A dynamic user grouping method for underwater sensor networks

Abstract

The application provides a dynamic user grouping method for an underwater sensor network, through a dynamic programming method, in a case of continuous iteration, a purpose of two-user grouping of all nodes in the network is achieved, user grouping of non-orthogonal multiple access of the underwater acoustic sensor network is realized, the number of users needing to be regrouped when a new user accesses or a channel condition of an existing user seriously changes is reduced, network required overhead is reduced, and applicability of the non-orthogonal multiple access in the underwater acoustic sensor network is improved, in a case of serious change of a state of the underwater acoustic channel or new user joining, a relay node in the network continues to achieve high communication capacity through regrouping of a small number of users, and high dynamic expandability and robustness are shown. Therefore, the application improves the usability of the non-orthogonal multiple access technology in the underwater acoustic sensor network, and promotes further development of the underwater acoustic sensor network.

Description

Technical Field

[0001] This invention relates to the field of underwater acoustic sensor network technology, and improves the non-orthogonal multiple access performance of underwater acoustic sensor networks through a dynamic user grouping method. Background Technology

[0002] Underwater acoustic sensor networks are a typical application of wireless sensor networks. As an ideal medium in the marine environment, they can monitor target sea areas over a wide area in real time and have broad application prospects in fields such as marine resource exploration, marine environmental monitoring, and marine security assurance.

[0003] The unique environment of underwater acoustic channels presents numerous new and significant challenges to multiple access in underwater acoustic sensor networks. These challenges include: limited transmission bandwidth, severe multipath propagation, significant and frequency-dependent propagation loss, long and variable transmission delay, complex environmental noise, significant Doppler effect, high energy consumption, high bit error rate, and spatiotemporal variation. These characteristics of the underwater acoustic environment lead to problems such as low spectral efficiency, low energy efficiency, low user fairness, low quality of service, severe near-far effect, and long transmission delay in the multiple access of underwater acoustic sensor networks, severely degrading network performance.

[0004] Non-Orthogonal Multiple Access (NOMA) is a promising candidate for future 5G and 6G wireless communication. Its main idea is that at the transmitting end, different users occupy the same time / frequency domain resources, and multiple user signals are directly superimposed and transmitted in the power domain through superposition coding. At the receiving end, serial interference cancellation is used to eliminate interference between different users, and data from different users is detected and decoded. It can be seen that, unlike traditional orthogonal multiple access technology, NOMA multiplexes data in the power domain. NOMA can not only reach the upper limit of multi-user channel capacity, but also offers advantages over traditional orthogonal multiple access methods, such as higher spectral and energy efficiency, better user fairness, higher cell edge throughput, massive connectivity, low latency and signaling overhead, and more relaxed channel feedback requirements. Therefore, the advantages of NOMA make it highly attractive for underwater acoustic sensor networks.

[0005] In power-domain NOMA (Normally Infinite Amplitude Modulation), user grouping has a significant impact on system throughput. Maintaining certain differences in channel conditions among users within a group maximizes throughput and facilitates signal decoding. Appropriate user grouping can effectively improve NOMA system capacity. Several common existing user grouping algorithms, such as greedy grouping, suboptimal grouping, and hybrid grouping, all share the basic idea of ​​first sorting users according to channel conditions and then pairing or grouping them according to certain rules or strategies. For example, greedy user grouping selects the users with the largest channel gain differences from the remaining users for pairing until all users are paired. However, existing algorithms require regrouping all users when new users join or when the channel conditions of existing users change drastically. This results in weak dynamic expansion performance and significant system overhead when the number of users is large, impacting network performance. Summary of the Invention

[0006] To overcome the shortcomings of existing technologies, this invention provides a dynamic user grouping method for underwater sensor networks.

[0007] To address the problems caused by the spatiotemporal variation of underwater acoustic channels, such as the indeterminacy of user communication channels over large scales and the potential for significant changes in user channel conditions over a period of time, and considering the needs of underwater acoustic sensor networks in complex marine environments, this invention provides a user grouping method that does not require pre-sorting of channel gain for all users. Through a dynamic programming approach, it achieves user grouping for non-orthogonal multiple access in underwater acoustic sensor networks, further reducing the number of users requiring regrouping when new users join or when the channel conditions of existing users change drastically. This reduces network overhead and improves the applicability of non-orthogonal multiple access in underwater acoustic sensor networks.

