An asynchronous periodic cooperative neighbor discovery method based on additional scanning window
By introducing a collaborative mechanism of additional scanning windows in the asynchronous periodic neighbor discovery method, the problems of long delay and low energy efficiency in the prior art are solved, and shorter neighbor discovery time and higher energy efficiency are achieved.
Patent Information
- Application Number
- CN202211064886.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-08-31
- Publication Date
- 2025-02-21
- Estimated Expiration
- 2042-08-31
AI Technical Summary
The existing asynchronous periodic neighbor discovery methods have shortcomings in reducing the average delay, especially in the application of collaboration mechanisms, resulting in low energy efficiency.
Asynchronous periodic collaborative neighbor discovery method based on additional scanning windows is adopted, broadcast packets are sent and scan windows are entered through periodic scheduling, and shorter additional scan windows are evenly distributed between conventional scanning windows to achieve collaborative neighbor discovery.
It effectively shortens the discovery time of two-way neighbors, optimizes the discovery time of one-sided neighbors, and improves energy efficiency, which is especially suitable for environments with sparse nodes.
Smart Images
Figure CN115460577B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of wireless communications, and in particular to an asynchronous periodic cooperative neighbor discovery method based on an additional scanning window. Background Art
[0002] Neighbor discovery is the first step for IoT devices to discover each other and build a network in an ad hoc network mode. Since most IoT devices are energy-constrained, it is unrealistic to keep the transceiver working for a long time. Therefore, an asynchronous periodic neighbor discovery method is often used. The broadcast node and the scanning node only enter the working state periodically, and the transceiver is turned off at other times to save energy. For asynchronous periodic neighbor discovery, different design methods and system parameters mean different average energy consumption levels, and the corresponding neighbor discovery delays are also different. For the same energy consumption level, the neighbor discovery method with shorter delay consumes less total energy and has higher energy efficiency. Therefore, designing a neighbor discovery method with shorter delay has become the focus of research.
[0003] Among the existing asynchronous periodic neighbor discovery methods, the SingleInt method (reference: Kindt PH, Narayanaswamy S, Saur M, "Optimizing BLE-Like Neighbor Discovery," IEEE Transactions on Mobile Computing, vol. 21, no. 5, pp. 1779-1797, 2022.) minimizes the worst-case latency of symmetric neighbor discovery under a given duty cycle, but does not consider using a collaborative mechanism to further reduce latency. The Griassdi method (reference: Kindt, PH, Yunge D, Reinerth G. Chakraborty, S. Griassdi, "Mutually assisted slotless neighbor discovery," in Proc. ACM / IEEE IPSN, 2017.) is based on a collaborative mechanism, but its scanning cycle is short, and the first neighbor discovery performance is poor in the random delay broadcast mode. Summary of the invention
[0004] The object of the present invention is to provide an asynchronous periodic cooperative neighbor discovery method based on an additional scanning window, which effectively reduces the average neighbor discovery delay.
[0005] The technical solution to achieve the purpose of the present invention is: an asynchronous periodic cooperative neighbor discovery method based on an additional scanning window, using a periodic scheduling method. Each node periodically sends broadcast packets to be discovered by neighbors, periodically enters the scanning window to receive broadcast packets to discover neighbors, and uses a periodic short scanning window to achieve cooperative neighbor discovery. Except for the time in the sending state and the scanning state, the nodes are in a dormant state to save energy.
[0006] Furthermore, the periodic scheduling method is as follows:
[0007] Each node periodically sends a broadcast packet, and the broadcast packet transmission time d a is a fixed value, and the broadcast period consists of the fixed value part T l The average broadcast period is T a =T l +r / 2. Each node periodically enters the scanning state to receive broadcast packets, and the scanning period is a fixed value T s , the length of the conventional scanning window is a fixed value d s There are (k-1) shorter additional scanning windows evenly distributed between two consecutive regular scanning windows to achieve cooperative neighbor discovery. The length of the additional scanning window is a fixed value d. e =αd a (α>1). Broadcast period T a With scanning period T s Therefore, during the normal scheduled periodic operation, it may be necessary to send broadcast packets in the regular scanning window or the additional scanning window. At this time, the scanning state is temporarily interrupted and the broadcast packet is sent. After the broadcast packet is sent, it is switched back to the scanning state, and the length of the scanning window is not compensated.
[0008] Furthermore, the method for implementing collaborative neighbor discovery using an additional scanning window is specifically as follows:
[0009] Each broadcast packet sent by node A will piggyback on the time when the next additional scan window of node A arrives. If node B receives a broadcast packet sent by node A, it will send an additional cooperative broadcast packet when the next additional scan window of node A arrives, so that it can be discovered by node A. If node A receives a cooperative broadcast packet with itself as the target and from node B, it realizes that the mutual discovery with node B has been completed.
