Cross-layer optimization routing protocol method and system for high-dynamic ad hoc network scene

Abstract

The invention discloses a cross-layer optimization routing protocol method and system for a high-dynamic ad hoc network scene. The method specifically comprises the following steps: firstly, establishing a cross-layer optimization model for a high-dynamic ad hoc network; then, physical layer link quality is calculated, and link quality evaluation is carried out in a windowed exponentially weighted moving average smoothing mode; counting the load condition of the MAC layer based on Kalman filtering, and carrying out updating interaction through a network maintenance message; and finally, constructing a comprehensive evaluation model, and realizing MPR selection based on a greedy algorithm for joint optimization of cross-layer resources. The system comprises a cross-layer optimization model construction module, a link quality evaluation module, an update interaction module and an MPR selection module. According to the invention, the problem of low network stability in a multi-node dynamic topology environment is solved, the end-to-end delay is reduced, the packet delivery success rate is improved, the routing overhead is reduced, and the adaptability to the dynamic environment is enhanced.

Description

Technical Field

[0001] This invention relates to the field of data link communication technology, and in particular to a cross-layer optimized routing protocol method and system for highly dynamic ad hoc network scenarios. Background Technology

[0002] Mobile Ad Hoc Network (MANET) is a wireless network that combines mobile communication and computer technologies. It requires no fixed infrastructure and achieves dynamic networking through a distributed structure. Its nodes function as both hosts and routers, employ multi-hop communication mechanisms to support free movement, and possess self-configuration, self-optimization, and self-healing capabilities. Its core features include dynamic topology, resource constraints, and distributed control. It is a communication network without a central control structure, primarily used in wireless environments.

[0003] MANETs, ​​by leveraging the characteristic of not requiring fixed infrastructure, allow nodes to autonomously establish networks, overcoming the difficulties faced by traditional wireless cellular networks due to geographical limitations. They enable rapid, convenient, and efficient deployment in a short time and are often used in tactical networks. For MANETs, ​​the research on routing protocols significantly impacts the network's outcome and quality. Currently, widely used routing protocols mainly fall into two categories: distance vector routing protocols and link-state routing protocols. Distance vector protocols include the Enhanced Interior Gateway Routing Protocol (EIGRP) and the Ad-hoc On-Demand Distance Vector (AODV). Link-state routing protocols include Optimized Link State Routing (OLSR) and Open Shortest Path First (OSPF). Research shows that OLSR generally outperforms OSPF-MDR in MANET environments, demonstrating its broader application prospects in large-scale wireless ad hoc network topologies.

[0004] OLSR is a table-driven, priori routing protocol. It achieves link listening and neighbor discovery through the exchange of HELLO handshake messages in a specific format between network nodes. After collecting basic link and neighbor information, it uses the MPR (Multi-Point Relay) mechanism to broadcast topology messages across the entire network, enabling the convergence of link information within the network. Finally, it uses the shortest path algorithm to generate the overall network routing information. The OLSR routing protocol updates routing status by periodically maintaining entries in neighbor tables, link tables, and topology tables.

[0005] As the application areas of MANET continue to expand, networks in various application scenarios face more challenges. Particularly in the military field, the stability, transmission latency, and bandwidth utilization of highly dynamic networks all need improvement. Because high-speed moving nodes frequently update their link states, single-path routing is prone to communication interruptions due to link disconnections, affecting network reliability. Traditional OLSR's MPR mechanism only focuses on the number of two-hop neighbors. If some low-quality links in the network are selected as MPRs, it will significantly reduce resource utilization and affect network performance. Furthermore, if the link load of the MPR node selected by this algorithm is high, it will lead to node congestion, resulting in a significant decrease in packet delivery success rate and increased latency.

[0006] To address the aforementioned issues, load balancing mechanisms have become an effective solution. However, maintaining various relevant metrics and accurate route calculations requires additional transmission overhead, storage, and computing resources, necessitating careful consideration for resource-constrained application platforms. Furthermore, different scenarios present varying requirements for network communication performance. Reducing end-to-end latency, route convergence time, and routing overhead, while improving packet delivery success rate in highly dynamic networking environments, has become a core challenge for current routing technologies.

