Backscatter wireless network optimal path routing method, device and storage medium based on online learning
By using KL-UCB algorithm and online learning methods in backscatter wireless communication networks, dynamically adjusting transmission mode and path selection, optimizing link delay and energy efficiency gain, the optimal path selection problem caused by unknown link delay is solved. , improves network performance and data transmission rate.
Patent Information
- Application Number
- CN202210992514.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-08-18
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2042-08-18
AI Technical Summary
In backscatter wireless communication networks, the link delay is unknown and difficult to determine, resulting in poor results in traditional shortest path algorithms and the inability to effectively select the optimal path, resulting in a degradation of network transmission performance.
Using the optimal path routing method of backscattering wireless network based on online learning, the KL-UCB algorithm is used to determine the link index value based on network parameters and working parameters, and by dynamically adjusting the transmission mode and path selection, link delay and energy efficiency gain are optimized. , realize the optimal path selection.
It improves network performance, increases data transmission rate and energy efficiency, improves network stability and model reliability, and solves the problem of optimal path selection in dynamic wireless communication networks.
Smart Images

Figure CN115361722B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of backscatter communication, and in particular to an optimal path routing method, a computer device and a storage medium for a backscatter wireless network based on online learning. Background Art
[0002] Backscatter technology is a new type of energy-saving communication technology that enables wireless nodes to transmit data without power supply. Specifically, backscatter nodes transmit their own information by choosing whether to reflect the radio frequency signal in the environment. Although this technology can significantly save energy, it is not as stable as the active transmission method due to the strength of the radio frequency signal in the environment. Therefore, nodes with dual-mode transmission methods can use active transmission to ensure the stability of data transmission and save energy through backscatter communication. In a wireless communication network using dual-mode nodes, information flow and energy flow exist simultaneously and can be converted into each other.
[0003] As an important area in the network, the optimal path routing strategy plays a unique role in improving data transmission rate and reducing energy consumption. In wireless communication networks using backscatter technology, finding the optimal path for data transmission is the key to improving network performance.
[0004] Wireless communication networks with active transmission and backscatter transmission nodes involve the mutual conversion of information flow and energy flow, and due to unreliable links and random access protocols, link delays will change randomly, which makes it very complicated to select the optimal path to transmit data. At present, for static networks with known link delays, most of them use Dijkstra algorithm, Floyd algorithm, Bellman-Ford algorithm and SPFA algorithm to determine the shortest path. However, due to the complex environment in which wireless communication networks are located, link delays are unknown and difficult to determine. These algorithms need to calculate link delays in real time and select the shortest path for data transmission, which makes the entire network overhead huge and the results are unreliable. At the same time, if the network structure or link delay changes during the process of selecting the shortest path, the selected path may not be the shortest path or the path may be interrupted, thereby reducing the data transmission performance of the entire network. Secondly, the traditional shortest path algorithm does not consider the energy efficiency gain of nodes in dual transmission mode (active transmission and backscatter transmission) and the mutual conversion of information flow and energy flow, that is, nodes can absorb energy from adjacent active transmission nodes and use active transmission signals as carriers to transmit data. Therefore, the shortest path in the routing strategy is not the optimal path in the wireless communication network. Since different transmission modes selected by nodes will cause different transmission delays and energy consumption, it is difficult to determine the optimal path. In summary, the traditional shortest path algorithm cannot solve the optimal path problem in the wireless communication network well. Summary of the invention
[0005] In view of the technical problems that the link delay in the current backscatter wireless communication network is unknown and difficult to determine, resulting in poor effect of applying the current shortest path algorithm, the purpose of the present invention is to provide a backscatter wireless network optimal path routing method, computer device and storage medium based on online learning.
