption of processor and other equipment resources is also large, resulting in relatively high cost.

In addition, there is also a Rich QoS solution proposed by NEC Corporation. Fig. 3 is the model figure of Rich QoS scheme, as shown in Fig. 3, QoS server 301 also comprises policy server 302 and directory server 303 and network management monitoring server 304 supporting with QoS server as key components, in this scheme, distribution The method of bandwidth is: the network management monitoring server 304 collects the original network topology data from the router in the bearer network, and stores the topology data collected in the directory server 303. Retrieve relevant data and obtain bandwidth, and then QoS server 301 reads the obtained result from policy server 302 and allocates bandwidth. The bandwidth acquisition process is as follows: use the traffic engineering statistical algorithm based on Multiprotocol Label Switching (MPLS, Multiprotocol Lable Switch) to obtain bandwidth, this method is obtained according to multiple parameters such as the length of the user data message and the round-trip time of the data Bandwidth that needs to be allocated.

The disadvantages of allocating path bandwidth in the above-mentioned Rich QoS scheme are: there is a large amount of network management data communication between the bearer control layer and the bearer network, the bearer control layer has a large amount of bandwidth calculation, and too many servers are involved in the hardware, which consumes a lot of equipment. resources; in addition, the method of obtaining bandwidth also requires the participation of a large number of parameters, and the program is complex, the calculation workload is very large, and the consumption of device resources such as processors is large, resulting in high cost, and it takes time to measure the round-trip time , so the real-time performance of this scheme is very poor.

Contents of the invention

In view of this, the main purpose of the present invention is to provide a method for allocating path bandwidth in the bearer control layer, thereby simplifying the steps of allocating bandwidth, reducing unnecessary waste of resources, and reducing costs.

In order to achieve the above object, the technical solution of the present invention is specifically implemented in the following way:

A method for allocating path bandwidth in the bearer control layer, characterized in that the method comprises:

a: During the resource request process in the bearer control layer, after receiving the connection resource application request, each hop bearer network resource manager selects a path according to the connection resource application request, and according to the maximum path label stack depth MTD in the bearer network and the total packet header length of the original data packet of the user service to obtain the maximum added overhead of the user data passing through the hop path; then according to the maximum added overhead, the bandwidth requested by the user included in the connection resource application request, and the The maximum peak packet length MPPL obtains the bandwidth occupied by the maximum newly added overhead; the sum of the acquired bandwidth occupied by the maximum newly added overhead and the bandwidth requested by the user is assigned to the selected path for each hop as a bandwidth value;

b: After the source bearer network resource manager of the resource request establishes all path connections according to the path selected by the bearer network resource manager of each hop, the bearer network resource manager of each hop uses the relative path label stack depth RTD and The total header length of the original data packet of the user service obtains the actual added overhead when the user data passes through each hop path; then obtains the actual added overhead when the user data passes through each hop path according to the actual added overhead, the bandwidth requested by the user, and MPPL The bandwidth occupied by the added overhead; and the sum of the bandwidth occupied by the actual added overhead when the user data passes through each hop path and the bandwidth requested by the user is used as the bandwidth value to replace the previously allocated bandwidth for each hop path.

In step a, the method for obtaining the maximum new overhead is:

The value obtained by taking MTD×4×2+the total packet header length of the original data packet of the user service is taken as the maximum added overhead.

In step a, the method for obtaining the bandwidth occupied by the maximum new overhead is:

The value obtained by the bandwidth requested by the user×the maximum added overhead/MPPL is used as the bandwidth occupied by the maximum added overhead.

After step b, described method also comprises the following steps:

c. When the bearer network resource managers of each hop receive a connection resource modification request, the bearer network resource managers of each hop obtain user data according to the RTD of each hop path and the total header length of the user service original data packet The actual new overhead when passing through each hop path; then according to the actual new overhead, the bandwidth requested by the user included in the connection resource modification request, and the bandwidth occupied by the actual new overhead when user data passes through each hop path obtained by MPPL ; Add the bandwidth occupied by the actual new overhead to the bandwidth requested by the user, and use the obtained sum value as the bandwidth value to replace the previously allocated bandwidth for each hop path.

