Network measurement optimization system and method based on detection target scheduling

Abstract

The invention discloses a network measurement optimization system and method based on detection target scheduling. The method comprises the following steps: obtaining detection configuration data of a detection scene; classifying each detection target in the detection target sequence into a corresponding network segment according to the belonging network segment sequence, organizing the detection targets by using a plurality of linked lists, outputting a detection target linked list, counting the total number of the detection targets of each network segment based on the measurement basic parameters, and outputting the total number of the detection targets; the number of distribution times of each network segment in a complete period is calculated, detection targets are taken out from the corresponding linked lists in sequence and filled into a scheduling array, and filling of integer parts is completed; and traversing all vacancies in positions which are not fully filled and unallocated detection target linked lists in each network segment in sequence, taking out residual detection targets from each network segment linked list in sequence for filling, and outputting a complete scheduling array which completes filling. According to the invention, the peak detection flow of the target network segment is reduced, and the network measurement effect is improved.

Description

Technical Field

[0001] The present invention relates to the field of network measurement, and in particular to a network measurement optimization system and method based on probe target scheduling. Background Art

[0002] The Internet is an important infrastructure in the world today and is indispensable in people's modern lives. At the personal level, mobile payment, social media, etc. have deeply changed people's daily lives; at the enterprise level, cloud computing and remote work have gradually become indispensable. Nowadays, the Internet has become an important part of the world. Therefore, the research on the Internet is also an important research direction in the academic community. And in today's network research, network measurement is an important field. Through various active and passive measurements, researchers can obtain information about addresses, topologies, ports, and services on the network, so as to conduct multi-faceted research.

[0003] In network measurement tasks, the address distribution of probe targets is often uneven, whether it is direct list scanning or measurement targets generated by various address generation algorithms. For example, in the Gasser Hitlist on March 23, 2025, the routing prefix with the most addresses accounted for 5.4% of the total number of addresses, while the addresses belonging to the top 10% of the routing prefixes accounted for 97.9% of the total number of addresses. This will lead to an excessive number of probe packets in a specific target network, and the uneven distribution of probe targets on the time axis exacerbates this situation.

[0004] A common network measurement process can be simply divided into three links: generating probe targets, sending probes, and receiving and parsing responses. Past research work often focused on the first link and ignored the optimization of other links. The present invention will optimize the link of sending probes, and propose a scheme for rearranging probe targets to alleviate the situation of excessive local traffic during network measurement.

[0005] In the link of sending probes in network measurement, very little work has been carried out on this link in past research work. The most representative one is ZMap that appeared in 2013. In ZMap, the author innovatively proposed stateless detection, thereby decoupling the sending and receiving processes in the network measurement process. Based on such a design, ZMap can complete the scanning of the entire IPv4 address space within 45 minutes. When scanning within a fixed prefix range, ZMap designed a scheme based on group theory to complete internal randomization. However, when facing a list scanning task, ZMap still executes a sequential scanning strategy and does not adjust the detection order.

[0006] If we focus on the related work of probe sending order, it can be found that past researchers did not conduct in-depth research on the probe sending order. Many studies directly performed list scans or added simple randomization during the scan. As time went by, many network measurement works that rely on specific detection strategies and require dynamic adjustment during the detection process emerged. For example, network measurement works such as 6Scan and 6Graph rely on traversals on tree-like data structures. These measurement works often cannot interfere significantly with the probe sending process. To sum up, it can be found that in past work, researchers did not pay attention to the sending order of detection targets.

[0007] As mentioned above, in existing network measurement works, there is a problem that detection targets are unevenly distributed in time and address space. When a large number of detection targets belonging to the same network segment are detected in a short period of time, it will cause excessive traffic in this local area, and then bring a series of problems, such as interfering with normal network activities in the network segment, triggering speed limits and affecting detection results, and causing detection behaviors to be discovered by mechanisms such as IDS. Summary of the Invention

[0008] The present invention aims to solve at least one of the technical problems in the related art to some extent.

[0009] The present invention proposes a network measurement optimization system based on detection target scheduling, designs a rearrangement scheme for detection targets, and reduces the peak traffic in the target network segment by rearranging detection targets according to a specific strategy. On the one hand, this can reduce the impact of detection behaviors on the target network, and at the same time can avoid speed limits in the target network segment to a certain extent, thereby improving the performance of network measurement.

[0010] Another object of the present invention is to propose a network measurement optimization method based on detection target scheduling.