[0006] On the one hand, an embodiment of the present invention includes an optimal path routing method for a backscatter wireless network based on online learning, comprising:
[0007] Acquire network parameters and working parameters of each routing node in the backscatter wireless network; the routing node can choose to work in active transmission mode or backscatter transmission mode; the network parameters represent the topology information of the routing node in the backscatter wireless network, and the working parameters represent the working state of the routing node;
[0008] Using the KL-UCB algorithm, determining the link index value of the link between each two of the routing nodes according to the network parameters and the operating parameters of each two of the routing nodes;
[0009] Determine a target link; the target link is a link with the smallest link indicator value among all links;
[0010] transmitting data through the target link to determine an actual path gain of the target link;
[0011] Determine the empirical path gain of the target link; the empirical path gain is the total path gain of transmitting data through the target link when the data transmission success rate of the target link is known;
[0012] Determining a performance indicator value corresponding to the target link according to a difference between the actual path gain and the empirical path gain;
[0013] When the performance indicator value reaches a performance threshold, the target link is determined as an optimal path.
[0014] Further, the determining the target link includes:
[0015] Determining a link set; the link set initially includes all links of the backscatter wireless network;
[0016] The target link is determined from all links included in the link set.
[0017] Furthermore, the online learning-based optimal path routing method for backscatter wireless networks also includes:
[0018] When the performance indicator value does not reach the performance threshold, deleting the target link from the link set;
[0019] The process returns to the step of determining the target link and starts execution.
[0020] Further, the use of the KL-UCB algorithm to determine the link index value of the link between every two of the routing nodes according to the network parameters and working parameters of every two of the routing nodes includes:
[0021] By formula
[0022]
[0023] Determine the link index value; wherein ij represents a link between the routing nodes numbered j, ω ij represents the link indicator value corresponding to link ij, n represents the number of the data packet sent through link ij, represents a vector composed of the corresponding data transmission success rates when data is transmitted through all the links in the backscatter wireless network in the active transmission mode, represents a vector of data transmission success rates corresponding to data transmission performed through each link in the backscattering wireless network in the backscattering transmission mode, t ij (n) represents the sum of the number of transmission attempts on link ij before transmitting the nth data packet through link ij, V i represents the set of neighbor nodes of routing node i in the backscatter wireless network, represents the set of back-hop nodes of routing node i in the backscatter wireless network, represents the set of ready back-hop nodes of routing node i in the backscatter wireless network, each node in the ready back-hop node set has completed the preparation of backscatter transmission conditions, I represents the number of bits contained in one of the data packets, and represents the parameters that satisfy the corresponding conditions in the formula, represents the probability that routing node i chooses to use active transmission mode, represents the probability that routing node i chooses to use the backscatter transmission mode, k, h and l represent the summation number, represents the empirical data transmission success rate corresponding to data transmission through link ij in active transmission mode, represents the empirical data transmission success rate corresponding to data transmission through link ij in the backscatter transmission mode, represents the energy consumed by routing node i in active transmission mode, represents the energy consumed by routing node i in the backscatter transmission mode, represents the receiving power of routing node l as the receiving node in link il in active transmission mode, represents the efficiency of routing node l absorbing energy from routing node i, KL() represents the KL divergence obtained by processing using the KL-UCB algorithm, and g(n) represents a function related to n.
[0024] Further, g(n)=log n+4log log n.
[0025] Further, the transmitting data through the target link to determine the actual path benefit of the target link includes:
[0026] By formula
[0027]
[0028] Determine the actual path benefit; wherein, represents an averaging operation, N represents the total number of data packets transmitted through the target link, and D(n) represents the actual path benefit brought by transmitting the nth data packet.
[0029] Further, the determining the empirical path benefit of the target link includes:
[0030] By formula ND θ (p * ) determining the experience path benefit;
[0031] in,
[0032]
[0033] p represents a link, p * Indicates the target link.
[0034] Further, determining the performance indicator value corresponding to the target link according to the difference between the actual path gain and the empirical path gain includes:
[0035] By formula
[0036]
[0037] Determine the performance indicator value.
[0038] On the other hand, an embodiment of the present invention also includes a computer device, including a memory and a processor, wherein the memory is used to store at least one program, and the processor is used to load the at least one program to execute the backscatter wireless network optimal path routing method based on online learning in the embodiment.
[0039] On the other hand, an embodiment of the present invention further includes a storage medium storing a program executable by a processor, wherein the program executable by the processor is used to execute the online learning-based backscatter wireless network optimal path routing method in the embodiment when executed by the processor.