[0007] The technical problem to be solved by this invention is: for highly dynamic self-organizing network scenarios, a cross-layer optimized dynamic routing protocol is proposed to solve the problems of low network stability in multi-node dynamic topology environments and MPR set optimization based on greedy algorithm, reduce routing overhead, effectively improve the data packet delivery success rate and reduce data packet latency. Summary of the Invention

[0008] The purpose of this invention is to provide a cross-layer optimized routing protocol method and system for highly dynamic ad hoc network scenarios, characterized by low routing overhead, high packet delivery success rate, and low packet latency.

[0009] The technical solution to achieve the purpose of this invention is: a cross-layer optimized routing protocol method for highly dynamic ad hoc network scenarios, comprising the following steps:

[0010] Step 1: Establish a cross-layer optimization model for highly dynamic self-organizing networks;

[0011] Step 2: Calculate the physical layer link quality Link quality is assessed using a windowed exponentially weighted moving average smoothing method.

[0012] Step 3: Statistically analyze the MAC layer load based on Kalman filter. And update and interact through network maintenance messages;

[0013] Step 4: Construct a comprehensive evaluation model to achieve cross-layer resource joint optimization based on greedy algorithm for MPR selection.

[0014] Furthermore, the establishment of the cross-layer optimization model for highly dynamic self-organizing networks described in step 1 is as follows:

[0015] Suppose there are K highly dynamic nodes in the network space, and each node collects the MAC layer load from the other nodes, denoted as K. Each node evaluates the physical layer link quality with the other nodes. The two-hop neighbor coverage of the MPR solution for each node is a set. The cross-layer optimization model for routing is then expressed as:

[0016]

[0017] in, The MPR selection is determined by the node energy corresponding to N one-hop neighbors. ; They are respectively The weighting coefficients.

[0018] Furthermore, the calculation of physical layer link quality described in step 2... Link quality is assessed using a windowed exponentially weighted moving average smoothing method, as detailed below:

[0019] Step 2.1: Calculate signal power :

[0020] Calculate the signal power received by the receiving antenna using Friis's transmission formula. :

[0021]

[0022] in, The transmitter's output power, measured in watts (W). This is the gain of the RF antenna, measured in dBi. This refers to the receiver antenna gain, expressed in dBi. The operating wavelength is in meters (m). , It's the speed of light. , It is frequency, and the unit is... ; The distance between the transmitter and receiver is expressed in meters (m). Free space path loss (FSPL) characterizes the attenuation of a signal during propagation.

[0023] Equivalent Isotropic Radiation Power It is part of the formula:

[0024]

[0025] The Friis formula measures the equivalent radiated power of a transmitting system in the direction of maximum radiation; therefore, it can be written as:

[0026]

[0027] Step 2.2: Calculate noise power :

[0028] If the total noise power at the receiver is expressed in terms of noise temperature, then the total noise power of the receiving system is... for:

[0029]

[0030] in, Boltzmann constant:

[0031] J / K

[0032] The system noise temperature, in Kelvin (K), includes external noise received by the antenna. and noise generated inside the receiver ,Right now:

[0033]

[0034] Noise generated inside the receiver The noise figure can be converted to noise temperature using the following formula:

[0035]

[0036] in This is a reference temperature. It is the noise figure;

[0037] The bandwidth of the receiving system, in units of That is, the channel bandwidth currently in use;

[0038] Step 2.3: Calculate the final signal-to-noise ratio (SNR):

[0039] Divide the received signal power by the noise power to obtain the final signal-to-noise ratio (SNR):

[0040]

[0041]

[0042] Expressed in decibels (dB):

[0043]

[0044] Step 2.4: Due to the time delay in the propagation of link quality, a weighted smoothing method is used to estimate the physical layer link quality based on the signal sequence containing noise. The smoothed sequence is obtained. The details are as follows:

[0045] First, for each data point Assign a weight This weight comes from the signal-to-noise ratio at that point. :

[0046] High SNR corresponds to high weight: if If it is considered a reliable signal, then Close to 1;

[0047] Low SNR corresponds to low weight: if If it is considered an unreliable signal, then Approaching 0;

[0048] Then, the SNR is converted into weights using a monotonic function, as shown in the formula:

[0049] or

[0050] Finally, using these weights for weighted smoothing, we obtain the physical layer link quality. :

[0051]

[0052] Furthermore, step 3 involves statistically analyzing the MAC layer load using Kalman filtering. It also updates and interacts through network maintenance messages, as detailed below:

[0053] Step 3.1, Define state variables:

[0054] node During the periodic transmission of handshake messages, the length of the PDU queue to be sent on the MAC layer interface is recorded and sampled. After multiple equally spaced recording and sampling, a time series set of load observations is obtained. ;set up Represents a node No. The observed value of the PDU queue length of the MAC interface when sending the Hello message for the first time. Indicates the first The rate of change of the secondary load, then the load The function is:

[0055]

[0056]

[0057] Step 3.2: Establish the state transition model:

[0058] State transition models describe how a state changes from one state to another. Evolved to Assuming the load change is approximately uniform, but there is an unpredictable process noise interference, then:

[0059]

[0060] Wherein, the state transition matrix for:

[0061]

[0062] in, It is the sampling time interval, matrix express:

[0063] New load = Old load + Rate of change * Time interval, the formula is:

[0064]

[0065] The state transition model is a uniform velocity model, meaning the new velocity equals the old velocity, and the formula is:

[0066]

[0067] Process noise It is a noise with a mean of 0 and a covariance matrix of Q; Q represents the abnormal expenditure of the sub-model, and the size of Q is the relevance confidence of this model. The larger Q is, the more the filter depends on the measured value; the smaller Q is, the more it depends on the predicted value.

[0068] Step 3.3: Establish the observation model:

[0069] The observation model reflects how to observe from the state Obtain the observation value That is, the actual measured load value, the formula is:

[0070]

[0071] in For the observation matrix:

[0072]

[0073] matrix This indicates that the load can only be directly observed. However, its rate of change cannot be directly observed;

[0074] The observation noise is a Gaussian white noise with a mean of 0 and a variance of R;

[0075] Step 3.4: Perform load forecasting and load updates.

[0076] initialization:

[0077]

[0078] based on The state at any given moment, prediction The state at any given moment:

[0079]

[0080] Update prediction error covariance:

[0081]

[0082] Get a pair Rough prediction of moment load and speed and its uncertainty ;

[0083] Step 3.5: Perform corrected estimation:

[0084] Actual measurements were performed to obtain the observed values ​​of the load. ;

[0085] The Kalman gain is calculated using the following formula:

[0086]

[0087] Kalman coefficient characterizes the new measurement value The degree of credibility;

[0088] Correcting the predicted values ​​with the measured values ​​yields the optimal posterior estimate:

[0089]

[0090] in, Known as a news item, it is the difference between a predicted value and a measured value.

[0091] Update the covariance of the posterior estimation error:

[0092]

[0093] Furthermore, the construction of the comprehensive evaluation model described in step 4, which realizes cross-layer resource joint optimization based on the greedy algorithm for MPR selection, is as follows:

[0094] Step 4.1: Traverse the set of one-hop neighbors of this node. and two neighbor sets Query the link quality, load status and node reachability of one-hop neighbors, and then record the node energy calculated according to the cross-layer optimization model in the MRP candidate neighbor table;

[0095] Step 4.2, Traverse all If a two-hop neighbor node exists, is there a uniquely reachable one-hop neighbor node? If so, add it to the MPR selection table.

[0096] Step 4.3: Query existing MPR nodes, from The two-hop neighbors covered by the MPR node are deleted in the process.

[0097] Step 4.4, Query Check if it is an empty set; if not, continue iterating. Based on the node energy calculated by the cross-layer optimization model, select the one-hop neighbor node with the highest energy and add it to the MPR selection table;

[0098] Step 4.5: Query existing MPR nodes, and then from... The two-hop neighbors covered by the MPR node are deleted in the process.

[0099] Step 4.6, repeat steps 4.4 and 4.5 until... Empty.

[0100] A cross-layer optimized routing protocol system for highly dynamic ad hoc network scenarios is disclosed. This system implements the aforementioned cross-layer optimized routing protocol method for highly dynamic ad hoc network scenarios. The system includes a cross-layer optimization model construction module, a link quality assessment module, an update interaction module, and an MPR selection module, wherein:

[0101] A cross-layer optimization model construction module is used to establish a cross-layer optimization model for highly dynamic self-organizing networks.

[0102] The link quality assessment module calculates the physical layer link quality. Link quality is assessed using a windowed exponentially weighted moving average smoothing method.

[0103] Update the interaction module to statistically analyze the MAC layer load based on a smoothing algorithm. And update and interact through network maintenance messages;

[0104] The MPR selection module constructs a comprehensive evaluation model to achieve cross-layer resource joint optimization based on a greedy algorithm for MPR selection.

[0105] A mobile terminal includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the cross-layer optimized routing protocol method for highly dynamic ad hoc network scenarios.