[0011] To achieve the above object, on the one hand, the present invention proposes a network measurement optimization system based on detection target scheduling, including:

[0012] A data acquisition module, configured to acquire detection configuration data of a detection scenario; wherein, the detection configuration data includes a detection target sequence, a network segment sequence to which the detection targets belong, and measurement basic parameters;

[0013] A network segment classification module, configured to classify each detection target in the detection target sequence into the corresponding network segment according to the network segment sequence to which the detection targets belong, organize the detection targets using multiple linked lists, output a detection target linked list, and count the total number of detection targets in each network segment based on the measurement basic parameters, and output the total number of detection targets;

[0014] An integer filling module is used to calculate the number of allocations of each network segment in a complete cycle based on the detection target linked list and the total number of detection targets, and to sequentially take out the detection targets from the corresponding linked list and fill them into the scheduling array to complete the filling of the integer part, so as to generate a preliminary periodic detection sequence;

[0015] The remainder filling module is used to traverse all empty positions in sequence for the positions that are not completely filled and the unassigned detection target linked lists in each network segment, and take out the remaining detection targets from the linked lists of each network segment in turn to fill them, and output a complete scheduling array that has been filled.

[0016] The network measurement optimization system based on detection target scheduling according to the embodiment of the present invention may also have the following additional technical features:

[0017] In one embodiment of the present invention, the data acquisition module is further used to acquire the sequence of the detection target as A i (1≤i≤n), the sequence length is n, and a probe data packet needs to be sent to each detection target; the network segment to which each detection target belongs is N i (1≤i≤n).

[0018] In one embodiment of the present invention, the integer filling module is further used for:

[0019] Fill in each part marked with M in the relationship diagram between the detection task scheduling period and the network segment allocation; M is calculated as follows: in That is, the total duration of the network measurement process;

[0020] The integer filling module generates a period array corresponding to the target network segment in each M period; the period array is filled in order to fill in M ​​for each network segment x. x x, fill in the detection target held by the network segment at the corresponding position of the period array for each M period, and the use of the detection target depends on the linked list generated in the network segment classification module.

[0021] In one embodiment of the present invention, the remainder filling module is further used for:

[0022] Fill each part marked as rem in the relationship diagram between the detection task scheduling period and the network segment allocation; traverse all the positions in rem in a preset order: first traverse the first position of each rem part sequentially, and then the second position. While traversing the empty positions sequentially, set another variable cur to traverse all the network segments and take out the detection targets corresponding to the network segments and put them into the above-mentioned empty positions; whenever the detection targets corresponding to cur are exhausted, point cur to the next network segment.

[0023] In one embodiment of the present invention, the system is further configured to:

[0024] Set a sliding window with a length of R, and the coverage range is [i, i + R - 1] (1 ≤ i ≤ n - R + 1). The detection targets in the sliding window correspond to the traffic sent within a certain second in the network measurement process, which is t x ; Since the part corresponding to M circulates at intervals of R according to the network segment to which it belongs, when only considering the M part, the number of network segment x in the sliding window is M x ; It is necessary to consider the number of occurrences of network segment x in the rem part; consider S x = M x T + r x (0 ≤ r x < T):

[0025] If r x = 0, there will be no detection targets belonging to network segment x in rem, that is

[0026] If r x ≠0, consider the situation where x appears in the rem part. Because r x < T, at most one x appears in each rem, and there is at least one rem without x; consider the process of filling rem, there is at least a gap of R between adjacent x; in the rem part, each sliding window will only have at most 1 occurrence of x, that is

[0027] The number of occurrences t of network segment x in the sliding window x Satisfy That is, to make each network segment x reach the theoretical optimum.

[0028] To achieve the above object, on the other hand, the present invention proposes a network measurement optimization method based on detection target scheduling, including:

[0029] Obtain the detection configuration data of the detection scenario; wherein, the detection configuration data includes a detection target sequence, a network segment sequence to which it belongs, and measurement basic parameters;

[0030] Classify each detection target in the detection target sequence into the corresponding network segment according to the network segment sequence to which it belongs, and organize the detection targets using multiple linked lists, output the detection target linked list, and count the total number of detection targets in each network segment based on the measurement basic parameters, and output the total number of detection targets;

[0031] Based on the detection target linked list and the total number of detection targets, the number of allocations of each network segment in a complete cycle is calculated, and the detection targets are taken out from the corresponding linked list in order to fill in the scheduling array, completing the filling of the integer part to generate a preliminary periodic detection sequence;

[0032] For the positions that are not completely filled and the unassigned detection target linked lists in each network segment, all empty positions are traversed in order, and the remaining detection targets are taken out from the linked lists of each network segment in turn to fill them, and the complete scheduling array with the filling is output.