[0010] Furthermore, the relationship between the duty cycle η and the system parameters is as follows:
[0011]
[0012] Furthermore, for a given coefficient α, duty cycle η, and the total number of neighbor nodes n within a one-hop range, the values of each system parameter are determined in the following manner:
[0013] Duty cycle c for regular scan window and broadcast, average neighbor discovery delay Scan cycle T s , average broadcast period T a and the regular scan window length d s They are given by the following formulas respectively:
[0014]
[0015]
[0016]
[0017]
[0018]
[0019] Substituting formula (2) into formula (3), we can obtain Search for the minimum in the interval [2,8 / (αη)+1] Substitute k into formula (2) to get c, and use c to calculate the scanning period T s , average broadcast period T a and the regular scan window length d s .
[0020] Compared with the prior art, the present invention has the following significant advantages: (1) it optimizes the unilateral neighbor discovery time; (2) it utilizes the additional scanning window to realize collaborative neighbor discovery, effectively shortening the bidirectional neighbor discovery time; and (3) it increases the number of additional scanning windows to compensate for the adverse effects caused by multiple nodes competing for the same additional scanning window. BRIEF DESCRIPTION OF THE DRAWINGS
[0021] Figure 1 It is a schematic diagram of a successful neighbor discovery event of the method proposed in the present invention.
[0022] Figure 2 It is a comparison of the cumulative distribution of delays of the method proposed in the present invention and the existing method when the duty cycle is 0.6% in a one-to-one neighbor discovery scenario.
[0023] Figure 3 It is a comparison of the average delay of the method proposed in the present invention and the existing method in a one-to-one neighbor discovery scenario.
[0024] Figure 4It is a comparison of the average delay of the method proposed in the present invention and the existing method in the group neighbor discovery scenario. DETAILED DESCRIPTION
[0025] The present invention proposes an asynchronous periodic cooperative neighbor discovery method based on an additional scanning window, which uses an additional scanning window to achieve cooperative neighbor discovery, thereby reducing the neighbor discovery time. The specific features are as follows:
[0026] Each node uses a periodic scheduling method: periodically sending broadcast packets to be discovered by neighbors, periodically entering the scanning window to receive broadcast packets to discover neighbors, and using periodic short scanning windows to achieve collaborative neighbor discovery. Except for the time in the sending state and scanning state, the nodes are in a dormant state to save energy.
[0027] The specific implementation of the present invention is as follows:
[0028] 1. The periodic scheduling method is as follows:
[0029] Each node periodically sends a broadcast packet, and the broadcast packet transmission time d a is a fixed value, and the broadcast period consists of the fixed value part T l The average broadcast period is T a =T l +r / 2. Each node periodically enters the scanning state to receive broadcast packets, and the scanning period is a fixed value T s , the length of the conventional scanning window is a fixed value d s There are (k-1) shorter additional scanning windows evenly distributed between two consecutive regular scanning windows to achieve cooperative neighbor discovery. The length of the additional scanning window is a fixed value d. e =αd a (α>1). Broadcast period T a With scanning period T s Therefore, during the normal scheduled periodic operation, it may be necessary to send broadcast packets in the regular scanning window or the additional scanning window. At this time, the scanning state is temporarily interrupted and the broadcast packet is sent. After the broadcast packet is sent, it is switched back to the scanning state, and the length of the scanning window is not compensated.
[0030] 2. The method for implementing cooperative neighbor discovery by using an additional scanning window is as follows:
[0031] Each broadcast packet sent by node A will piggyback on the time when the next additional scan window of node A arrives. If node B receives a broadcast packet sent by node A, it will send an additional cooperative broadcast packet when the next additional scan window of node A arrives, so that node A can receive its broadcast packet within the additional scan window. If node A receives a cooperative broadcast packet with itself as the target and from node B, it realizes that mutual discovery with node B has been completed.
[0032] 3. The relationship between the duty cycle η and the system parameters is as follows:
[0033]
[0034] 4. For a given coefficient α, duty cycle η, and the total number of neighbor nodes n within one hop, the values of each system parameter are determined in the following way:
[0035] Duty cycle c for regular scan window and broadcast, average neighbor discovery delay Scan cycle T s , average broadcast period T a and the regular scan window length d s They are given by the following formulas respectively:
[0036]
[0037]
[0038]
[0039]
[0040]
[0041] Substituting formula (2) into formula (3), we can obtain Search for the minimum in the interval [2,8 / (αη)+1] Substitute k into formula (2) to get c, and use c to calculate the scanning period T s , average broadcast period T a and the regular scan window length d s .