The method for obtaining the actual new overhead when user data passes through each hop path is: judge whether the message length of the largest data packet of the user service is greater than the maximum path transmission unit PMTU of the current hop path, and if so, transfer the current hop path RTD × 4 × 2 + the total header length of the original data packet of the user service as the actual new overhead when the user data passes through each hop path; otherwise, the value obtained by the RTD × 4 of the current hop path As the actual new overhead when the user data passes through each hop path.

The packet length of the maximum data packet of the user service is: the maximum peak packet length+4×the RTD of each hop path.

The method for obtaining the bandwidth occupied by the actual new overhead when the user data passes through each hop path is as follows:

The value obtained by the bandwidth requested by the user × the actual added overhead/the maximum peak packet length is used as the bandwidth occupied by the actual added overhead when the user data passes through each hop path.

The total packet header length of the original user service data packet is: the sum of the packet header lengths of all layers that the user data packet passes through. The message headers of each layer include: a link layer message header and an IP message header.

Because the method of the present invention utilizes the resource network bearer to independently allocate the path bandwidth, thereby saving equipment resources, and the method of the present invention can obtain the information required for each The bandwidth allocated by the jump path greatly reduces the complexity of obtaining the bandwidth, and the workload is small, thereby saving a large amount of processor resources and greatly reducing the cost; in addition, the method of the present invention is faster and does not need to measure data Round-trip time, so real-time is good.

Description of drawings

Figure 1 is a diagram of a differentiated service model with an independent bearer control layer;

Fig. 2 is the bandwidth agent model diagram of Qbone;

Fig. 3 is a model diagram of the Rich QoS scheme;

FIG. 4 is a flow chart of completing a resource request in a bearer network;

Fig. 5 is a common message format diagram of a user service original data packet;

Fig. 6 is when (MPPL+4*RTD)＜=PMTU when the user service data message format diagram;

FIG. 7 is a format diagram of user service data packets when (MPPL+4×RTD)>PMTU.

Detailed ways

The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

When the bearer control layer selects a path for a user service connection, it needs to allocate the path bandwidth according to the user's resource request. The method of the present invention is mainly based on the user's message length, requested bandwidth, and path information on the bearer network resource manager. , to allocate bandwidth for the selected path.

The path described in this embodiment refers to a Label Switch Path (LSP, Label Switch Path). Each bearer network resource manager in the bearer control layer can select the management domain under the jurisdiction of the bearer network manager for the service connection requested by the user. LSP in the LSP, obtain the bandwidth resources to be allocated on each hop LSP, and allocate bandwidth for each hop LSP according to the obtained result.

Figure 4 is a flow chart of completing a resource request in the bearer network. As shown in Figure 4, the process of applying for connection resources or modifying and adjusting resources for a business connection usually includes the following steps:

a: CA sends a connection resource application request to the source bearer network resource manager, that is, bearer network resource manager 1. The connection resource application request includes the bandwidth RB requested by the user, and the source bearer network resource manager receives the connection resource application request Finally, select an LSP, and allocate reserved bandwidth on each selected LSP, and then send a connection resource application request to the next-hop bearer network resource manager;

b: After the current bearer network resource manager receives the connection resource application request, it selects an LSP and allocates reserved bandwidth on each selected LSP hop. If the current bearer network resource manager is the destination bearer network resource manager of the resource request device, that is, the bearer network resource manager n, returns a connection resource application response to the bearer network resource manager of the next hop, and performs step c; otherwise, sends a connection resource application request to the bearer network resource manager of the next hop, and returns to step b;

c: The current bearer network resource manager receives the connection resource application response. If the current bearer network resource manager is the source bearer network resource manager of the resource request, all LSP connections of the service connection are established according to the LSP information in the resource response. And return a connection resource application response to the CA; otherwise, return a connection resource application response to the bearer network resource manager of the previous hop, and return to step c.