[0033] The network measurement optimization system and method based on detection target scheduling of the embodiments of the present invention can not only improve the performance of network measurement and reduce the impact of network measurement behavior on the target network, but also reduce the target network's perception of the detection behavior based on reducing the detection traffic to the target network segment. In many traffic monitoring works for network measurement, the traffic scale per unit time is an important basis for these IDS to judge scanning behavior. Therefore, reducing the detection traffic to the target network segment can also improve the concealment of network measurement behavior. The present invention designs a scheduling algorithm for the detection target and rearranges the order of the detection targets when sending probes, thereby reducing the peak detection traffic to the target network segment as much as possible, thereby improving the effect of network measurement.

[0034] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0036] Figure 1 is a structural diagram of a network measurement optimization system based on detection target scheduling according to an embodiment of the present invention;

[0037] Figure 2 is a data structure diagram of a network measurement optimization system based on detection target rearrangement according to an embodiment of the present invention;

[0038] Figure 3 is a diagram showing the relationship between the detection task scheduling period and the network segment allocation according to an embodiment of the present invention;

[0039] Figure 4 is a flow chart of a network measurement optimization method based on detection target scheduling according to an embodiment of the present invention. DETAILED DESCRIPTION

[0040] It should be noted that, without conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the drawings and in conjunction with the embodiments.

[0041] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts shall fall within the protection scope of the present invention.

[0042] The network measurement optimization system and method based on probe target scheduling proposed according to the embodiments of the present invention will be described below with reference to the drawings.

[0043] Figure 1 is a structural diagram of a network measurement optimization system based on probe target scheduling according to an embodiment of the present invention, as Figure 1 shown, the system 10 includes:

[0044] A data acquisition module 100, configured to acquire probe configuration data of a probe scenario; wherein, the probe configuration data includes a probe target sequence, a network segment sequence to which the probe target belongs, and measurement basic parameters;

[0045] A network segment classification module 200, configured to classify each probe target in the probe target sequence into the corresponding network segment according to the network segment sequence to which the probe target belongs, organize the probe targets using multiple linked lists, output a probe target linked list, and count the total number of probe targets in each network segment based on the measurement basic parameters, and output the total number of probe targets;

[0046] An integer filling module 300, configured to calculate the allocation times of each network segment in a complete cycle based on the probe target linked list and the total number of probe targets, and sequentially take out probe targets from the corresponding linked list and fill them into a scheduling array to complete the filling of the integer part, so as to generate a preliminary periodic probe sequence;

[0047] A remainder filling module 400, configured to sequentially traverse all empty positions for the positions that are not fully filled and the probe target linked lists that are not allocated in each network segment, and sequentially take out the remaining probe targets from the linked lists of each network segment for filling, and output a complete scheduling array after completion of filling.

[0048] Specifically, the present invention proposes a network measurement optimization system based on probe target rearrangement, as Figure 2 shown, the system is composed of a data acquisition module, a network segment classification module, an integer filling module, and a remainder filling module.

[0049] Among them, the data acquisition module: The sequence of detection targets is A i (1 ≤ i ≤ n), the sequence length is n, and a probe data packet needs to be sent for each detection target; the network segment to which each detection target belongs is N i (1 ≤ i ≤ n), and this parameter will change according to the conditions of the problem.

[0050] Exemplarily, the most straightforward solution is to directly truncate the address of the detection target to / 64, and use this prefix as the network segment to which the detection target belongs in the present invention; if the differences in network segment sizes in different scenarios are considered, the AS to which the detection target belongs can be used as the network segment; in the topology detection scenario, TTL can also be introduced into the setting process of this parameter.

[0051] Meanwhile, the data acquisition module also includes a series of basic parameters for network measurement, such as the packet sending rate R during the network detection process, with the unit of packets per second (i.e., pps), and the batch size during the packet sending process, etc. Next, the other three modules of this system will be introduced in detail.