[0040] The beneficial effects of the present invention are as follows: the optimal path routing method for a backscatter wireless network based on online learning in the embodiment models the routing problem of a backscatter wireless network with unknown link delay into a combined multi-armed slot machine problem, and by accumulating known path benefit feedback, instantly selects the path with the highest benefit for data transmission in a dynamically changing network, thereby improving network performance, increasing energy efficiency, and increasing data transmission rate; for a wireless communication network with nodes in different transmission modes, the probability of each node selecting a different transmission mode is obtained respectively, and the probability is related to the data transmission success rate, and the average data transmission success rate of each outgoing link is obtained by weighting, thereby obtaining the average link delay; using the average link delay as a link indicator, the stability of the network is improved, and the reliability of the model is increased; the problem of difficulty in finding the optimal path in real time in the current dynamic wireless communication network with signal-energy coupling is solved, and at the same time, the single path evaluation standard in the online learning algorithm is improved, the calculation method of the link delay is optimized, and the network throughput is improved. BRIEF DESCRIPTION OF THE DRAWINGS
[0041] Figure 1 A schematic diagram of a network structure in which an optimal path routing method for a backscatter wireless network based on online learning can be applied in an embodiment;
[0042] Figure 2 A flow chart of an optimal path routing method for a backscatter wireless network based on online learning in an embodiment;
[0043] Figure 3 Another flow chart of the optimal path routing method for a backscatter wireless network based on online learning in an embodiment. DETAILED DESCRIPTION
[0044] Figure 1 It is a schematic diagram of a wireless communication network capable of using an active transmission mode and a backscatter communication mode. The wireless communication network shown in the figure can apply the backscatter wireless network optimal path routing method based on online learning in this embodiment.
[0045] Reference Figure 1 On the left side, node B uses active transmission to communicate with C, and provides energy and carrier to neighboring nodes G, F, and D. Nodes G and F do not transmit data because their stored energy cannot meet the energy required by their transmission mode, and they only absorb energy. Node D has enough energy, so it uses the active transmission signal in the environment as a carrier for backscatter transmission. In this figure, the information flow transmitted by the BC link is converted into energy flow on BG, BF, and BD, and the energy flow on BD is converted into information flow on DE. It can be seen that the energy flow and information flow in the network are converted into each other, and the transmission mode selection of the node is closely related to the link parameters and network topology. Therefore Figure 1The network shown can be called a signal-energy coupling network. Figure 1 As shown on the right, when node B changes from active transmission mode to backscatter transmission mode, the neighboring nodes cannot absorb energy and cannot use the active transmission signal as a carrier, so communication is impossible. The data transmission success rate of link BC will also change in response to different transmission modes.
[0046] The network is modeled as a directed graph with the following characteristics: 1. All nodes in the node set V are dual-mode nodes, that is, they can perform both active transmission and backscatter transmission, and can transmit data with multiple nodes at the same time. 2. All edges in the edge set E are directed edges, and each edge ij has a different unknown data success rate. and Represent the successful data transmission rate in active mode and backscattering mode respectively. 3. If the receiving power of a receiving node receiving data from a sending node is greater than the receiving sensitivity of the receiving node, there is a directed edge from the sending node to the receiving node, and the two nodes are adjacent. 4. There is only one source node and one sink node in the graph, which are responsible for transmitting and receiving data packets, and the intermediate nodes are only responsible for forwarding.
[0047] Since the transmission mode selection probability of the nodes is different, the corresponding transmission mode probability of each transmission link is also different. In order to more accurately describe the link status, the average data transmission success rate of the link after probability weighting is introduced. Assume that each node i has a different probability of active transmission or backscatter transmission, denoted as and Then the average data transmission success rate of the directed edge sent by node i can be expressed as In summary, according to the characteristics of geometric distribution, the average delay of link ij can be obtained as
[0048] In order to evaluate the link characteristics more comprehensively, it is necessary to consider the energy efficiency gain of the link sending node to the adjacent node on the basis of considering the average link delay. First, let the information flow rate on link ij be f ij , assuming that a data packet carries 1 bit of data, according to the different transmission modes of the nodes, there are and The energy consumed by active transmission of node i is The power consumed by backscatter transmission is And the energy efficiency absorbed by node j from node i is in represents the receiving power of receiving node j in link ij. In summary, the energy efficiency gain of the link sending node can be obtained as
[0049]
[0050] Where V i , , They represent the neighboring node set, the back-hop node set and the back-hop node set of node i that are ready for backscatter transmission (sufficient energy, only lacking carrier). The first part on the right side of the equation represents the weighting of the active transmission probability and its corresponding energy efficiency gain adopted by the sending node i, while the second part represents the weighting of the backscatter probability and its corresponding energy efficiency gain. This indicator can accurately and comprehensively measure the energy efficiency gain of each link sending node for the entire network. In summary, the evaluation index of the link should be measured by the average link delay and the energy efficiency gain of the link sending node. The average link delay should be as small as possible, while the energy efficiency gain of the sending node should be as large as possible.
