Wireless sensor network multipath routing protocol combining energy saving and load balancing

Abstract

The invention discloses a wireless sensor network multipath routing protocol combining energy saving and load balancing, and belongs to the field of wireless sensor network routing protocols. The existing wireless sensor network has the characteristics of large network data volume, limited node energy, large network scale and the like, but the problems of monitoring holes and transmission path breakage of the multipath routing protocol of the wireless sensor network cannot be solved by comprehensively considering the problems. A wireless sensor network multipath routing protocol combining energy saving and load balancing comprises the steps that a fitness function is designed based on a routing protocol AOMDV-FF of node minimum energy and path load, and a hop count is replaced to serve as a basis for judging whether routing is good or not; on the basis of a routing protocol AOMDV-FF, a plurality of returned routes are recombined by using a bee colony algorithm, and a multipath routing protocol AOMDV-FA based on the bee colony algorithm is designed. According to the method, the speed of searching the optimal path can be increased, the algorithm calculation speed is reduced, and the effects of saving energy consumption and prolonging the network life cycle are achieved.

Description

Technical Field

[0001] The invention relates to a wireless sensor network multipath routing protocol, and in particular to a wireless sensor network multipath routing protocol combining energy saving and load balancing. Background Art

[0002] For wireless sensor networks, a large number of wireless sensor nodes is the basic premise for forming a network. Sensor nodes have the advantages of low power consumption, low cost, and small size. Nodes communicate with each other through short-range wireless communication technology. However, nodes are often deployed in areas that are difficult for people to reach and have limited energy. As a result, once the power of the node is exhausted, it will be killed immediately. Dead nodes cannot participate in monitoring and transmission tasks. This will cause holes in the monitoring area, and the path passing through this node will also be disconnected. Therefore, in wireless sensor networks, the issue of node energy should be considered as the focus, and the survival time of the node should be extended as much as possible to avoid energy holes.

[0003] During the working process of the node, the transmission and reception of data consumes most of the energy, while the energy consumption of data processing is much smaller. The frequency of data transmission and reception is mainly controlled by the routing protocol. Therefore, designing an energy-efficient routing protocol plays a great role in improving the performance of wireless sensor networks. Researchers have designed many types of routing protocols based on the characteristics of wireless sensor networks. The common ones are cluster routing protocols and multipath routing protocols. The cluster routing protocol divides the nodes into multiple clusters, which are managed uniformly by the cluster head. The cluster head node is responsible for collecting the data of all nodes in its cluster and transmitting it to other clusters or base station nodes. Other ordinary nodes only need to upload data to the cluster head node of the cluster to which it belongs, which reduces energy consumption. The multipath routing protocol starts the route discovery process once and saves multiple paths that can communicate. The route discovery process is restarted only when all the saved paths are disconnected. This protocol can reduce the number of times route discovery is started, reduce the amount of information transmitted in the network, and save node energy. For example, mobile wireless sensor networks, wireless sensor networks in marine environments, wireless sensor networks in volcanic environments, and wireless sensor networks in military applications.

[0004] At present, many researchers have conducted a lot of research on routing protocols for wireless sensor networks and made great progress. Smail et al. proposed the AOMR-LM protocol. The algorithm introduced an energy-aware mechanism, calculated the energy level according to the remaining energy of the node, selected and classified the path according to the energy level of the node, extended the network life cycle, and reduced the end-to-end delay. Xia Li and Song Zhi proposed an algorithm based on load and location information. The algorithm uses a delay algorithm and a congestion control mechanism to automatically select idle nodes and establish a new link to transmit data, reducing the packet loss rate and end-to-end delay, and improving network utilization. Mohd and Chiranjeev proposed the QMLB multipath protocol based on FMLB to address the network congestion problem. It uses uncongested or less congested paths to transmit data, distributes network traffic among multiple paths according to the minimum queue length, and reduces network delay. Han et al. proposed a cross-layer optimized node-disjoint multipath routing protocol. When forwarding nodes at the routing layer, they select according to the neighbor node information of the next two hops. The protocol shortens the average path length without consuming a lot of energy. Prakash proposed a fault-tolerant routing algorithm based on ant colony optimization, which uses the ant colony algorithm to optimize the various parameters of QOS. The algorithm can find error-prone nodes in the route discovery phase, thereby skipping these nodes from the selected path. Compared with the existing non-ant colony algorithm, the QOS of this algorithm is improved by 20-30%. Taha et al. improved the routing selection mechanism of the traditional AOMDV protocol and proposed the FF-AOMDV routing algorithm. The algorithm designed a fitness function based on energy consumption and distance, and selected the path with the lowest energy consumption and the shortest distance as the main transmission path. It has good performance in terms of throughput and packet delivery rate. However, although the protocol has improved the function of selecting paths, it is not sufficiently considered in the route discovery phase, resulting in no obvious improvement in the network life cycle.

[0005] Most of the above studies only considered the impact of energy or load on data transmission, without combining the two. Overloaded nodes in the network often cause data retransmission, which further increases energy consumption.