[0106] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the cross-layer optimized routing protocol method for highly dynamic ad hoc network scenarios.

[0107] A computer device includes a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the cross-layer optimized routing protocol method for highly dynamic ad hoc network scenarios.

[0108] A computer program product includes computer instructions for causing a computer to execute the cross-layer optimized routing protocol method for highly dynamic ad hoc network scenarios.

[0109] Compared with the prior art, the present invention has the following significant advantages: (1) It proposes a cross-layer optimized dynamic routing protocol for highly dynamic self-organizing network scenarios, which solves the problem of low network stability in multi-node dynamic topology environments; (2) It adopts link state detection and load prediction, which effectively avoids low-quality links and congested nodes, reduces end-to-end latency, and improves packet delivery success rate; (3) Based on greedy algorithm-based MPR set optimization, it reduces the redundancy of MPR nodes, reduces the repeated sending of topology messages, and reduces routing overhead; (4) It retains physical signal characteristics and combines the high correlation parameters of the physical layer and link layer to enhance adaptability to dynamic environments. Attached Figure Description

[0110] Figure 1 This is a flowchart illustrating a cross-layer optimized routing protocol method for highly dynamic ad hoc network scenarios according to the present invention.

[0111] Figure 2 This is a schematic diagram of a simulation scenario in an embodiment of the present invention.

[0112] Figure 3 The figure shows the simulation results of the end-to-end delay in an embodiment of the present invention.

[0113] Figure 4 The figure shows the simulation results of the routing convergence time in an embodiment of the present invention.

[0114] Figure 5 This is a simulation result diagram of the packet delivery success rate in an embodiment of the present invention.

[0115] Figure 6This is a simulation result diagram of routing overhead in an embodiment of the present invention. Detailed Implementation

[0116] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0117] Combination Figure 1 This invention provides a method for a cross-layer optimized dynamic routing protocol (CR-LODP) for highly dynamic ad hoc network scenarios, comprising the following steps:

[0118] Step 1: Establish a cross-layer optimization model for highly dynamic self-organizing networks;

[0119] Step 2: Calculate the physical layer link quality Link quality is assessed using a windowed exponentially weighted moving average smoothing method.

[0120] Step 3: Statistical analysis of MAC layer load based on Kalman filtering And update and interact through network maintenance messages;

[0121] Step 4: Construct a comprehensive evaluation model to realize MPR selection based on a greedy algorithm for cross-layer resource joint optimization.

[0122] As a specific example, step 1, which establishes a cross-layer optimization model for highly dynamic ad hoc networks, is as follows:

[0123] Assuming there are K highly dynamic moving nodes in the space, the MAC layer load collected from each node is... -1, assess the physical layer link quality of each node. The two-hop neighbor coverage of its MPR solution is set as follows: Assume that the cross-layer optimization model of routing can be expressed as:

[0124]

[0125] in, The energy of the nodes corresponding to N one-hop neighbors determines the MPR selection. , These are the weighting coefficients.

[0126] As a specific example, step 2 involves calculating the physical layer link quality. Link quality is assessed using a windowed exponentially weighted moving average smoothing method, as detailed below:

[0127] The analysis of the entire link follows the path of the signal from the transmitter to the receiver. The signal power and noise power are calculated step by step below.

[0128] 1. Calculation of signal power (P_rx_signal)

[0129] Signal power received by the receiving antenna Given by Friis's transmission formula:

[0130]

[0131] in:

[0132] : Transmitter output power (W);

[0133] Transmit antenna gain (dimensionless, usually expressed in dBi);

[0134] Receiver antenna gain (dimensionless, usually expressed in dBi);

[0135] Operating wavelength (m) c is the speed of light (3 × 10⁻⁶) 8 m / s), f It is frequency ( );

[0136] R: Distance between the transmitter and receiver (m);

[0137] Free space path loss (FSPL) is a very important term that characterizes the attenuation of a signal during propagation.

[0138] The equivalent isotropic radiated power (EIRP) is part of the formula:

[0139]

[0140] EIRP measures the equivalent radiated power of a transmitting system in the direction of maximum radiation. Therefore, Friis's formula can also be written as:

[0141]

[0142] 2. Calculation of noise power

[0143] The total noise power at the receiver end is usually expressed using a very important concept—noise temperature (T). Total noise power of the receiving system for:

[0144]

[0145] in:

[0146] k: Boltzmann constant; J / K;

[0147] System noise temperature (K);

[0148] It includes external noise received by the antenna (such as cosmic noise and atmospheric noise) and noise generated inside the receiver (thermal noise from amplifiers, mixers, etc.).