[0008] To address the spatiotemporal uncertainties inherent in underwater acoustic environments, and the potential for new node additions or significant changes in channel conditions for existing nodes within underwater acoustic sensor networks, this paper proposes a dynamic user grouping method for underwater sensor networks. This method utilizes dynamic programming to achieve a low-complexity solution, thereby reducing network overhead. Results show that the method disclosed in this invention effectively reduces system overhead by grouping underwater acoustic sensor network nodes into users.

[0009] The purpose of this invention is to address the problem in underwater acoustic sensor networks where changes in the channel state of existing nodes or the access of new users due to underwater acoustic channel issues can affect existing packets and degrade network performance. This invention proposes a user node grouping method for underwater acoustic sensor networks.

[0010] The steps of the technical solution adopted by the present invention to solve its technical problem are as follows:

[0011] Step 1: Determine the initial user segmentation range;

[0012] Define the maximum number of users in a single group as k max The initial user partitioning interval g for different intervals is calculated, and the result is obtained by calculating the channel gain state of existing grouped users:

[0013]

[0014] Where, k≤k max -1, m and n are the number of users grouped within two adjacent grouping intervals, |h i | 2 Let |h| be the current channel gain for user i. j | 2 Let be the current channel gain for user j;

[0015] When the set of grouped users is empty, the first k... max One user is initially grouped into groups, with each group containing one user.

[0016] After the partitioning is completed, the partitioned intervals for all users are represented as follows:

[0017]

[0018] Furthermore, a minimum difference threshold of δ is defined between users. When the channel gain difference between two users is less than δ, the two users cannot be grouped into the same group.

[0019] ||h i | 2 -|h j | 2 |≤δ (3)

[0020] After the partitioning is completed, the user interval is represented as follows:

[0021]

[0022] Step 2: Determine the state transition equation and define the intra-group channel difference;

[0023] When the (i+1)th node is connected, the state transition equation is used to determine |h i+1 |2 Can the intra-group channel difference of the worst group be increased?

[0024] If the sum of channel gain differences can be increased, replace the user n recorded in the current group with the (i+1)th user, and place this user n into the ungrouped interval, then proceed to step four.

[0025] If the sum of channel gain differences cannot be increased, proceed to step three;

[0026] Step 3: Determine if the partitioned set contains the number of grouped users n. <k max Grouping, if there exists n <k max Determine if the current user meets the access requirements:

[0027] If the partitioned set does not contain the number of grouped users n <k max If a user is grouped in the correct group or does not meet the access conditions, the (i+1)th user will be placed in the ungrouped set.

[0028] Step 4: Determine if there are any users in the ungrouped set who can be grouped:

[0029] A greedy grouping algorithm is used, that is, to determine the k in the current ungrouped user set. max Do users with the largest differences in channel conditions meet the grouping condition of formula (8)?

[0030]

[0031] That is, the channel gain difference between adjacent users must be greater than the minimum grouping threshold δ, and the channel gains of the users must be in different grouping intervals;

[0032] If there exists a user group that meets the grouping criteria of formula (8), add all users in this user group as a group to the already grouped set;

[0033] Step 5: Iterate through all newly added users, repeating steps 2 to 4 until all nodes are grouped or cannot be grouped. After each new grouping is completed, the user partitioning interval g needs to be recalculated to complete the update.

[0034] Step Six: Output the completed grouped set:

[0035]

[0036] There will ultimately be i user groups, and each user group will have at most k users. max One user;

[0037] After completing user grouping for all users in the underwater acoustic sensor network, different transmit powers are allocated to all users within the same group, ultimately enabling all users in the same group to communicate simultaneously on the same time-frequency resources.

[0038] The channel differences among different users within the same group determine the upper limit of NOMA system performance. By matching new users with users who have already been grouped, the regrouping of already grouped users is avoided, reducing the additional overhead caused when network nodes change. This achieves the goal of improving the dynamic scalability of underwater acoustic sensor networks in the face of channel changes and reducing additional overhead.

[0039] The state transition equation is:

[0040] dp[i+1]=max{dp[i],(dp[i-1],|h i+1 | 2 (5)

[0041] Where dp[i] represents the sum of channel differences when i nodes are grouped together.