[0006] (1) In wireless sensor networks, the AOMDV routing protocol only relies on the number of hops as the criterion for judging the quality of the path, without considering information such as node energy and load, resulting in excessive energy consumption.

[0007] (2) In wireless sensor networks, the AOMDV routing protocol adopts a flooding path establishment process, which easily leads to incomplete discovery of available transmission nodes and poor transmission effect of the established path.

[0008] (3) In the research on signal reconstruction in wireless sensor networks, although the data acquisition and reconstruction system based on compressed sensing can save node energy, the signal reconstruction end still has the problem of low reconstruction efficiency.

[0009] The present invention proposes a new fitness function based on node energy and path load to replace the number of hops as the basis for judging the path, and proposes the AOMDV-FF protocol based on the AOMDV routing protocol. Then, the artificial bee colony algorithm is used to further screen the returned path and the AOMDV-FA protocol is proposed. The protocol proposed by the present invention tries to use nodes with more residual energy and less load to transmit data, so that low-energy nodes have a longer life cycle, extend the life time of the link, reduce the frequency of routing initiation in the network, and has a good performance in wireless sensor networks. Summary of the invention

[0010] Wireless sensor networks have the characteristics of large network data volume, limited node energy, and large network scale. The monitoring holes and transmission path breakage problems of wireless sensor network multipath routing protocols have not been solved. At the same time, there are also problems such as excessive energy consumption of sensor network nodes, unbalanced node loads, and low signal reconstruction rate. The existing technology has not been able to comprehensively consider the above problems. The purpose of the present invention is to solve the above problems and propose a wireless sensor network multipath routing protocol that combines energy saving and load balancing.

[0011] The above purpose is achieved through the following technical solutions:

[0012] A multipath routing protocol for wireless sensor networks combining energy saving and load balancing, the method includes designing a fitness function based on the routing protocol AOMDV-FF with minimum node energy and path load, replacing the number of hops as the basis for judging the quality of the route; reorganizing the returned multiple routes using the bee colony algorithm based on the routing protocol AOMDV-FF, and designing a multipath routing protocol based on the routing protocol AOMDV-FF. bee Multipath routing protocol based on swarm algorithm AOMDV-FA.

[0013] Furthermore, the fitness function includes an energy fitness function and a queue fitness function; wherein:

[0014] The energy fitness is expressed as F e , the expression is:

[0015]

[0016] F e represents the energy fitness function, F min represents the energy value of the node with the minimum energy on the path, F full Represents the initial energy value of the node. The initial energy of all nodes in the network is the same; F eIt is defined by the ratio of the energy value of the node with the lowest energy in the path to the initial energy value of the node, which characterizes the overall situation of the path energy;

[0017] The queue fitness function is expressed as F q , the expression is:

[0018]

[0019] Set U contains all intermediate nodes between the current node and the destination node, Q i is the cache queue length of node i, H i is the number of hops from the current node to the destination node, Q full is the capacity of the node cache queue. Assume that the capacity of each node queue is the same, that is, the queue can cache a maximum of data packets. The data packet is transmitted to the destination node in a hop-by-hop manner on the path. The load on the node far from the destination node will eventually be transmitted to the destination node via the node close to the destination node. The load on the node far from the destination node consumes more energy on the path. q The calculation of F is firstly to take the product of the queue length of the current node and the number of hops to the destination node as the load of the current node, and then to normalize the load of all nodes on the path; when forwarding RREP in the routing response process, the load of the intermediate nodes passed is accumulated. The larger the load of the path, the higher the load of F. q The smaller the value of is, the smaller the value of F will be, and the possibility of establishing a route will also be reduced;

[0020] The fitness function is expressed as F, and the expression is: F = F e +F q .

[0021] Furthermore, the bee colony algorithm is used to reorganize the returned multiple routes, and the process of designing a multipath routing protocol AOMDV-FA based on the bee colony algorithm includes:

[0022] First, the route discovery method is a process of finding multiple routes from the source node S to the destination node D, specifically:

[0023] (1) Start the route discovery phase and broadcast RREQ to find a path for communication;

[0024] (2) The RREP data packet returns to establish multiple available paths and store their fitness values;

[0025] (3) The leader bee discovers a new path near the available paths. If the new path has a higher fitness, it will replace the old path. Otherwise, it will abandon the new path.

[0026] (4) The leader bee informs the follower bee of the path information and the follower bee chooses the path in a roulette wheel manner;

[0027] (5) The leading bee and the following bee exchange their identities, completing one iteration;

[0028] (6) The new leader bee continues to search for a new path;

[0029] (7) If the nectar source is not updated after n iterations, the scout bees are started to search for a new path;

[0030] The paths finally stored are the three paths with the highest fitness values;

[0031] Second, the process of calculating the fitness value of the path is specifically:

[0032] (1) Improve the routing table;

[0033] The energy, load and fitness information of the corresponding path are stored in the improved routing table, and the energy, load and fitness information of the corresponding path are transmitted back in the RREP reply packet, and two information bits of node energy and path load are added on the basis of the RREP packet of the original AOMDV protocol;

[0034] (2) Calculate the fitness function using RREP;

[0035] When the target node receives the RREQ sent by the source node, it sends a RREP to the source node along the reverse path; when the intermediate node receives the RREP, it compares its own remaining energy with the response energy stored in the RREP. If its own energy is less than the response energy, it stores the remaining energy in the response energy item. Otherwise, if its own energy is greater than or equal to the response energy, no operation is performed; then, the product of the node's load number and the number of hops from the target node is calculated, and the product is stored in the response load in an accumulated manner;

[0036] (3) When the source node receives the RREP, it records the current path and stores the path and its fitness value together;

[0037] Third, use the bee swarm algorithm to optimize the path.