[0149] The noise performance of a receiver is also commonly represented by the noise figure (NF), which can be converted into noise temperature. ,in =290K is the reference temperature.

[0150] Total system noise temperature .

[0151] B: Bandwidth of the receiving system ( This refers to the channel bandwidth you are using.

[0152] 3. Final Signal-to-Noise Ratio (SNR) Calculation

[0153] Dividing the received signal power by the noise power yields the final SNR:

[0154]

[0155]

[0156] Expressed in decibels (dB):

[0157] When calculating the link budget, it is most convenient to convert all items to decibels before performing addition and subtraction operations.

[0158]

[0159] Since the propagation of link quality has a certain time delay, a weighted smoothing method is used to estimate the physical layer link quality in order to reduce errors.

[0160] Suppose there is a signal sequence containing noise. We hope to obtain a smoothed sequence. .

[0161] First, it is necessary to consider each data point Assign a weight This weight is derived from the signal-to-noise ratio at that point. .

[0162] High SNR → High weight: If If it is considered a reliable "signal", then Close to 1.

[0163] Low SNR → Low weight: If If it is considered a reliable "signal", then Close to 0

[0164] A common method for converting SNR into weights is to use a monotonic function, for example:

[0165] or

[0166] We now use these weights to perform weighted smoothing to obtain the physical layer link quality. :

[0167]

[0168] As a specific example, step 3 describes the statistical analysis of the MAC layer load based on Kalman filtering. Specifically:

[0169] When performing communication assurance tasks, highly dynamic terminal networks need to transmit multiple large-volume data streams, leading to increased load and even congestion on some node links. This results in packet loss and increased end-to-end latency, ultimately impacting the overall communication performance of the UAV network. This invention proposes a Kalman filter-based approach to statistically analyze MAC layer load for optimized routing. A state-space model is then developed to model the link load.

[0170] ① Definition of state variables

[0171] node When periodically sending handshake messages, the length of the PDU queue to be sent on the MAC layer interface is recorded and sampled. After multiple equally spaced recording and sampling, a time series set of load observations can be obtained. .set up Represents a node The observed length of the PDU queue to be sent on the MAC interface during the k-th Hello message transmission. This represents the rate of change of the load in the k-th iteration; the load... The function is:

[0172]

[0173]

[0174] ② State transition model (dynamic model)

[0175] This model describes how states change The evolution leads to the assumption that the load change is roughly uniform, but there is an unpredictable disturbance (process noise).

[0176]

[0177] State transition matrix F:

[0178]

[0179] This is the sampling time interval, represented by this matrix:

[0180] New load = Old load + Rate of change * Time interval )

[0181] New speed = old speed ( (This is assumed to be a uniform velocity model.)

[0182] Process noise : is a variable with a mean of 0 and a covariance matrix of Q. Q represents the abnormal expenditure of this model, such as a sudden and significant change in business traffic; the size of Q is the relevance confidence of this model; the larger Q is, the more the filter depends on the measured value; the smaller Q is, the more it depends on the predicted value.

[0183] ③ Observation model

[0184] The observation model reflects how to observe from the state Only the observations were obtained (i.e., the actual measured load value)

[0185]

[0186] :

[0187]

[0188] This matrix indicates that only the load can be directly observed. However, its rate of change cannot be directly observed.

[0189] Observation noise : is a Gaussian white noise with a mean of 0 and a variance of R. It represents measurement error (such as the inaccuracy of SNMP sampling). The larger the R value, the less reliable the measurement.

[0190] Next, load forecasting and load updates will be performed:

[0191] initialization:

[0192]

[0193] (The initial value can be set roughly, and the filter will converge quickly.)

[0194] Based on the state at time k-1, predict the state at time k:

[0195]

[0196] Update of prediction error covariance:

[0197]

[0198] Now, we have a "coarse" prediction of the load and speed at time k. and its uncertainty .

[0199] Next, we will make a revised estimate:

[0200] Actual measurements were performed to obtain the observed values ​​of the load. .

[0201] Calculate Kalman gain

[0202]

[0203] Kalman coefficients are like "weights". To what extent should the new measurements be trusted? Instead of predicting .

[0204] Correcting the predicted values ​​with the measured values ​​yields the optimal posterior estimate:

[0205]

[0206] Items in parentheses Known as a news item, it is the difference between a predicted value and a measured value.