[0051] The information flow and energy flow in the signal-energy coupling network are frequently transformed into each other, the link delay is unstable, and the energy efficiency gain of each link is different. The selection of the optimal path has a decisive influence on the improvement of network throughput and network performance. The entire wireless network can be modeled as a directed graph, in which the source node has only outgoing edges, that is, it only sends data packets but does not receive data packets, and the sink node has only incoming edges, that is, it only receives data packets but does not send data packets. The network adopts a semi-bandit feedback scheme, that is, the source node formulates a routing strategy and receives the revenue feedback of each link in the selected path. The process of transmitting data packets on each link can be approximated as a geometric distribution, that is, transmission failure will be retransmitted until success. In the initial state, the data transmission success rate of each link is unknown. As the number of transmitted data packets increases, each link obtains an estimated data transmission success rate. According to the definition of geometric distribution, link delay can be expressed as the inverse of the data transmission success rate.
[0052] Most current routing strategies use the shortest path delay as the measurement indicator of the optimal path, ignoring the energy efficiency gain of each link sending node in the path. This patent proposes a new optimal path evaluation method. Specifically, each link maintains its own evaluation index, which consists of link delay and energy efficiency gain. Similarly, the path also has a corresponding path index, which is the sum of the evaluation indicators of the links that make up the path.
[0053] Based on the above principle, in this embodiment, a backscatter wireless network optimal path routing method based on online learning is designed. Figure 2 ,The optimal path routing method for backscatter wireless network based on online learning includes the following steps:
[0054] S1. Obtain network parameters and operating parameters of each routing node in the backscatter wireless network;
[0055] S2. Using the KL-UCB algorithm, the link index value of the link between each two routing nodes is determined according to the network parameters and operating parameters of each two routing nodes;
[0056] S3. Determine the target link; the target link is the link with the smallest link index value among all links;
[0057] S4. Transmitting data through the target link to determine the actual path gain of the target link;
[0058] S5. Determine the empirical path benefit of the target link; the empirical path benefit is the total path benefit of transmitting data through the target link when the data transmission success rate of the target link is known.
[0059] S6. Determine the performance index value corresponding to the target link according to the difference between the actual path gain and the empirical path gain;
[0060] S7. When the performance indicator value reaches the performance threshold, the target link is determined as the optimal path.
[0061] In this embodiment, the backscatter wireless network includes multiple routing nodes, each routing node can choose to work in active transmission mode or backscatter transmission mode. Each routing node has its own corresponding network parameters and working parameters.
[0062] In step S1, the network parameters represent the topological information of the corresponding routing node in the backscatter wireless network, and the working parameters represent the working state of the corresponding routing node. Specifically, the network parameters may include the node set V, edge set E, path set PA of the backscatter wireless network where the routing node is located, and the neighbor node set and back-hop node set of the routing node; the working parameters may include the probability of the routing node choosing to use the active transmission mode when working, the probability of choosing to use the backscatter transmission mode, the energy consumed in the active transmission mode, the energy consumed in the backscatter transmission mode, the total number of data packets sent during a link communication, the number of bits contained in each data packet, the link benefit expectation and confidence interval, etc.