[0038] Furthermore, the process of optimizing the path by using the bee colony algorithm is specifically:

[0039] Initialization: The honey source represents the available path i, that is, the return path. The number of honey sources is defined according to the number of available paths. The number of leading bees, following bees, and scout bees is defined. The number of leading bees is equal to the number of following bees, which is equal to the number of honey sources.

[0040] Leading bees update nectar sources: The leading bees first calculate the fitness value of the available path i, and then search for new nectar sources near the nectar source. If the fitness value of the new nectar source is greater than the existing path, the new path will be used to replace the current path;

[0041] Search honey source rules: among the intermediate nodes of multiple paths returned by the protocol, randomly select one node to replace one intermediate node of the current path;

[0042] The leader bee returns to the hive to exchange information: The leader bee returns to the hive to exchange nectar source information with all the follower bees. At this time, the follower bee and the leader bee exchange their identities. The follower bee becomes the new leader bee and selects the path memory of the leader bee with probability p. The probability p is calculated by roulette. The original leader bee becomes a follower bee and loses all its memory. After one iteration, the new leader bee will continue to update the nectar source.

[0043]

[0044] Generate scout bees: When the nectar source has not become a better nectar source after reaching the threshold L after N iterations, the current nectar source is stored in descending order of fitness value, and scout bees are generated at this time; the scout bees remember the path with the smallest fitness value, and randomly select j nodes from the intermediate nodes of the multiple paths returned by the protocol to replace the intermediate nodes of the existing path, where j is 1 / 3 of the current path length; calculate the fitness value of the new path, and if the fitness value is greater than the fitness value of the current path, replace the path with the new path; if the fitness value of the new path is less than the fitness value of the current path, repeatedly generate new paths until the fitness value of the path is greater than the current path, and at most 10 paths are repeatedly generated.

[0045] The beneficial effects of the present invention are:

[0046] 1. The present invention proposes a fitness function for selecting a path, combining node energy and load, and can judge whether the node's energy and load are suitable for data transmission through the value of the fitness function, which can avoid low-energy or high-load nodes from transmitting data and extend the node's survival time. Design an energy-saving and load-balanced multipath routing protocol, improve the routing criterion, use the fitness value based on the minimum node energy and path load instead of the hop count as the standard for judging the quality of the route, improve the routing protocol's packet delivery rate, throughput and other parameters, and reduce network energy consumption.

[0047] 2. In combination with the characteristics of wireless sensor network routing protocols, a transmission path optimization mechanism based on the artificial bee colony algorithm is designed to split and reorganize the intermediate nodes of multiple paths returned by multipath routing, so as to solve the problem that the flooding-style transmission path establishment is prone to fall into local optimality.

[0048] 3. Study the signal reconstruction theory based on compressed sensing in wireless sensor networks, analyze and compare the advantages and disadvantages of various reconstruction algorithms, combine the greedy algorithm and convex optimization algorithm, and design a reconstruction algorithm with a higher reconstruction rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1It is a graph showing the change of packet delivery rate with simulation time when the maximum moving speed involved in the simulation verification part of the present invention is 12 m / s;

[0050] Figure 2 It is the simulation result of the packet delivery rate of five routing protocols under different maximum moving speeds involved in the simulation verification part of the present invention;

[0051] Figure 3 It is a simulation result diagram of the throughput varying with time involved in the simulation verification part of the present invention;

[0052] Figure 4 It is a simulation result diagram of the throughput involved in the simulation verification part of the present invention changing with the maximum moving speed;

[0053] Figure 5 It is a simulation result diagram of the average network delay over time involved in the simulation verification part of the present invention;

[0054] Figure 6 It is the simulation result of the average network delay of five routing protocols under different maximum moving speeds involved in the simulation verification part of the present invention;

[0055] Figure 7 It is a simulation result diagram of energy consumption varying with time involved in the simulation verification part of the present invention;

[0056] Figure 8 It is a simulation result diagram of energy consumption varying with the number of nodes involved in the simulation verification part of the present invention;

[0057] Fig. 9 It is a diagram of the relationship between routing overhead and time involved in the simulation verification part of the present invention. DETAILED DESCRIPTION

[0058] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Specific implementation method one:

[0060] The present embodiment is a wireless sensor network multipath routing protocol that combines energy saving and load balancing. The method includes designing a fitness function based on the routing protocol AOMDV-FF of the minimum node energy and path load, replacing the number of hops as the basis for judging the quality of the route, and solving the problem that the routing criteria of the original routing protocol are not comprehensive; on the basis of the routing protocol AOMDV-FF, the bee colony algorithm is used to reorganize the returned multiple routes, and the multipath routing protocol AOMDV-FA based on the bee colony algorithm is designed to make up for the shortcomings of flooding-based path establishment. Specific implementation method 2:

[0062] Different from the first specific implementation mode, in this implementation mode, a wireless sensor network multipath routing protocol combining energy saving and load balancing is described, wherein the fitness function includes an energy fitness function and a queue fitness function; wherein:

[0063] The energy fitness is expressed as F e , the expression is:

[0064]

[0065] F e represents the energy fitness function, F min represents the energy value of the node with the minimum energy on the path, F full Indicates the initial energy value of the node. By default, the initial energy of all nodes in the network is the same. When any node on the path dies, the entire path will be interrupted, so F e It is defined by the ratio of the energy value of the node with the lowest energy in the path to the initial energy value of the node, and it characterizes the overall situation of the path energy.

[0066] The queue fitness function is expressed as F q , the expression is:

[0067]

[0068] Set U contains all intermediate nodes between the current node and the destination node, Q i is the cache queue length of node i, H i is the number of hops from the current node to the destination node, Q full is the capacity of the node cache queue, assuming that the capacity of each node queue is the same, that is, the maximum number of packets that can be cached in the queue. The data packet is transmitted to the destination node in a hop-by-hop manner on the path. The load on the node far away from the destination node will eventually be transmitted to the destination node via the node close to the destination node. Therefore, the load on the node far away from the destination node consumes more energy on the path. F qThe calculation of F is firstly to take the product of the queue length of the current node and the number of hops to the destination node as the load of the current node, and then normalize the load of all nodes on the path. In the process of routing response, the load of the intermediate nodes passed by is accumulated when forwarding RREP. The larger the load of the path, the higher the load of F. q The smaller the value of is, the smaller the value of F will be, and the possibility of establishing a route will also be reduced.

[0069] The fitness function is expressed as F, and the expression is: F = F e +F q .

[0070] An algorithm for calculating the path load is used to calculate the node load value based on the capacity of the node cache queue and the number of hops from the node to the destination node. The load values ​​of all nodes on the path are accumulated as the load fitness value of the entire path. The energy value of the node with the minimum energy on the path is used as the energy fitness value of the path. A new fitness function is then proposed, which is the load fitness value plus the energy fitness value. The fitness function is used instead of the number of hops as the basis for judging the quality of path communication, solving the problem of excessive energy consumption of existing routing protocols. Specific implementation method three:

[0072] Different from the second specific implementation mode, the present implementation mode is a wireless sensor network multipath routing protocol combining energy saving and load balancing, and uses the bee colony algorithm to reorganize the returned multiple routes. In the process of designing the multipath routing protocol AOMDV-FA based on the bee colony algorithm, in the traditional bee colony algorithm, the initial honey source is randomly generated, and the role of the scout bee is to find a new honey source far away from the existing honey source, in order to prevent the optimization process from falling into the local optimal solution. However, in the case of the present invention, the returned available path is defined as the initial honey source, which ensures the rationality of the initial honey source and does not require a large amount of energy to search for new honey sources. Therefore, in this protocol, the role of the scout bee is to speed up the search for the optimal path when the path update is slow. The present invention uses the intermediate nodes of the return path as the available node set of the leading bee and the scout bee because in the actual layout of the wireless sensor network, these intermediate nodes are often close to each other and can easily reach the communication distance. Using these intermediate nodes for replacement can make the generated new path have a greater possibility of connection, and also reduce the amount of calculation of the algorithm. The bee colony algorithm is an optimization algorithm that simulates the foraging behavior of natural bees. The algorithm is often used to solve optimization problems, especially continuous optimization problems. At the same time, because the bee colony algorithm has the characteristics of few parameters and simple calculation, it is very suitable for nodes with limited energy. In the field of wireless sensor networks, artificial bee colony algorithms have long been widely used, and are often used in data collection, node positioning, data transmission, coverage, etc. Including: designing routing discovery methods, calculating the fitness value of the path, and using the bee colony algorithm to optimize the path; among them,

[0073] First, the route discovery method is a process of finding multiple routes from the source node S to the destination node D, specifically:

[0074] (1) Start the route discovery phase and broadcast RREQ to find a path for communication;

[0075] (2) The RREP data packet returns to establish multiple available paths and store their fitness values;

[0076] (3) The leader bee discovers a new path near the available paths. If the new path has a higher fitness, it will replace the old path. Otherwise, it will abandon the new path.