[0207] Update the covariance of the posterior estimation error:

[0208]

[0209] As a specific example, step 4 describes the construction of a comprehensive evaluation model, which realizes MPR selection based on a greedy algorithm for cross-layer resource joint optimization, as detailed below:

[0210] Multi-point relay (MPR) is a key mechanism in OLSR routing. Through MPR, the topology message is broadcast across the entire network. MPR nodes spread the received topology message to their neighboring nodes. Other nodes not selected as MPR nodes do not need to forward the message, reducing the excessive broadcasting of topology messages in the network, improving the efficiency of routing information, and reducing the routing overhead of the entire network. Topology message forwarding is performed by designated routing nodes on each communication link, without all nodes participating in flooding, thereby reducing the broadcast overhead of routing information. Ultimately, the link information within the network converges, and then the shortest path algorithm is used to generate the routing information for the entire network.

[0211] OLSR routing uses a greedy mechanism to filter MPR nodes:

[0212] ① Traverse the one-hop neighbor set of this node and two neighbor sets The system queries the link quality, load status, and node reachability of one-hop neighbors, and then calculates the node energy according to the cross-layer optimization model, recording it in the MRP candidate neighbor table.

[0213] ② Traverse all For each two-hop neighbor node in the MPR selection table, if there exists a uniquely reachable one-hop neighbor node, add it to the MPR selection table.

[0214] ③ Query existing MPR nodes, from which The two-hop neighbors covered by it are deleted.

[0215] ④Query Is it an empty set? If not, continue iterating. Based on the node energy calculated by the cross-layer optimization model, select the one-hop neighbor node with the highest energy and add it to the MPR selection table.

[0216] ⑤ Query existing MPR nodes and then... The two-hop neighbors covered by it are deleted.

[0217] ⑥ Repeat steps ④ and ⑤ until... Empty.

[0218] This invention also provides a cross-layer optimized routing protocol system for highly dynamic ad hoc network scenarios. This system implements the aforementioned cross-layer optimized routing protocol method for highly dynamic ad hoc network scenarios. The system includes a cross-layer optimization model construction module, a link quality assessment module, an update interaction module, and an MPR selection module, wherein:

[0219] A cross-layer optimization model construction module is used to establish a cross-layer optimization model for highly dynamic self-organizing networks.

[0220] The link quality assessment module calculates the physical layer link quality. Link quality is assessed using a windowed exponentially weighted moving average smoothing method.

[0221] Update the interaction module to statistically analyze the MAC layer load based on a smoothing algorithm. And update and interact through network maintenance messages;

[0222] The MPR selection module constructs a comprehensive evaluation model to achieve cross-layer resource joint optimization based on a greedy algorithm for MPR selection.

[0223] The present invention also provides a mobile terminal, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the cross-layer optimized routing protocol method for highly dynamic ad hoc network scenarios.

[0224] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the steps in the cross-layer optimized routing protocol method for highly dynamic ad hoc network scenarios.

[0225] The present invention also provides a computer device, including: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the cross-layer optimized routing protocol method for highly dynamic ad hoc network scenarios.

[0226] The present invention also provides a computer program product, including computer instructions, which are used to cause a computer to execute the cross-layer optimized routing protocol method for highly dynamic ad hoc network scenarios.

[0227] Example

[0228] This embodiment uses OPNET network simulation to verify the effectiveness and parameter estimation performance of the cross-layer optimized routing protocol method for highly dynamic ad hoc network scenarios proposed in this invention.

[0229] The simulation scenario in this embodiment is as follows: Figure 2 As shown, a 1*16 node network is constructed, with each node moving at a speed of 1440km / h. Figure 3 The figure shows the simulation results of the end-to-end delay. Figure 4 The graph shows the simulation results for the routing convergence time. Figure 5 The graph shows the simulation results for package delivery success rate. Figure 6 The simulation results for routing overhead are shown in the figure.

[0230] Simulation results show that, compared to the traditional OLSR algorithm, the CR-LODP algorithm adds link state detection and load prediction, effectively avoiding low-quality links and congested nodes, significantly reducing end-to-end latency, and improving packet delivery success rate by about 10%. Since the CR-LODP algorithm needs to exchange load information in the HELLO message, it can reduce the redundancy of MPR nodes and reduce the repeated transmission of topology messages, thus the routing overhead is slightly lower than that of the traditional OLSR algorithm. In the efficient data exchange process, high-dynamic networking proceeds smoothly, and the routing convergence time is reduced by 15%. Therefore, it can be concluded that the CR-LODP algorithm preserves physical signal characteristics and, combined with highly correlated parameters at the physical and link layers, enhances adaptability to dynamic environments and effectively solves the redundancy problem of the traditional MPR algorithm.