[0052] In an embodiment of the present invention, the network segment classification module. This module is responsible for classifying each detection target in the detection target sequence into its corresponding network segment, and organizing these detection targets using multiple linked lists, with each network segment corresponding to a linked list. Since the network segment N to which each detection target belongs i can have different calculation methods under different detection tasks, so N i will directly become the input of the entire solution rather than being calculated within the solution. Using linked lists for organization is to facilitate subsequent modules to retrieve each detection target according to the network segment to which it belongs for rearrangement. Since a linked list insertion needs to be performed for each of the n detection targets, the time complexity of this process is O(n). Meanwhile, this module will also complete a series of basic numerical calculations, including the statistics of the total number of detection targets in each network segment, etc.

[0053] In summary, the network segment classification module will complete the organization of the detection targets and make some basic preparations for the subsequent two modules, while the following two modules will complete the rearrangement of the entire detection target sequence.

[0054] In an embodiment of the present invention, the integer filling module. This module is responsible for filling Figure 3 each part marked as M. And the calculation method of M is: Among them That is, the total duration of the network measurement process. The integer filling module will first generate a period array corresponding to the target network segments in each M cycle. The filling method of the period array is that for each network segment x, M is filled in sequence xx. Then, for each M period, the detection target held by the network segment at the corresponding position in the period array is filled in. The use of the detection target depends on the linked list generated in the network segment classification module. The operation mode of this part is Figure 3 The upper part of the diagram is demonstrated, and the specific implementation scheme is described in the following algorithm, as shown in Table 1.

[0055] Table 1

[0056]

[0057] In one embodiment of the present invention, the remainder filling module. This module is responsible for filling Figure 3 Each part marked as rem in the network. The present invention traverses all the positions in rem in a certain order: first, the first position of each rem part is sequentially traversed, followed by the second position, and so on. While sequentially traversing these empty positions, the present invention sets another variable cur, which will traverse all network segments and continuously take out the detection targets corresponding to the network segment and put them into the above-mentioned empty positions. Whenever the detection targets corresponding to cur are exhausted, cur is pointed to the next network segment. In this way, the network can be reasonably filled. Figure 3 All the rem parts in the . This part operates in Figure 3 The specific implementation scheme is described in the following algorithm.

[0058] Table 2

[0059]

[0060] It can be found that the integer filling module and the remainder filling module only traverse the entire detection target sequence once, and the time consumption for each position in the detection target sequence is O(1), so the time complexity of the whole scheme is O(n), which is a negligible time overhead for common network measurement work.

[0061] Next, the present invention will explain that this optimization scheme can achieve the theoretical optimum peak flow rate for all target network segments. Consider a sliding window of length R, whose coverage range is [i,i+R-1] (1≤i≤n-R+1). The detection target in this sliding window corresponds to the flow rate emitted within a certain second in the network measurement process. The present invention sets it as t x Considering the complete detection target sequence in the middle of the figure above, it is obvious that the detection targets in the sliding window can be divided into the part corresponding to M and the part corresponding to rem. Since the part corresponding to M cycles at intervals of R according to the network segment to which it belongs, when only considering the M part, the number of network segments x in the sliding window should be M xNext, it is necessary to consider the number of occurrences of network segment x in the rem part. Consider S x = M x T + r x (0 ≤ r x < T):

[0062] If r x = 0, then there will be no detection target belonging to network segment x in rem, that is

[0063] If r x ≠ 0, then considering the situation where x appears in the rem part, it can be found that because r x < T, at most one x will appear in each rem, and there is at least one rem without x; considering the process of filling rem, there will be at least an interval of R between adjacent x's. Therefore, in the rem part, each sliding window will have at most 1 occurrence of x, that is

[0064] In summary, the number of occurrences t of network segment x in the sliding window x satisfies That is, the present invention can make each network segment x reach the theoretical optimum.

[0065] The beneficial effects of the present invention are as follows:

[0066] 1) By adding the rearrangement of detection targets in the probe sending link of network measurement, the present invention can evenly distribute the detection traffic for a single network segment throughout the entire detection process, and thereby reduce the peak value of the detection traffic received by the target network segment during the detection process. By reducing the peak value of the detection traffic, the present invention can reduce the impact of network measurement behavior on normal network activities in the target network segment, which is helpful for maintaining the network environment.

[0067] 2) There are a large number of speed limit behaviors in the Internet, and network measurement is often affected by them. By reducing the peak value of the detection traffic, the present invention can help network measurement behavior bypass some speed limits in the Internet, thereby obtaining more effective responses in network measurement, and further achieving better detection results and improving the efficiency of network measurement.