[0077] (4) The leader bee informs the follower bee of the path information and the follower bee chooses the path in a roulette wheel manner;

[0078] (5) The leading bee and the following bee exchange their identities, completing one iteration;

[0079] (6) The new leader bee continues to search for a new path;

[0080] (7) If the nectar source is not updated after n iterations, the scout bees are started to search for a new path;

[0081] The paths finally stored are the three paths with the highest fitness values;

[0082] Second, the process of calculating the fitness value of the path is specifically:

[0083] (1) Improve the routing table;

[0084] Add corresponding information items. In the AOMDV-FA protocol, each node has its own routing table. The routing table is composed of one or more routing table items, and each routing table item records the routing information to different destination addresses. Since the AOMDV-FA protocol is a multi-path routing protocol, there may be multiple paths to the same destination address. The information of these multiple paths constitutes a path list and is stored in the corresponding routing table item. The present invention stores the energy, load, suitability and other information of the corresponding path in the improved routing table, and transmits back the energy, load, suitability and other information of the corresponding path in the RREP reply packet. Therefore, it is also necessary to modify the RREP and add two information bits of node energy and path load on the basis of the RREP packet of the original AOMDV protocol;

[0085] (2) Calculate the fitness function using RREP;

[0086] When the target node receives the RREQ sent by the source node, it sends a RREP to the source node along the reverse path; when the intermediate node receives the RREP, it compares its own remaining energy with the response energy stored in the RREP. If its own energy is less than the response energy, it stores the remaining energy in the response energy item. Otherwise, if its own energy is greater than or equal to the response energy, no operation is performed; then, the product of the node's load number and the number of hops from the target node is calculated, and the product is stored in the response load in an accumulated manner;

[0087] Here we follow Algorithm 1.

[0088] (3) When the source node receives the RREP, it records the current path and stores the path and its fitness value together;

[0089] Third, use the bee swarm algorithm to optimize the path; after each route discovery phase, the bee swarm algorithm is used to optimize the generated path to prevent the established path from having too low fitness. Specific implementation method four:

[0091] Different from the third specific implementation mode, the present implementation mode is a wireless sensor network multipath routing protocol combining energy saving and load balancing, and the path optimization process using the bee colony algorithm is specifically:

[0092] Initialization: The honey source represents the available path i, that is, the return path. The number of honey sources is defined according to the number of available paths. The number of leading bees, following bees, and scout bees is defined. The number of leading bees is equal to the number of following bees, which is equal to the number of honey sources.

[0093] Leading bees update nectar sources: The leading bees first calculate the fitness value of the available path i, and then search for new nectar sources near the nectar source. If the fitness value of the new nectar source is greater than the existing path, the new path will be used to replace the current path;

[0094] Search honey source rules: among the intermediate nodes of multiple paths returned by the protocol, randomly select one node to replace one intermediate node of the current path;

[0095] The leader bee returns to the hive to exchange information: The leader bee returns to the hive to exchange nectar source information with all the follower bees. At this time, the follower bee and the leader bee exchange their identities. The follower bee becomes the new leader bee and selects the path memory of the leader bee with probability p. The probability p is calculated by roulette. The original leader bee becomes a follower bee and loses all its memory. After one iteration, the new leader bee will continue to update the nectar source.

[0096]

[0097] Generate scout bees: When the nectar source has not become a better nectar source after reaching the threshold L after N iterations, the current nectar source is stored in descending order of fitness value, and scout bees are generated at this time; the scout bees remember the path with the smallest fitness value, and randomly select j nodes from the intermediate nodes of the multiple paths returned by the protocol to replace the intermediate nodes of the existing path, where j is 1 / 3 of the current path length; calculate the fitness value of the new path, and if the fitness value is greater than the fitness value of the current path, replace the path with the new path; if the fitness value of the new path is less than the fitness value of the current path, repeatedly generate new paths until the fitness value of the path is greater than the current path, and at most 10 paths are repeatedly generated.

[0098] The fitness function is described as:

[0099]

[0100] The bee colony algorithm is described as:

[0101]

[0102]

[0103] The method of the present invention is simulated and verified:

[0104] 1. Simulation environment and parameter design

[0105] In order to verify the effectiveness of the protocol proposed in this paper, this paper uses NS2 software to perform simulations in different simulation environments. Table 1 shows the various parameters of the experiment.

[0106]

[0107]

[0108] 2. Performance indicators

[0109] The following indicators are used for performance evaluation:

[0110] (1) Packet Delivery Rate (PDR): refers to the ratio of the number of data packets successfully received by all destination nodes to the number of data packets sent by all source nodes. This indicator indicates the quality of the network's transmission performance. The larger the indicator, the less data packet loss and the more reliable the network. PDR is calculated as follows:

[0111]

[0112] Among them, P d represents the number of data packets delivered, P s is the number of packets sent.

[0113] (2) Throughput (TH): refers to the total number of bits successfully transmitted by the network, measured in megabits per second. The larger the index, the greater the amount of data transmitted per unit time and the less packet loss.

[0114]

[0115] (3) End-to-end delay (EED): refers to the average transmission time required for a destination node in the network to successfully receive a data packet. This indicator reflects the efficiency of network data transmission. The smaller the average network delay, the lower the network transmission time and the lighter the congestion.