[0231] The above are merely preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A cross-layer optimized routing protocol method for highly dynamic ad hoc network scenarios, characterized in that, Includes the following steps: Step 1: Establish a cross-layer optimization model for highly dynamic self-organizing networks; Step 2: Calculate the physical layer link quality Link quality is assessed using a windowed exponentially weighted moving average smoothing method. Step 3: Statistical analysis of MAC layer load based on Kalman filtering And update and interact through network maintenance messages; Step 4: Construct a comprehensive evaluation model to achieve cross-layer resource joint optimization based on greedy algorithm for MPR selection.

2. The cross-layer optimized routing protocol method for highly dynamic ad hoc network scenarios according to claim 1, characterized in that, Step 1, which involves establishing a cross-layer optimization model for highly dynamic ad hoc networks, is detailed below: Suppose there are K highly dynamic nodes in the network space, and each node collects the MAC layer load from the other nodes, denoted as K. Each node evaluates the physical layer link quality with the other nodes. The two-hop neighbor coverage of the MPR solution for each node is a set. The cross-layer optimization model for routing is then expressed as: ； in, The MPR selection is determined by the node energy corresponding to N one-hop neighbors. ; They are respectively The weighting coefficients.

3. The cross-layer optimized routing protocol method for highly dynamic ad hoc network scenarios according to claim 2, characterized in that, Step 2 describes the calculation of physical layer link quality. Link quality is assessed using a windowed exponentially weighted moving average smoothing method, as detailed below: Step 2.1: Calculate signal power : Calculate the signal power received by the receiving antenna using Friis's transmission formula. : ； in, The transmitter's output power, measured in watts (W). This is the gain of the RF antenna, measured in dBi. This refers to the receiver antenna gain, expressed in dBi. The operating wavelength is in meters (m). , It's the speed of light. , It is frequency, and the unit is... ; The distance between the transmitter and receiver is expressed in meters (m). Free space path loss (FSPL) characterizes the attenuation of a signal during propagation. Equivalent Isotropic Radiation Power It is part of the formula: ； The Friis formula measures the equivalent radiated power of a transmitting system in the direction of maximum radiation; therefore, it can be written as: ； Step 2.2: Calculate noise power : If the total noise power at the receiver is expressed in terms of noise temperature, then the total noise power of the receiving system is... for: ； in, Boltzmann constant: J / K; The system noise temperature, in Kelvin (K), includes external noise received by the antenna. and noise generated inside the receiver ,Right now: ； Noise generated inside the receiver The noise figure can be converted to noise temperature using the following formula: ； in This is a reference temperature. It is the noise figure; The bandwidth of the receiving system, in units of That is, the channel bandwidth currently in use; Step 2.3: Calculate the final signal-to-noise ratio (SNR): Divide the received signal power by the noise power to obtain the final signal-to-noise ratio (SNR): ； ； Expressed in decibels (dB): ； Step 2.4: Due to the time delay in the propagation of link quality, a weighted smoothing method is used to estimate the physical layer link quality based on the signal sequence containing noise. The smoothed sequence is obtained. The details are as follows: First, for each data point Assign a weight This weight comes from the signal-to-noise ratio at that point. : High SNR corresponds to high weight: if If it is considered a reliable signal, then Close to 1; Low SNR corresponds to low weight: if If it is considered an unreliable signal, then Approaching 0; Then, the SNR is converted into weights using a monotonic function, as shown in the formula: or ; Finally, using these weights for weighted smoothing, we obtain the physical layer link quality. : 。 4. The cross-layer optimized routing protocol method for highly dynamic ad hoc network scenarios according to claim 3, characterized in that, Step 3 describes the statistical analysis of the MAC layer load based on Kalman filtering. It also updates and interacts through network maintenance messages, as detailed below: Step 3.1, Define state variables: node During the periodic transmission of handshake messages, the length of the PDU queue to be sent on the MAC layer interface is recorded and sampled. After multiple equally spaced recording and sampling, a time series set of load observations is obtained. ;set up Represents a node No. The observed value of the PDU queue length of the MAC interface when sending the Hello message for the first time. Indicates the first The rate of change of the secondary load, then the load The function is: ； ； Step 3.2: Establish the state transition model: State transition models describe how a state changes from one state to another. Evolved to Assuming the load change is approximately uniform, but there is an unpredictable process noise interference, then: ； Wherein, the state transition matrix for: ； in, It is the sampling time interval, matrix express: New load = Old load + Rate of change * Time interval, the formula is: ； The state transition model is a uniform velocity model, meaning the new velocity equals the old velocity, and the formula is: ； Process noise It is a noise with a mean of 0 and a covariance matrix of Q; Q represents the abnormal expenditure of the sub-model, and the size of Q is the relevance confidence of this model. The larger Q is, the more the filter depends on the measured value; the smaller Q is, the more it depends on the predicted value. Step 3.3: Establish the observation model: The observation model reflects how to observe from the state Obtain the observation values That is, the actual measured load value, the formula is: ； in For the observation matrix: ； matrix This indicates that the load can only be directly observed. However, its rate of change cannot be directly observed; The observation noise is a Gaussian white noise with a mean of 0 and a variance of R; Step 3.4: Perform load forecasting and load updates. initialization: ； based on The state at any given moment, prediction The state at any given moment: ； Update prediction error covariance: ； Get a pair Rough prediction of moment load and speed and its uncertainty ; Step 3.5: Perform corrected estimation: Actual measurements were performed to obtain the observed values ​​of the load. ; The Kalman gain is calculated using the following formula: ； Kalman coefficient characterizes the new measurement value The degree of credibility; Correcting the predicted values ​​with the measured values ​​yields the optimal posterior estimate: ； in, Known as a news item, it is the difference between a predicted value and a measured value. Update the covariance of the posterior estimation error: 。 5. The cross-layer optimized routing protocol method for highly dynamic ad hoc network scenarios according to claim 4, characterized in that, Step 4 describes the construction of a comprehensive evaluation model to achieve cross-layer resource joint optimization based on a greedy algorithm for MPR selection, as follows: Step 4.1: Traverse the set of one-hop neighbors of this node. and two neighbor sets Query the link quality, load status and node reachability of one-hop neighbors, and then record the node energy calculated according to the cross-layer optimization model in the MRP candidate neighbor table; Step 4.2, Traverse all If a two-hop neighbor node exists, is there a uniquely reachable one-hop neighbor node? If so, add it to the MPR selection table. Step 4.3: Query existing MPR nodes, from The two-hop neighbors covered by the MPR node are deleted in the process. Step 4.4, Query Check if it is an empty set; if not, continue iterating. Based on the node energy calculated by the cross-layer optimization model, select the one-hop neighbor node with the highest energy and add it to the MPR selection table; Step 4.5: Query existing MPR nodes, and then from... The two-hop neighbors covered by the MPR node are deleted in the process. Step 4.6, repeat steps 4.4 and 4.5 until... Empty.