[0063] In step S2, the KL-UCB algorithm used is a deformation and generalization of the classic UCB algorithm. The KL-UCB algorithm determines the confidence interval through KL divergence. The index of the classic UCB algorithm is generalized to the SP-CMAB problem, and the algorithm idea is as follows: 1. Assume that the initial probability distribution of each link is the same, that is, the expectation is the same. 2. Each link has its confidence interval, the upper bound of the initial interval is the same, and the selection benefit of each link is likely to be within the confidence interval. Using the upper bound of the confidence interval of the expected benefit as the link index, the corresponding path index is the sum of the upper bounds of the confidence intervals of each link on the path. 3. Select a path with the largest path index for data transmission, and obtain link benefit feedback and path benefit feedback. If the benefit feedback of a certain link is less than the initial expectation, after recalculation and updating, the benefit expectation of the link will be less than the initial expectation. Similarly, if the link benefit feedback is greater than the initial expectation, the updated link benefit expectation is greater than the initial expectation. At the same time, due to the increase in the number of samples, the length of the confidence interval obtained by KL divergence becomes smaller. 4. Each time a data packet is sent, step 3 is repeated and the link index is updated. If a path is repeatedly selected, the confidence interval length of the link on the path will become smaller and smaller, and the corresponding path index may be smaller than other path indexes, which provides opportunities for the selection of other paths. When the number of sent data packets is large enough, the algorithm converges and tends to select the path with the highest benefit for data transmission.
[0064] When executing step S2, that is, using the KL-UCB algorithm to determine the link index value of the link between every two routing nodes according to the network parameters and working parameters of every two routing nodes, it can be specifically calculated by the following formula:
[0065]
[0066] ij represents a link between routing nodes numbered j, ω ij represents the link indicator value corresponding to link ij, n represents the number of data packets sent through link ij, It represents the vector composed of the corresponding data transmission success rates when data is transmitted through all links in the backscatter wireless network in the active transmission mode. In the backscatter transmission mode, data is transmitted through each link in the backscatter wireless network, and the corresponding data transmission success rate is a vector, t ij (n) represents the sum of the number of transmission attempts on link ij before transmitting the nth data packet through link ij, V i represents the set of neighbor nodes of routing node i in the backscatter wireless network, represents the set of back-hop nodes of routing node i in the backscatter wireless network, It represents the set of ready back-hop nodes of routing node i in the backscatter wireless network. Each node in the ready back-hop node set has completed the preparation of backscatter transmission conditions (for example, it has obtained sufficient energy to complete backscatter transmission, but lacks carrier and does not start backscatter transmission temporarily). I represents the number of bits contained in a data packet. and represents the parameters that satisfy the corresponding conditions in the formula, represents the probability that routing node i chooses to use active transmission mode, represents the probability that routing node i chooses to use the backscatter transmission mode, k, h and l represent the summation number, represents the empirical data transmission success rate corresponding to data transmission through link ij in active transmission mode, represents the empirical data transmission success rate corresponding to data transmission through link ij in the backscatter transmission mode, represents the energy consumed by routing node i in active transmission mode, represents the energy consumed by routing node i in the backscatter transmission mode, represents the receiving power of routing node l as the receiving node in link il in active transmission mode, represents the efficiency of routing node l absorbing energy from routing node i, KL() represents the KL divergence obtained by processing using the KL-UCB algorithm, and g(n) represents a function related to n. Specifically, g(n)=log n+4loglog n can be set. The link index represented by this formula is the upper bound of the confidence interval, and the confidence interval is derived from the KL divergence.
[0067] The ω calculated by the above formula ij , represents the link index value corresponding to the link between routing node i and routing node j, i.e., link ij, in the backscatter wireless network. For the entire backscatter wireless network, two routing nodes can be arbitrarily selected as routing node i and routing node j, respectively, so that the link index value corresponding to the link between the two routing nodes can be calculated by the above formula.
[0068] In this embodiment, when executing step S3, that is, the step of determining the target link, the following steps may be specifically performed:
[0069] S301. Determine a link set;
[0070] S302. Determine a target link from all the links included in the link set.
[0071] In step S301, when a link set is initially established, the link set includes all links of the backscatter wireless network. By executing subsequent steps, some links may be added or deleted from the link set, thereby changing the links included in the link set.