After the resource manager of the source bearer network establishes all the LSP connections of service connections, it needs to modify and adjust the previously reserved bandwidth according to the information of all LSPs; after that, when the resource manager of the source bearer network receives the When a modification and adjustment request is made for the reserved bandwidth, it is necessary to modify and adjust the previously reserved bandwidth according to the modification and adjustment request. The process of adjusting the bandwidth of these two modifications is the same, and the specific steps are as follows:

d: The resource manager of the source bearer network, according to the bandwidth requested by the user included in the connection resource application request, or the bandwidth requested by the user included in the connection resource modification request, is previously reserved by the source bearer network resource manager The bandwidth is modified and adjusted, and a connection resource modification request is sent to the next source bearer network resource manager;

e: After the current bearer network resource manager receives the connection resource modification request, it modifies and adjusts the bandwidth previously reserved by the current bearer network resource manager. If the current bearer network resource manager is the destination bearer network resource of the resource request The manager, that is, the bearer network resource manager n, returns a connection resource modification response to the bearer network resource manager of the next hop, and performs step f; otherwise, sends a connection resource modification request to the bearer network resource manager of the next hop, and returns to step e ;

f: The current bearer network resource manager receives a connection resource modification response. If the current bearer network resource manager is the source bearer network resource manager of the resource request, it returns a connection resource modification response to the CA; otherwise, it hops up to the bearer network The resource manager returns a connection resource modification response, and returns to step f.

In the whole process of applying for resources or modifying resources, the bearer network resource manager should allocate bandwidth on each hop LSP for the user's service connection, and the method of the present invention is to obtain the required bandwidth on each hop LSP, and according to To allocate bandwidth:

The present embodiment takes IP message transmission as an example to illustrate the method of the present invention. At first, explain what determines the required bandwidth of each hop LSP. FIG. As shown in FIG. 5 , the message of the original data packet of the user service includes: a link layer message header 501 , an IP message header 502 and user service payload data 503 . The link layer message header 501 is the message header of the link layer, the IP message header 502 is the message header added when the message passes through the IP protocol layer, and the user service payload data 503 is the user's business data. The bandwidth RB requested by the user is determined according to the above-mentioned bandwidth occupied by the original data packet of the user service.

When user service data is transmitted through the LSP in the bearer network, the message format of the user service data changes accordingly. Figure 6 is a format diagram of the user service data message at this time. As shown in Figure 6, the user data message includes : link layer message header 501, LSP label stack 601, IP message header 502 and user service payload data 503. In the bearer network, each hop LSP has its own label, and the self-label of each hop LSP and the labels of the previous few hops LSP are stored in the label stack 601 of the hop LSP, and the number of labels stored in the LSP label stack 601 is determined by Relative path Tag stack depth (RTD, Relative path Tag stack Depth) is expressed, that is, in the entire LSP set, the number of LSP hops that each hop LSP has passed relative to the initial CN. For example, the RTD of the first-hop LSP is 1, the RTD of the second-hop LSP is 2, the RTD of the third-hop LSP is 3, and so on. In the differentiated services model with an independent bearer control layer, there is also a specification attribute for the bearer network resource manager, that is, the maximum path tag stack depth (MTD, Max path Tag stack depth), MTD indicates that the service connection is in the entire LSP set , the maximum number of LSP hops allowed to pass, and the value of MTD can be defined according to the scale of the network.

The LSP label stack 601 is a newly added overhead and also occupies bandwidth, so the bandwidth occupied by the LSP label stack 601 should be included when allocating bandwidth. The number of bytes occupied by the newly added overhead is the length of the LSP label stack 601, and the length of the LSP label stack 601=RTD×the number of bytes occupied by each label, and because in the user data service message, the label stack The byte length of each label is 4 bytes, so the length of the LSP label stack is RTD×4.

Since the bandwidth is not only related to the length of the message, but also related to the transmission frequency of the message, and the length of the message transmitted every moment is also variable, so the maximum peak packet length (MPPL, MaxPeak Packet Length) is used to determine Indicates the message that occupies the largest bandwidth in the user's service connection, and the MPPL is the maximum message length of a single data packet under the peak bandwidth of the user's service connection; The data packet is the maximum transmission unit (MTU, Max Transfer Unit); between CN and CN, among all the LSPs that the business connection can pass through, the data packet with the smallest MTU value is the maximum path transmission unit (PMTU, Path Max Transfer Unit).

In peak conditions, the message length of the largest data packet of the user service is: the original message length + the newly added overhead length, that is, (MPPL+4×RTD). When (MPPL+4×RTD)<=PMTU, as shown in Figure 6, compared to the user's original ordinary data packet shown in Figure 5, the new overhead of the data packet is the LSP label stack 601, so when obtaining the bandwidth To include the bandwidth occupied by the added overhead, the added overhead is: the length of the LSP label stack 601 , that is, RTD×4.