6. A cross-layer optimized routing protocol system for highly dynamic ad hoc network scenarios, characterized in that, This system is used to implement the cross-layer optimized routing protocol method for highly dynamic ad hoc network scenarios as described in any one of claims 1 to 5. The system includes a cross-layer optimization model construction module, a link quality assessment module, an update interaction module, and an MPR selection module, wherein: A cross-layer optimization model construction module is used to establish a cross-layer optimization model for highly dynamic self-organizing networks. The link quality assessment module calculates the physical layer link quality. Link quality is assessed using a windowed exponentially weighted moving average smoothing method. Update the interaction module to statistically analyze the MAC layer load based on Kalman filtering. And update and interact through network maintenance messages; The MPR selection module constructs a comprehensive evaluation model to achieve cross-layer resource joint optimization based on a greedy algorithm for MPR selection.

7. A mobile terminal, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the cross-layer optimized routing protocol method for highly dynamic ad hoc network scenarios as described in any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps in the cross-layer optimized routing protocol method for highly dynamic ad hoc network scenarios as described in any one of claims 1 to 5.

9. A computer device, characterized in that, include: The system includes a memory and a processor, which are interconnected. The memory stores computer instructions, and the processor executes the computer instructions to perform the cross-layer optimized routing protocol method for highly dynamic ad hoc network scenarios as described in any one of claims 1 to 5.

10. A computer program product, characterized in that, It includes computer instructions, which are used to cause a computer to execute the cross-layer optimized routing protocol method for highly dynamic ad hoc network scenarios as described in any one of claims 1 to 5.

Images