[0042] The sum of the intra-group channel differences is:

[0043]

[0044] The judgment |h i+1 | 2 The steps to increase the intra-group channel difference of the worst group are as follows:

[0045] Assuming the sum of channel differences among users in the m-th group is minimized, if replacing a user in the m-th group with the (i+1)-th user, determine whether the sum of channel differences in the current group increases:

[0046]

[0047] And record the grouped user n that has obtained the maximum sum of channel differences after replacement.

[0048] The access conditions are as follows:

[0049]

[0050] That is, the channel gain condition of the (i+1)th user is within the partitioned interval, and there are currently no users accessing this partitioned interval.

[0051] The beneficial effect of this invention lies in the proposed dynamic user grouping method for underwater sensor networks. In complex and ever-changing underwater environments, without requiring prioritization of channel gain for all users, it achieves the goal of grouping all nodes in the network into two-user groups through dynamic programming and iterative processes. Thus, even in the face of significant changes in the underwater acoustic channel state or the addition of new users, network relay nodes can maintain high communication capacity by regrouping a small number of users, demonstrating extremely high dynamic scalability and robustness. Therefore, this invention improves the usability of non-orthogonal multiple access technology in underwater acoustic sensor networks, promotes the further development of underwater acoustic sensor networks, and provides a technological foundation for the development of my country's integrated air-space-sea network. Attached Figure Description

[0052] Figure 1 This is a flowchart of the overall grouping method of the present invention.

[0053] Figure 2 This is a performance comparison chart of the grouping algorithm of this invention.

[0054] Figure 3 For the comparison chart of regrouped numbers, Figure 3 (a) A comparison chart of the number of new user connections regrouped. Figure 3 (b) The channel state occurs for a single user. Detailed Implementation

[0055] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0056] Taking a two-user group in an underwater acoustic sensor network as an example, the corresponding grouping implementation scheme is given:

[0057] The overall method block diagram of the present invention is as follows: Figure 1 As shown.

[0058] Step 1: Two user groups can be divided into strong and weak users. Determine the initial strong / weak user division interval, denoted as g. Calculate the channel gain state of the existing user groups:

[0059]

[0060] Among them, |h i | 2 Let m be the current channel gain for user i, and m be the number of users already grouped.

[0061] When the set of grouped users is empty, the first and second users are used as the initial completed groups.

[0062] After the partitioning is completed, the strong and weak user intervals are represented as follows:

[0063]

[0064] In addition, a minimum difference threshold is defined between users. When the channel gain difference between two users is less than a certain value, the two users cannot be grouped into the same group.

[0065] ||h i | 2 -|h j | 2 |≤δ (12)

[0066] Step 2: Determine the state transition equation:

[0067] dp[i+1]=max{dp[i],(dp[i-1],|h i+1 | 2 (13)

[0068] Where dp[i] represents the sum of channel differences when i user groups have been grouped.

[0069] Define the sum of channel differences within two user packets as:

[0070] dp[i]=∑d i =∑||h i | 2 -|h j | 2 (14)

[0071] When the (i+1)th user connects, the state transition equation is used to determine whether the intra-group channel difference of the worst group can be increased.

[0072] Assuming the sum of channel differences among users in the m-th group is the minimum, determine whether replacing a user in the m-th group with the (i+1)-th user can increase the sum of channel differences in the current group and record the grouped user that achieves the maximum value after replacement. If the sum of channel gain differences can be increased, replace the recorded user with the (i+1)-th user and place this user in the ungrouped interval, then proceed to step four.

[0073] Otherwise, proceed to step three;

[0074] Step 3: Determine if there exists a grouped user set containing only one user. If it exists, determine if the (i+1)th user meets the access conditions.

[0075]

[0076] That is, the channel gain condition of the (i+1)th user is within the defined interval;

[0077] Otherwise, add the (i+1)th user to the ungrouped set;

[0078] Step 4: Determine whether there are groupable user pairs in the unpartitioned set. Use a greedy grouping algorithm to determine whether two users with the largest difference in channel conditions meet the grouping condition.

[0079]

[0080] If a user group meets the criteria, add it to the group set.

[0081] Step 5: Iterate through all newly added users, repeating steps 2, 3, and 4 until all users are grouped or cannot be grouped further. After each new grouping, the strong / weak user classification interval needs to be updated.