[0116]

[0117] Where n represents the total number of messages received by the destination node, EED i Indicates the time each message is received minus the time it is sent.

[0118] (4) Energy consumption (EC): refers to the total energy consumed by nodes in the network during the simulation time. The lower the energy consumption, the more energy-efficient the network.

[0119]

[0120] Among them, I i represents the initial energy of the node, E i represents the remaining energy of the node, and n is the total number of nodes in the network.

[0121] (5) Routing Overhead (RO): refers to the ratio of the number of routing control packets sent by all nodes in the network to the number of data packets sent by all nodes. This indicator reflects the utilization rate of effective network resources. The smaller the indicator, the less additional network overhead, indicating a higher effective utilization rate of network resources.

[0122]

[0123] Among them, RO p Indicates the number of routing packets sent by all nodes in the network, DT p Indicates the number of valid data packets sent by all nodes in the network.

[0124] 3. Simulation results

[0125] (1) The comparison diagram of packet delivery rate is represented as follows: Figure 1 , Figure 2 shown.

[0126] Figure 1The following is a graph showing the packet delivery rate as the simulation time changes when the maximum moving speed is 12m / s. It can be seen from the figure that as the simulation time increases, the packet delivery rates of the five routing protocols all decrease. This is because as the simulation time increases, the energy of the nodes becomes less and less, and the number of dead nodes increases, resulting in more and more data packets required for route discovery and route maintenance. This also increases the traffic in the network, making the congestion in the network more serious and reducing the packet delivery rate of the network. The AOMDV-FA and AOMDV-FF proposed in this paper take into account the impact of energy and load on path transmission efficiency when establishing routes, which can avoid data transmission by nodes with low energy and high load, so the packet delivery rate has better performance.

[0127] Figure 2 The following are the simulation results of the packet delivery rates of five routing protocols under different maximum moving speeds. It can be seen from the figure that as the moving speed increases, the packet delivery rates of the five routing protocols decrease. This is because movement will cause changes in the network topology and cause the transmission path to be broken. The greater the moving speed, the more serious the impact, so the packet delivery rate decreases. The AODV protocol is a single-path protocol. After the path is interrupted, it cannot be sent through other paths and can only re-perform the route discovery process. Therefore, the packet delivery rate is the lowest and is also the fastest to be affected by speed. Compared with the AOMDV protocol, the other three multi-path routing protocols consider more comprehensive criteria when establishing routes, and the established paths have higher stability. This group delivery The packet delivery rate of AOMDV-FA protocol is the highest because it takes into account The impact of energy and load on the path is reduced, so it is less affected by the increase in mobile speed. At the same time, the protocol optimizes the path discovery process so that the established path has more energy and less load, so the packet delivery rate of the AOMDV-FA protocol is also higher than that of AOMDV-FF.

[0128] (2) A comparison of the throughput and time relationship, such as Figure 3 , 4 shown.

[0129] Figure 3 The simulation result graph of throughput changing with time is shown in Figure 2. Figure 4 The simulation result of throughput changing with the maximum mobile speed is shown in Figure 2. Throughput is related to factors such as network topology, routing protocol, and data flow. Figure 3 It can be seen from the figure that the throughput of the network using these five protocols to transmit data decreases over time. This is because the data traffic in the network will increase over time, which increases the probability of packet loss or retransmission in the network. At the same time, the death of some nodes will cause the path to be interrupted, which will lead to a decrease in network throughput. Figure 4It can be seen that the throughput decreases as the node movement speed increases. This is because the faster the node moves, the faster the network topology environment changes, and the probability of path breakage will also increase. Compared with the single-path protocol, the multi-path protocol can significantly improve the network throughput. The FF-AOMDV protocol takes into account the impact of node energy on network stability, so it has a significant improvement compared to the AOMDV protocol. The AOMDV-FA protocol and the AOMDV-FF protocol both consider the impact of node energy and load on network transmission data, so they perform slightly better than FF-AOMDV. The reason why the AOMDV-FA protocol is better than the AOMDV-FF protocol is that it uses a swarm algorithm to optimize the path, which can obtain a path with higher fitness. The throughput of the AOMDV-FA protocol is increased by 32% and 9% compared to the AOMDV protocol and the AOMDV-FF protocol, respectively.

[0130] (3) The relationship between end-to-end delay and time is shown in the figure below: Figure 5 , 6 shown.

[0131] Figure 5 This is the simulation result of the average network delay over time. As can be seen from the figure, the delay increases with the increase of simulation time. This is because the number of data packets in the network will increase over time. At the same time, due to the death of some nodes, the router needs to find a new path suitable for transmission, which will increase the network delay. Since the AOMDV-FF protocol increases the size of the data packet, the routing transmission speed will decrease, and the probability of network congestion will increase. Therefore, the average network delay will be slightly greater than the FF-AOMDV protocol. However, the AOMDV-FF protocol will give priority to nodes with small loads to transmit data, which enhances the stability of the link. As the simulation time increases, the delay simulation results of the two protocols are similar. The AOMDV-FA protocol has a larger delay because it adds the process of optimizing the path by the swarm algorithm. However, since the path it establishes is more stable, its overall delay is still smaller than the AOMDV protocol.