When (MPPL+4×RTD)>PMTU, the current LSP does not allow the data packet of the user service to pass through, so at this time, the data packet must be fragmented, that is, the payload data of the user service in a data packet is divided into Load two data packets, and these two data packets can pass through the current LSP, as shown in Figure 7, the user's original data packet is divided into data packet 701 and data packet 702, data packet 701 includes: link layer message Header 501, LSP label stack 601, IP packet header 502 and the first part 703 of user service payload data 503; Data packet 702 includes: link layer packet header 501, LSP label stack 601, IP packet header 502 and user The remainder 704 of the traffic payload data 503 . Compared with the user's original data packet shown in Figure 5, the LSP label stack 601 in the data packet 701 is an additional overhead of 1, and the link layer header 501, LSP label stack 601 and IP packet in the data packet 702 The header 502 is the new overhead 2, so these two parts of the new overhead should be taken into account when obtaining the bandwidth, so the new overhead of the entire data packet is: new overhead 1 + new overhead 2 = LSP label stack length × 2+ (link layer header length+IP header length), where the length of the LSP label stack is: RTD×the number of bytes occupied by each label, that is, RTD×4.

As mentioned above, the data packet transmitted in the bearer network adds new overhead compared with the original data packet of the user, and these new overheads occupy a part of the bandwidth. Therefore, the bandwidth allocated for the current hop LSP is obtained by formula (1):

Bandwidth allocated for the current hop LSP = RB+ΔRB (1)

In formula (1), the unit of bandwidth is bit/second (bps), RB is the bandwidth requested by the user, and ΔRB is the bandwidth occupied by the new overhead. Since the new overhead will be different in different situations, the value of ΔRB is different in different situations, and the corresponding current LSP bandwidth is also different, as follows:

When allocating reserved bandwidth for each hop LSP in the above step a and step b, since it is impossible to fully determine all the final LSP connections of this service connection, in order to ensure that the reserved bandwidth allocated for the current hop LSP is sufficient for use, so at this time The bandwidth that needs to be reserved for the current hop LSP is obtained according to the maximum new cost, that is, the maximum new cost is: the sum of the above new cost 1 and the new cost 2, and the depth of the label stack of the current hop LSP is the largest MTD, such as formula (2):

ΔRB＝RB×Maximum new overhead/MPPL (2)

In formula (2), the added overhead is the number of bytes occupied by the largest added overhead, that is, 4×MTD×2+(link layer header length+IP header length).

After the entire LSP is established successfully, the bearer network resource manager of each hop knows the RTD of each hop LSP, so it will make a resource modification and adjustment for the entire set of previously applied LSPs, that is, the RTD of each previous hop LSP Modify and adjust the reserved bandwidth; or modify and adjust the previously reserved bandwidth for the LSP due to other reasons, for example, the resource manager of the bearer network may receive a connection resource modification request from the CA, so the original bandwidth needs to be modified and adjusted . At this time, in order to obtain the bandwidth allocated for the current hop LSP more accurately, the bandwidth is obtained and allocated according to the precise new overhead. There are two situations at this time:

If (MPPL+4×RTD)<=PMTU, then:

ΔRB＝RB×accurate new overhead/MPPL (3)

As shown in FIG. 6 , the added overhead described in formula (3) is the length of the label stack of the current hop LSP, that is, 4×RTD.

If (MPPL+4×RTD)>PMTU, then:

ΔRB＝RB×accurate new overhead/MPPL (4)

As shown in Figure 7, the new overhead described in formula (4) is: the number of bytes occupied by the label stack of the current hop LSP × 2 + (link layer header length + IP header length), that is (4 × RTD×2+(link layer header length+IP header length)).

In this embodiment, the user's net business data is in the second layer, that is, the IP layer. If the user's net business data is in the third layer and above layer protocols, the above-mentioned IP header length should be replaced by: IP header length+three layers and The sum of headers of layers above three layers.