[0042] Example 1
[0043] The first embodiment of the present invention is described in detail as follows. The network simulation is run on the OMNeT++ platform, using the C++ language, and the parameter settings do not affect the generality. In the simulation, the broadcast packet transmission time is set to d a =3.76×10 -4s, the ratio of the additional scanning window to the broadcast packet transmission time is set to α = 1.5. The random part r of the broadcast cycle is randomly selected in the interval [0, 10ms] each time. This setting is consistent with the random delay broadcast mode of low-power Bluetooth. The duty cycle η is selected between 0.6% and 1.2%. The starting time of each simulation is considered to be the moment when the nodes enter the communication range of each other. The first broadcast time of each node is in the interval [0, T a ] is randomly selected, and the first scanning time is in the interval [0, T s ] are randomly selected. For each duty cycle, 2000 samples are simulated and the average value is calculated.
[0044] Figure 2 This is a comparison chart of the cumulative distribution results of neighbor discovery delays of the method proposed in the present invention, the SingleInt method, and the Griassdi method when the duty cycle is 0.6% in the one-to-one neighbor discovery scenario in Example 1. It can be seen from the figure that the proposed method is superior to the SingleInt method and the Griassdi method in terms of both average delay and two-way discovery delay.
[0045] Figure 3 This is a comparison chart of the average delay results of the method proposed in the present invention, the SingleInt method and the Griassdi method in the one-to-one neighbor discovery scenario in Implementation 1. It can be seen from the figure that for different duty cycles, the proposed method is better than the SingleInt method and the Griassdi method, and the average delay can be reduced by about 26% compared with the SingleInt method. It can be seen that the proposed method can effectively reduce the neighbor discovery delay through the collaborative neighbor discovery mechanism.
[0046] Figure 4 It is a comparison chart of the average delay results of the method proposed in the present invention, the SingleInt method and the Gríassdi method in the group neighbor discovery scenario in Implementation 1. It can be seen from the figure that as the number of one-hop neighbor nodes increases, the average delay of each method will increase, but the proposed method is more affected. When the number of nodes is 4, 8, 12 and 16, the average delay of the proposed method is about 25%, 18%, 13% and 5% lower than that of the SingleInt method, and the advantage of the proposed method gradually decreases. It can be seen that the proposed method is mainly suitable for environments with sparse nodes.
Claims
1. An asynchronous periodic cooperative neighbor discovery method based on an additional scanning window, characterized in that: Each node uses a periodic scheduling method: it periodically sends broadcast packets to be discovered by neighbors, periodically enters the scanning window to receive broadcast packets to discover neighbors, and uses periodic short scanning windows to achieve collaborative neighbor discovery; except for the time in the sending state and scanning state, the nodes are in a dormant state to save energy; The periodic scheduling method is as follows: Each node periodically sends a broadcast packet, and the broadcast packet transmission time d a is a fixed value, and the broadcast period is composed of the fixed value part T l The average broadcast period is T a =T l +r / 2; Each node periodically enters the scanning state to receive broadcast packets, and the scanning period is a fixed value T s , the length of the conventional scanning window is a fixed value d s ; k-1 shorter additional scanning windows are evenly distributed between two consecutive regular scanning windows to achieve collaborative neighbor discovery, and the length of the additional scanning window is a fixed value d e =αd a , where α>1; average broadcast period T a With scanning period T s Different, if a broadcast packet needs to be sent within the scanning window during the normal scheduled periodic operation, the scanning state is temporarily interrupted and the broadcast packet is sent. After the broadcast packet is sent, it is switched back to the scanning state, and the length of the scanning window is not compensated.
2. The asynchronous periodic cooperative neighbor discovery method based on additional scanning window according to claim 1, characterized in that: The method of implementing cooperative neighbor discovery using an additional scanning window is as follows: Each broadcast packet sent by node A will carry the arrival time of node A's next additional scanning window. If node B receives a broadcast packet sent by node A, it will send an additional cooperative broadcast packet when node A's next additional scanning window arrives, so that it can be discovered by node A. If node A receives a cooperative broadcast packet with itself as the target and from node B, it realizes that the mutual discovery with node B has been completed.
3. The asynchronous periodic cooperative neighbor discovery method based on additional scanning window according to claim 2, characterized in that: The neighbor discovery energy consumption of a node is expressed by the duty cycle η, which is defined as the ratio of the time in working state to the total time. The relationship between the duty cycle η and the system parameters is obtained.
4. The method for asynchronous periodic cooperative neighbor discovery based on additional scanning window according to claim 3, characterized in that: For a given coefficient α, duty cycle η, and the total number of neighbor nodes n within one hop, the values of each system parameter are determined in the following way: Duty cycle c for regular scan window and broadcast, average neighbor discovery delay Scan cycle T s , average broadcast period T a and the regular scan window length d s They are given by the following formulas respectively: Substituting formula (2) into formula (3), we can obtain Search for the minimum in the interval [2,8 / (αη)+1] Substitute k into formula (2) to get c, and use c to calculate the scanning period T s , average broadcast period T a and the regular scan window length d s .