[0132] Figure 6 It is the simulation result of the average network delay of five routing protocols under different maximum moving speeds. As can be seen from the figure, with the increase of packet sending rate, the delay of the five routing protocols increases. This is because when the moving speed increases, the network topology is very easy to change, which can easily cause the existing path to break, thereby increasing the network delay. Multipath routing protocols save multiple backup paths and can significantly reduce network delay. In a high-speed moving network, the route discovery process will be frequently started. The route discovery process of the AOMDV-FA protocol takes longer than that of the AOMDV-FF protocol and the FF-AOMDV protocol. Therefore Will have a larger delay.

[0133] (4) Energy consumption comparison chart Figure 7 and 8 shown.

[0134] Figure 7 The figure is a simulation result of energy consumption changing over time. Energy saving is one of the main goals of the protocol in this paper. It can be seen from the figure that the energy consumption of the AOMDV-FA protocol proposed in this paper is less than that of other protocols. When establishing a path, the algorithm proposed in this paper will calculate the minimum energy of the nodes in the path and use it as the energy fitness function of the path to avoid using low-energy nodes for data transmission, so that the established path has higher stability. On the other hand, the algorithm in this paper adopts the idea of ​​congestion control and uses the path load as the load fitness function, which can reduce the probability of retransmission in the network, which also reduces the overall energy consumption of the network. Other protocols do not take into account the problem of network congestion.

[0135] Figure 8 The following is a simulation result diagram showing the energy consumption changing with the number of nodes. In order to ensure consistent node density, simulation experiments with different numbers of nodes use different area sizes. When the number of simulated nodes is 60, the network topology area is 1000m*1000m. In order to avoid excessive congestion in a network with too many nodes, which may lead to node death and affect the accuracy of the experiment, the simulation time is set to 200s. A comprehensive comparison of the two simulation result diagrams shows that the AOMDV-FA protocol saves 26% and 12% of energy compared to AOMDV and FF-AOMDV, respectively. The main reason why AOMDV-FA has a better energy-saving effect than the FF-AOMDV protocol is that it avoids data transmission by nodes with overload, which makes the established transmission path more stable and less likely to fail due to node energy exhaustion.

[0136] (5) The relationship between routing cost and time, such as Fig. 9 shown.

[0137] The number of routing overhead messages is related to parameters such as network stability, routing algorithm, network scale, and network topology.

[0138] Fig. 9The figure below shows the change of routing overhead with simulation time. It can be seen from the figure that as the simulation time increases, the routing overhead of the four routing protocols is increasing. This is because as the simulation time increases, the number of routing maintenance packets is gradually increasing, and the number of dead nodes is also increasing. When all the paths saved by the route fail, the routing discovery process will be started, increasing the number of routing messages in the network. AODV is a single-path routing protocol, and the number of packets used for routing maintenance is much less than that of multi-path routing protocols, so the routing overhead is smaller than that of other multi-path routing protocols. However, once its path fails and there is no backup path to replace it, the routing discovery process can only be restarted, so the routing overhead grows very fast. The AOMDV-FA protocol proposed in this paper uses a path with more node residual energy and less load for transmission, which prolongs the path survival time and reduces the frequency of routing discovery, so it has a smaller routing overhead than other multi-path protocols.

[0139] in conclusion:

[0140] Aiming at the limited energy of nodes in wireless sensor networks, this paper proposes a routing protocol based on node energy and path load. In this protocol, a fitness function is used as the routing criterion, which takes into account the impact of energy and load on the transmission data, making the established path more stable. At the same time, the bee colony algorithm is used to split and reorganize the established path to find a path with a higher fitness value. NS2 is used to test the five protocols of AOMDV-FA, AOMDV-FF, FF-AOMDV, AOMDV, and AODV under different simulation times, node speeds, and node numbers. The simulation was performed using grouping The test is based on five indicators: delivery rate, throughput, end-to-end delay, energy consumption, and routing overhead. AOMDV-FA has the best overall performance.

[0141] The embodiments of the present invention disclose preferred embodiments, but are not limited thereto. A person skilled in the art can easily understand the spirit of the present invention based on the above embodiments and make different extensions and changes. However, as long as they do not deviate from the spirit of the present invention, they are all within the protection scope of the present invention.

Claims

1. A multipath routing protocol for wireless sensor networks combining energy saving and load balancing, Features: The method includes designing a fitness function based on the routing protocol AOMDV-FF of the minimum node energy and path load, replacing the hop count as the basis for judging the quality of the route; Based on the routing protocol AOMDV-FF, the bee swarm algorithm is used to reorganize the returned multiple routes, and a multi-path routing protocol AOMDV-FA based on the bee swarm algorithm is designed.