[0068] 3) Reducing the peak value of the detection traffic for the target network segment can also reduce the perception of the target network for this detection behavior. In many traffic monitoring works for network measurement, the traffic scale within a unit time is an important basis for these IDSs to judge scanning behaviors. Therefore, reducing the detection traffic for the target network segment can also improve the concealment of network measurement behavior.

[0069] According to the network measurement optimization system based on probe target scheduling in an embodiment of the present invention, by evenly distributing the probe traffic for a single network segment throughout the entire probing process, the peak value of the probe traffic in the target network segment is reduced. This enables all target network segments to achieve a theoretically optimal traffic peak. At the same time, the time complexity of this solution is only O(n), and the additional time overhead added to the network measurement process is almost negligible. The present invention is the first method to optimize network measurement in the IPv6 scenario from the perspective of probe target scheduling, and reduces the peak value of the probe traffic in the target network segment by rearranging the probe targets during the probe sending process. Finally, by designing a scheduling algorithm for the probe targets, the order of the probe targets is rearranged when sending probes, thereby minimizing the peak probe traffic to the target network segment and improving the effect of network measurement.

[0070] To implement the above embodiment, as Figure 4 shown, the present embodiment also provides a network measurement optimization method based on probe target scheduling, including:

[0071] S1. Obtain the probe configuration data of the probing scenario; wherein, the probe configuration data includes a probe target sequence, a network segment sequence to which the probe target belongs, and measurement basic parameters;

[0072] S2. Classify each probe target in the probe target sequence into the corresponding network segment according to the network segment sequence to which the probe target belongs, and organize the probe targets using multiple linked lists, output the probe target linked list, and count the total number of probe targets in each network segment based on the measurement basic parameters, and output the total number of probe targets;

[0073] S3. Based on the probe target linked list and the total number of probe targets, calculate the allocation times of each network segment in a complete cycle, and sequentially take out the probe targets from the corresponding linked list and fill them into the scheduling array to complete the filling of the integer part, so as to generate a preliminary periodic probing sequence;

[0074] S4. Traverse all empty positions in sequence for the positions that are not fully filled and the probe target linked lists that are not allocated in each network segment, and sequentially take out the remaining probe targets from the linked lists of each network segment for filling, and output the complete scheduling array after filling.

[0075] Further, obtaining the probe configuration data of the probing scenario includes:

[0076] Obtain the sequence of probe targets as A i (1 ≤ i ≤ n), the sequence length is n, and a probe data packet needs to be sent for each probe target; the network segment to which each probe target belongs is N i (1 ≤ i ≤ n).

[0077] Furthermore, based on the detection target linked list and the total number of detection targets, the number of allocations of each network segment in a complete cycle is calculated, and the detection targets are sequentially taken out from the corresponding linked list and filled into the scheduling array to complete the filling of the integer part, so as to generate a preliminary periodic detection sequence, including:

[0078] Fill in each part marked with M in the relationship diagram between the detection task scheduling period and the network segment allocation; M is calculated as follows: in That is, the total duration of the network measurement process;

[0079] Generate a period array corresponding to the target network segment in each M period; the period array is filled in by filling M in order for each network segment x x x, fill in the detection target held by the network segment at the corresponding position of the period array for each M period, and the use of the detection target depends on the linked list generated in the network segment classification module.

[0080] Furthermore, the unfilled positions and the unallocated detection target linked lists in each network segment are sequentially traversed through all the empty positions, and the remaining detection targets are sequentially taken out from each network segment linked list for filling, and a complete scheduling array with the filling completed is output, including:

[0081] Fill each part marked as rem in the relationship diagram between the detection task scheduling period and the network segment allocation; traverse all the positions in rem in a preset order: first traverse the first position of each rem part sequentially, and then the second position. While traversing the empty positions sequentially, set another variable cur to traverse all the network segments and take out the detection targets corresponding to the network segments and put them into the above-mentioned empty positions; whenever the detection targets corresponding to cur are exhausted, point cur to the next network segment.