[0072] In step S302, for all the links included in the link set (the link set may be in the initial state, that is, including all the links of the backscatter wireless network, or some links may be deleted from it), each link can be used as the link ij in the formula to calculate the corresponding link index value ω ij , the link corresponding to the smallest link index value can be selected as the target link.
[0073] In step S4, data is transmitted through the target link, that is, the link corresponding to the smallest link index value. This process can be expressed by the following formula:
[0074] When executing step S4, that is, transmitting data through the target link and determining the actual path benefit of the target link, it can be specifically calculated by the following formula:
[0075]
[0076] in, represents the averaging operation, N represents the total number of data packets transmitted through the target link, and D(n) represents the actual path benefit brought by transmitting the nth data packet.
[0077] When executing step S5, that is, determining the empirical path benefit of the target link, it can be specifically calculated by the following formula:
[0078] By formula ND θ (p * ) Determine the experience path benefits;
[0079] in,
[0080]
[0081] p represents a link, p * represents the target link, D θ (p * ) in the calculation formula, and ω ij The same symbols involved in the calculation formula have the same meaning. Specifically, ω ij In the calculation formula of , routing node i and routing node j can represent any routing node, and D θ (p *) in the calculation formula, routing node i and routing node j can represent the two routing nodes connected by the target link.
[0082] When executing step S6, that is, determining the performance indicator value corresponding to the target link according to the difference between the actual path gain and the empirical path gain, it can be calculated by the following formula:
[0083]
[0084] In this embodiment, the performance indicator value R(N) corresponding to the target link is the actual path benefit Subtract the experience path gain ND θ (p * ) is the difference obtained.
[0085] The performance index value R(N) obtained by executing step S6 means the difference between the average of the total benefits of randomly transmitting data packets obtained by the routing method in this embodiment and the total benefits of transmitting data packets using the optimal path under the condition of knowing all the link information (link transmission data success rate). The performance index value R(N) can also be called regret and can be used as a performance indicator to measure the routing algorithm.
[0086] After obtaining the performance index value R(N) corresponding to the target link, step S7 may be performed to compare the performance index value R(N) with a preset performance threshold. If the performance index value R(N) is not less than the performance threshold, that is, the performance index value R(N) reaches the performance threshold, it indicates that the performance index value R(N) corresponding to the target link is large enough, and the target link may be determined as the optimal path in the backscatter wireless network.
[0087] In this embodiment, the principles of steps S1-S7 are:
[0088] (1) For wireless communication networks where information flow and energy flow are mutually converted, since the link delay is unknown, the routing problem of such wireless communication networks is modeled as a combined multi-armed bandit (CMAB) problem; the problem of online learning of optimal paths in a signal-energy coupled network solved by the present invention can be referred to as the SP-CMAB problem; the language description of the MAB problem can be expressed as follows: Suppose a player enters a casino and faces a row of slot machines with the same appearance; but the expected return and expected loss of each slot machine are different and unknown. The player hopes to obtain the highest return by playing the slot machine, and the slot machine selection strategy is required; and the CMAB problem is a variation of the MAB problem, that is, the player not only plays one slot machine at a time, but also plays another slot machine at a time. Instead of playing one slot machine, you play a group of slot machines, so as to receive the benefits of one group, and finally find the combination selection strategy of each group of slot machines; if the player is regarded as the maker of routing strategy, for a network, each link is regarded as a slot machine, the probability of successful data packet transmission is the probability of obtaining benefits, and a path from the source to the sink is a group of slot machines, then the SP-CMAB problem can be modeled as a combined multi-armed slot machine problem; when the source transmits a new data packet, the routing strategy maker accumulates the known path benefit feedback and instantly selects the path with the highest benefit for data transmission in the dynamically changing network, thereby improving network performance, increasing energy efficiency, and increasing data transmission rate;
[0089] (2) When the link delay is unknown and difficult to determine, an online learning method is proposed to obtain the optimal path, optimize the routing strategy, make up for the defect that the existing approximate algorithm is not suitable for dynamic wireless communication networks, improve network performance, and increase network throughput;
[0090] (3) The shortest path algorithm is improved by using a comprehensive evaluation index of link delay plus energy efficiency gain, which can more accurately find the optimal path in the wireless communication network, improve the data transmission rate, and take into account the network energy efficiency;
[0091] (4) For wireless communication networks with nodes using different transmission modes, the probability of each node selecting a different transmission mode is obtained respectively, and the probability is linked to the data transmission success rate. The average data transmission success rate of each outgoing link is weighted to obtain the average link delay. The average link delay is used as the link indicator to improve the stability of the network and increase the reliability of the model.