2. A wireless sensor network multipath routing protocol combining energy saving and load balancing according to claim 1, Features: The fitness function includes an energy fitness function and a queue fitness function; wherein: The energy fitness is expressed as F e , the expression is: F e represents the energy fitness function, F min represents the energy value of the node with the minimum energy on the path, F full Represents the initial energy value of the node. The initial energy of all nodes in the network is the same; F e It is defined by the ratio of the energy value of the node with the lowest energy in the path to the initial energy value of the node, which characterizes the overall situation of the path energy; The queue fitness function is expressed as F q , the expression is: Set U contains all intermediate nodes between the current node and the destination node, Q i is the cache queue length of node i, H i is the number of hops from the current node to the destination node, Q full is the capacity of the node cache queue. Assume that the capacity of each node queue is the same, that is, the queue can cache a maximum of data packets. The data packet is transmitted to the destination node in a hop-by-hop manner on the path. The load on the node far from the destination node will eventually be transmitted to the destination node via the node close to the destination node. The load on the node far from the destination node consumes more energy on the path. q The calculation of F is firstly to take the product of the queue length of the current node and the number of hops to the destination node as the load of the current node, and then to normalize the load of all nodes on the path; when forwarding RREP in the routing response process, the load of the intermediate nodes passed is accumulated. The larger the load of the path, the higher the load of F. q The smaller the value of is, the smaller the value of F will be, and the possibility of establishing a route will also be reduced; The fitness function is expressed as F, and the expression is: F = F e +F q .

3. A wireless sensor network multipath routing protocol combining energy saving and load balancing according to claim 2, Features: The process of using the bee colony algorithm to reorganize the returned multiple routes and designing the multipath routing protocol AOMDV-FA based on the bee colony algorithm includes: First, the route discovery method is a process of finding multiple routes from the source node S to the destination node D, specifically: (1) Start the route discovery phase and broadcast RREQ to find a path for communication; (2) The RREP data packet returns to establish multiple available paths and store their fitness values; (3) The leader bee discovers a new path near the available paths. If the new path has a higher fitness, it will replace the old path. Otherwise, it will abandon the new path. (4) The leader bee informs the follower bee of the path information and the follower bee chooses the path in a roulette wheel manner; (5) The leading bee and the following bee exchange their identities, completing one iteration; (6) The new leader bee continues to search for a new path; (7) If the nectar source is not updated after n iterations, the scout bees are started to search for a new path; The paths finally stored are the three paths with the highest fitness values; Second, the process of calculating the fitness value of the path is specifically: (1) Improve the routing table; The energy, load and fitness information of the corresponding path are stored in the improved routing table, and the energy, load and fitness information of the corresponding path are transmitted back in the RREP reply packet, and two information bits of node energy and path load are added on the basis of the RREP packet of the original AOMDV protocol; (2) Calculate the fitness function using RREP; When the target node receives the RREQ sent by the source node, it sends a RREP to the source node along the reverse path; when the intermediate node receives the RREP, it compares its own remaining energy with the response energy stored in the RREP. If its own energy is less than the response energy, it stores the remaining energy in the response energy item. Otherwise, if its own energy is greater than or equal to the response energy, no operation is performed; then, the product of the node's load number and the number of hops from the target node is calculated, and the product is stored in the response load in an accumulated manner; (3) When the source node receives the RREP, it records the current path and stores the path and its fitness value together; Third, use the bee swarm algorithm to optimize the path.

4. A wireless sensor network multipath routing protocol combining energy saving and load balancing according to claim 3, Features: The process of optimizing the path by using the bee colony algorithm is specifically as follows: Initialization: The honey source represents the available path i, that is, the return path. The number of honey sources is defined according to the number of available paths. The number of leading bees, following bees, and scout bees is defined. The number of leading bees is equal to the number of following bees, which is equal to the number of honey sources. Leading bees update nectar sources: The leading bees first calculate the fitness value of the available path i, and then search for new nectar sources near the nectar source. If the fitness value of the new nectar source is greater than the existing path, the new path will be used to replace the current path; Search honey source rules: among the intermediate nodes of multiple paths returned by the protocol, randomly select one node to replace one intermediate node of the current path; The leader bee returns to the hive to exchange information: The leader bee returns to the hive to exchange nectar source information with all the follower bees. At this time, the follower bee and the leader bee exchange their identities. The follower bee becomes the new leader bee and selects the path memory of the leader bee with probability p. The probability p is calculated by roulette. The original leader bee becomes a follower bee and loses all its memory. After one iteration, the new leader bee will continue to update the nectar source. Generate scout bees: When the nectar source has not become a better nectar source after reaching the threshold L after N iterations, the current nectar source is stored in descending order of fitness value, and scout bees are generated at this time; the scout bees remember the path with the smallest fitness value, and randomly select j nodes from the intermediate nodes of multiple paths returned by the protocol to replace the intermediate nodes of the existing path, where j is 1 / 3 of the current path length; calculate the fitness value of the new path, and if the fitness value is greater than the fitness value of the current path, replace the path with the new path; if the fitness value of the new path is less than the fitness value of the current path, repeatedly generate new paths until the fitness value of the path is greater than the current path, and at most 10 paths are repeatedly generated.

Images