[0082] Step 6: Output the completed group set G:

[0083] G=({|h 1,1 |,|h 1+2 |},{|h 2,1 |,|h 2,2 |},……,{|h i,1 |,|h 1,2 |}) (17)

[0084] After completing user grouping for all users in the underwater acoustic sensor network, different transmit powers are allocated to all users within the same group, ultimately enabling all users in the same group to communicate simultaneously on the same time-frequency resources.

[0085] Figure 2 This is a performance comparison chart of various groups when the number of users n=100.

[0086] Figure 3 (a) and Figure 3 (b) is a comparison chart of the number of users that need to be regrouped when a new user accesses the network and when the channel status of a single user changes.

Claims

1. A dynamic user grouping method for underwater sensor networks, characterized in that... Includes the following steps: Step 1: Determine the initial user segmentation range; Define the maximum number of users in a single group as k max The initial user partitioning interval g for different intervals is calculated, and the result is obtained by calculating the channel gain state of existing grouped users: Where, k≤k max -1, m and n are the number of users grouped within two adjacent grouping intervals, |h i | 2 Let |h| be the current channel gain for user i. j | 2 Let be the current channel gain for user j; When the set of grouped users is empty, the first k... max One user is initially grouped into groups, with each group containing one user. After the partitioning is completed, the partitioned intervals for all users are represented as follows: Furthermore, a minimum difference threshold of δ is defined between users. When the channel gain difference between two users is less than δ, the two users cannot be grouped into the same group. ||h i | 2 -|h j | 2 |≤δ (3) After the partitioning is completed, the user interval is represented as follows: Step 2: Determine the state transition equation and define the intra-group channel gain difference; When the (i+1)th node is connected, the state transition equation is used to determine |h i+1 | 2 Can the intra-group channel gain difference of the worst group be increased? If the sum of channel gain differences can be increased, replace the user n recorded in the current group with the (i+1)th user, and place this user n into the ungrouped interval, then proceed to step four. If the sum of channel gain differences cannot be increased, proceed to step three; Step 3: Determine if the partitioned set contains the number of grouped users n. <k max Grouping, if there exists n <k max Determine if the current user meets the access requirements: If the partitioned set does not contain the number of grouped users n <k max If a user is grouped in the correct group or does not meet the access conditions, the (i+1)th user will be placed in the ungrouped set. Step 4: Determine if there are any users in the ungrouped set who can be grouped: A greedy grouping algorithm is used, that is, to determine the k in the current ungrouped user set. max Do users with the largest differences in channel conditions meet the grouping condition of formula (8)? ||h i | 2 -|h j | 2 |>δ and|h i | 2 >g>|h j | 2 (8) That is, the channel gain difference between adjacent users must be greater than the minimum grouping threshold δ, and the channel gains of the users must be in different grouping intervals; If there exists a user group that meets the grouping criteria of formula (8), add all users in this user group as a group to the already grouped set; Step 5: Iterate through all newly added users, repeating steps 2 to 4 until all nodes are grouped or cannot be grouped. After each new grouping is completed, the user partitioning interval g needs to be recalculated to complete the update. Step Six: Output the completed grouped set: There will ultimately be i user groups, and each user group will have at most k users. max One user; After completing user grouping for all users in the underwater acoustic sensor network, different transmit powers are allocated to all users within the same group, ultimately enabling all users in the same group to communicate simultaneously on the same time-frequency resources.

2. The dynamic user grouping method for underwater sensor networks according to claim 1, characterized in that: The state transition equation is: dp[i+1]=max{dp[i],(dp[i-1],|h i+1 | 2 )} (5) Where dp[i] represents the sum of channel gain differences when i nodes are grouped together.

3. The dynamic user grouping method for underwater sensor networks according to claim 1, characterized in that: The intra-group channel gain difference is:

4. The dynamic user grouping method for underwater sensor networks according to claim 1, characterized in that: The judgment |h i+1 | 2 The steps to increase the intra-group channel gain difference of the worst group are as follows: Assuming the sum of channel gain differences among users in the m-th group is minimized, if replacing a user in the m-th group with the (i+1)-th user increases the sum of channel gain differences among the current groups, determine whether this increases the sum of channel gain differences. And record the grouped user n that has the maximum sum of channel gain differences after replacement.

5. The dynamic user grouping method for underwater sensor networks according to claim 1, characterized in that: The access conditions are as follows: That is, the channel gain condition of the (i+1)th user is within the partitioned interval, and there are currently no users accessing this partitioned interval.

Images