Under normal circumstances, the method described in the above-mentioned embodiments is used to obtain and allocate path bandwidth, but the method described in the present invention can also only use formula (2) to obtain the bandwidth occupied by the new overhead, and calculate the bandwidth to be allocated accordingly , and will not modify and adjust the allocated bandwidth afterwards. Although this implementation manner is relatively simple and has a small amount of calculation, the accuracy is not high, and bandwidth resources are likely to be wasted. This implementation mode is applicable to services with high time requirements and low bandwidth requirements, but generally, this implementation mode is not used.

The above is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person familiar with the technology can easily think of changes or replacements within the technical scope disclosed in the present invention. , should be covered within the protection scope of the present invention.

Claims (8)

1. A method for allocating path bandwidth in the bearer control layer, characterized in that the method comprises: a: During the resource request process in the bearer control layer, after receiving the connection resource application request, each hop bearer network resource manager selects a path according to the connection resource application request, and according to the maximum path label stack depth MTD in the bearer network and the total packet header length of the original data packet of the user service to obtain the maximum added overhead of the user data passing through the hop path; then according to the maximum added overhead, the bandwidth requested by the user included in the connection resource application request, and the The maximum peak packet length MPPL obtains the bandwidth occupied by the maximum newly added overhead; the sum of the acquired bandwidth occupied by the maximum newly added overhead and the bandwidth requested by the user is assigned to the selected path for each hop as a bandwidth value; b: After the source bearer network resource manager of the resource request establishes all path connections according to the path selected by the bearer network resource manager of each hop, the bearer network resource manager of each hop uses the relative path label stack depth RTD and The total header length of the original data packet of the user service obtains the actual added overhead when the user data passes through each hop path; then obtains the actual added overhead when the user data passes through each hop path according to the actual added overhead, the bandwidth requested by the user, and MPPL The bandwidth occupied by the added overhead; and the sum of the bandwidth occupied by the actual added overhead when the user data passes through each hop path and the bandwidth requested by the user is used as the bandwidth value to replace the previously allocated bandwidth for each hop path. 2. The method according to claim 1, characterized in that, in step a, the method for obtaining the maximum newly added overhead is: The value obtained by taking MTD×4×2+the total packet header length of the original data packet of the user service is taken as the maximum added overhead. 3. The method according to claim 1, characterized in that, in step a, the method for obtaining the bandwidth occupied by the maximum newly added overhead is: The value obtained by the bandwidth requested by the user×the maximum added overhead/MPPL is used as the bandwidth occupied by the maximum added overhead. 4. The method according to claim 1, characterized in that, after step b, said method further comprises the following steps: c. When the bearer network resource managers of each hop receive a connection resource modification request, the bearer network resource managers of each hop obtain user data according to the RTD of each hop path and the total header length of the user service original data packet The actual new overhead when passing through each hop path; then according to the actual new overhead, the bandwidth requested by the user included in the connection resource modification request, and the bandwidth occupied by the actual new overhead when user data passes through each hop path obtained by MPPL ; Add the bandwidth occupied by the actual new overhead to the bandwidth requested by the user, and use the obtained sum value as the bandwidth value to replace the previously allocated bandwidth for each hop path. 5. The method according to claim 1 or 4, characterized in that the method for obtaining the actual new overhead when user data passes through each hop path is: judging whether the message length of the largest data packet of the user service is greater than the current If it is the maximum path transmission unit PMTU of the hop path, the value obtained by the current hop path RTD×4×2+the total packet header length of the user service original data packet is used as the actual value when the user data passes through each hop path Added overhead; otherwise, the value obtained by RTD×4 of the current hop path is used as the actual added overhead when the user data passes through each hop path. 6. The method according to claim 5, wherein the packet length of the maximum data packet of the user service is: maximum peak packet length+4×RTD of each hop path. 7. The method according to claim 1 or 4, wherein the method for obtaining the bandwidth occupied by the actual new overhead when the user data passes through each hop path is as follows: The value obtained by the bandwidth requested by the user × the actual added overhead/the maximum peak packet length is used as the bandwidth occupied by the actual added overhead when the user data passes through each hop path. 8. The method according to claim 1, wherein the total header length of the original user service data packet is: the sum of the header lengths of all layers that the user data packet passes through. 9. The method according to claim 8, wherein the packet headers of each layer include: a link layer packet header and an IP packet header.

Images