[0082] Furthermore, the method further comprises:

[0083] Set a sliding window of length R, with a coverage range of [i,i+R-1](1≤i≤n-R+1). The detection target in the sliding window corresponds to the traffic sent within a certain second in the network measurement process, which is t x ; Since the part corresponding to M cycles at intervals of R according to the network segment to which it belongs, when only the M part is considered, the number of network segments x in the sliding window is M x ; It is necessary to consider the number of occurrences of network segment x in the rem part; Consider S x =M x T+r x (0≤r x <T):

[0084] If r x= 0, there will be no detection targets belonging to network segment x in rem, that is,

[0085] If r x ≠0, consider the case where x appears in the rem part, because r x <T,每个rem中最多出现一个x,且至少有一个rem中没有x;考虑填充rem的过程,相邻的x之间至少会间隔R;rem部分中,每个滑动窗口只会出现至多1次x,也即

[0086] The number of times segment x appears in the sliding window is t x satisfy That is to say, each network segment x can reach the theoretical optimum.

[0087] According to the network measurement optimization method based on detection target scheduling of the embodiment of the present invention, the detection traffic for a single network segment is evenly distributed to the entire detection process, thereby reducing the peak value of the detection traffic in the target network segment. The traffic peaks faced by all target network segments can reach the theoretical optimum. At the same time, the time complexity of this solution is only O(n), and the additional time overhead added in the network measurement process can be almost ignored. The present invention is the first method to optimize network measurement in an IPv6 scenario from the perspective of detection target scheduling, and reduces the peak value of the detection traffic in the target network segment by rearranging the detection targets in the link of probe sending. Finally, by designing a scheduling algorithm for the detection targets, the order of the detection targets is rearranged when sending the probes, thereby reducing the peak detection traffic to the target network segment as much as possible, thereby improving the effect of network measurement.

[0088] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0089] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include at least one of such features. In the description of the present invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise specifically defined.

Claims

1. A network measurement optimization system based on detecting target scheduling, characterized in that Including: A data acquisition module, configured to acquire detection configuration data of a detection scenario; wherein, the detection configuration data includes a detection target sequence, a network segment sequence to which the detection targets belong, and measurement basic parameters; A network segment classification module, configured to classify each detection target in the detection target sequence into the corresponding network segment according to the network segment sequence to which the detection targets belong, organize the detection targets using multiple linked lists, output a detection target linked list, and count the total number of detection targets in each network segment based on the measurement basic parameters, and output the total number of detection targets; An integer filling module, configured to calculate the allocation times of each network segment in a complete cycle based on the detection target linked list and the total number of detection targets, and take out the detection targets from the corresponding linked list in sequence and fill them into a scheduling array to complete the filling of the integer part, so as to generate a preliminary periodic detection sequence; A remainder filling module, configured to traverse all empty positions in sequence for the positions that are not completely filled and the detection target linked lists that are not allocated in each network segment, and sequentially take out the remaining detection targets from the linked lists of each network segment for filling, and output a complete scheduling array after filling.

2. The system according to claim 1, characterized in that, The data acquisition module is further configured to acquire that the sequence of the detection target is A i (1 ≤ i ≤ n), the sequence length is n, and a probe data packet needs to be sent for each detection target; the network segment to which each detection target belongs is N i (1 ≤ i ≤ n).

3. The system according to claim 1, wherein The integer filling module is further configured to: Fill in each part marked as M in the diagram showing the relationship between the detection task scheduling period and network segment allocation; the calculation method of M is as follows: where that is, the total duration of the network measurement process; The integer filling module generates a period array corresponding to the target network segment in each M period; the period array is filled in order to fill in M ​​for each network segment x. x x, fill in the detection target held by the network segment at the corresponding position of the period array for each M period, and the use of the detection target depends on the linked list generated in the network segment classification module.

4. The system according to claim 1, wherein The remainder filling module is further configured to: Fill each part marked as rem in the graph representing the relationship between the detection task scheduling period and the network segment allocation; traverse all positions in rem in a preset order: first traverse the first position of each rem part in sequence, then the second position, and while traversing the empty positions in sequence, set another variable cur, which will traverse all network segments and take out the detection targets corresponding to the network segments and put them into the above-mentioned empty positions; whenever the detection targets corresponding to cur are exhausted, cur will point to the next network segment.

5. The system according to claim 1, wherein The system is further configured to: Set a sliding window with a length of R, covering the range [i, i + R - 1] (1 ≤ i ≤ n - R + 1). The detection targets in the sliding window correspond to the traffic sent within a certain second in the network measurement process, which is t x ; Since the part corresponding to M cycles at an interval of R according to the network segment it belongs to, when only considering the M part, the number of occurrences of network segment x in the sliding window is M x ; It is necessary to consider the number of occurrences of network segment x in the rem part; consider S x = M x T + r x (0 ≤ r x < T): If r x = 0, there will be no detection targets belonging to network segment x in rem, that is If r x ≠ 0, consider the case where x appears in the rem part. Since r x < T, at most one x appears in each rem, and there is at least one rem without x. Consider the process of filling the rem. There will be at least a gap of R between adjacent x's. In the rem part, each sliding window will have at most one occurrence of x, that is The number of occurrences t of network segment x in the sliding window x Meet That is, to make each network segment x reach the theoretical optimum.

6. A network measurement optimization method based on detecting target scheduling, characterized in that, Including: Acquire detection configuration data of a detection scenario; wherein, the detection configuration data includes a detection target sequence, a network segment sequence to which the detection targets belong, and measurement basic parameters; Classify each detection target in the detection target sequence into the corresponding network segment according to the network segment sequence to which the detection targets belong, organize the detection targets using multiple linked lists, output a detection target linked list, and count the total number of detection targets in each network segment based on the measurement basic parameters, and output the total number of detection targets; Calculate the allocation times of each network segment in a complete cycle based on the detection target linked list and the total number of detection targets, and take out the detection targets from the corresponding linked list in sequence and fill them into a scheduling array to complete the filling of the integer part, so as to generate a preliminary periodic detection sequence; Traverse all empty positions in sequence for the positions that are not completely filled and the detection target linked lists that are not allocated in each network segment, and sequentially take out the remaining detection targets from the linked lists of each network segment for filling, and output a complete scheduling array after filling.

7. The method according to claim 6, characterized in that, Acquire detection configuration data of a detection scenario, including: Obtain the sequence of detection targets as A i (1 ≤ i ≤ n), the sequence length is n, and a probe data packet needs to be sent for each detection target; the network segment to which each detection target belongs is N i (1 ≤ i ≤ n).

8. The method according to claim 6, wherein The calculation of the allocation times of each network segment in a complete cycle based on the detection target linked list and the total number of detection targets, and taking out the detection targets from the corresponding linked list in sequence and filling them into a scheduling array to complete the filling of the integer part, so as to generate a preliminary periodic detection sequence, includes: Fill the part marked as M in the diagram showing the relationship between the detection task scheduling period and network segment allocation; the calculation method of M is as follows: where i.e., the total duration of the network measurement process; Generate a period array corresponding to the target network segment in each M period; the period array is filled in by filling M in order for each network segment x x x, fill in the detection target held by the network segment at the corresponding position of the period array for each M period, and the use of the detection target depends on the linked list generated in the network segment classification module.

9. The method according to claim 6, characterized in that, Traverse all empty positions in sequence for the positions that are not fully filled and the unallocated detection target linked lists in each network segment, and sequentially take out the remaining detection targets from the linked lists of each network segment for filling, and output a complete scheduling array after filling, including: Fill the part marked as rem in the diagram showing the relationship between the detection task scheduling period and network segment allocation; traverse all positions in rem in a preset order: first traverse the first position of each rem part in sequence, then the second position. While traversing the empty positions in sequence, set another variable cur, traverse all network segments, and take out the detection targets corresponding to the network segments and put them into the above-mentioned empty positions; whenever the detection targets corresponding to cur are exhausted, point cur to the next network segment.

10. The method according to claim 6, wherein The method further includes: Set a sliding window of length R, with a coverage range of [i, i + R - 1] (1 ≤ i ≤ n - R + 1). The detection targets in the sliding window correspond to the traffic sent within a certain second in the network measurement process, which is t x ; Since the part corresponding to M cycles at intervals of R according to the network segment it belongs to, when only considering the M part, the number of the network segment x in the sliding window is M x ; It is necessary to consider the number of occurrences of the network segment x in the rem part; Consider S x = M x T + r x (0 ≤ r x < T): If r x = 0, there will be no detection targets belonging to network segment x in rem, that is If r x ≠ 0, consider the case where x appears in the rem part. Since r x < T, there is at most one x in each rem, and there is at least one rem without x; consider the process of filling the rem. There is at least a gap of R between adjacent x's; in the rem part, each sliding window will have at most one occurrence of x, that is The number of occurrences t of network segment x in the sliding window x Meet That is, to make each network segment x reach the theoretical optimum.

Images