[0092] In this embodiment, refer to Figure 3 , the optimal path routing method for backscatter wireless network based on online learning also includes the following steps:
[0093] S8. When the performance indicator value does not reach the performance threshold, the target link is deleted from the link set, and the process returns to the step of determining the target link, that is, step S3 is re-executed.
[0094] Reference Figure 3 , when the performance index value R(N) calculated in step S6 is less than the performance threshold, that is, the performance index value R(N) does not reach the performance threshold, it indicates that the performance index value R(N) corresponding to the target link is not large enough, and the current target link cannot be determined as the optimal path in the backscatter wireless network, and step S8 is executed to delete the target link from the link set. In this way, the latest link set does not include the current target link, and a new target link can be screened out according to the rule of the smallest link index value.
[0095] After executing step S8, jump to step S3 and start again in the order of steps S3, S4, S5, and S6. By executing step S3, a new target link can be screened out from the latest link set. Then, the steps S4-S6 executed can determine the performance index value corresponding to the new target link and determine whether the new target link can be used as the optimal path in the backscatter wireless network. If the new target link can be used as the optimal path in the backscatter wireless network, then any step in steps S1-S8 can be no longer executed; if the new target link cannot be used as the optimal path in the backscatter wireless network, then continue to execute step S8, delete the new target link from the link set, and then jump to step S3 to start a new round of steps until the optimal path is selected from the updated link set, or the optimal path that meets the conditions cannot be determined. After executing multiple rounds of steps S3-S8, the link set becomes an empty set, and then any step in steps S1-S8 is no longer executed.
[0096] By executing step S8, when the optimal path is not found after executing steps S1-S7 once, other links in the backscatter wireless network can be further searched until the optimal path is determined, thereby increasing the possibility of finding the optimal path and improving the efficiency and success rate of the optimal path routing method for the backscatter wireless network based on online learning.
[0097] A computer program for executing the optimal path routing method for a backscatter wireless network based on online learning in this embodiment can be written and written into a computer device or a storage medium. When the computer program is read out and executed, the optimal path routing method for a backscatter wireless network based on online learning in this embodiment is executed, thereby achieving the same technical effect as the optimal path routing method for a backscatter wireless network based on online learning in the embodiment.
[0098] It should be noted that, unless otherwise specified, when a feature is referred to as "fixed" or "connected" to another feature, it may be directly fixed or connected to another feature, or it may be indirectly fixed or connected to another feature. In addition, the descriptions of up, down, left, right, etc. used in this disclosure are only relative to the relative positional relationship of the components of the present disclosure in the accompanying drawings. The singular forms of "a", "said" and "the" used in this disclosure are also intended to include the plural forms, unless the context clearly indicates other meanings. In addition, unless otherwise defined, all technical and scientific terms used in this embodiment have the same meaning as those generally understood by technicians in this technical field. The terms used in the specification of this embodiment are only for describing specific embodiments, not for limiting the present invention. The term "and / or" used in this embodiment includes any combination of one or more related listed items.
[0099] It should be understood that, although the term first, second, third etc. may be adopted to describe various elements in the present disclosure, these elements should not be limited to these terms. These terms are only used to distinguish the same type of elements from each other. For example, without departing from the scope of the present disclosure, the first element may also be referred to as the second element, and similarly, the second element may also be referred to as the first element. The use of any and all examples or exemplary language ("for example", "such as" etc.) provided by the present embodiment is only intended to better illustrate embodiments of the present invention, and unless otherwise required, the scope of the present invention will not be limited.
[0100] It should be appreciated that embodiments of the present invention may be implemented or enforced by computer hardware, a combination of hardware and software, or by computer instructions stored in a non-transitory computer-readable memory. The method may be implemented in a computer program using standard programming techniques - including a non-transitory computer-readable storage medium configured with a computer program, wherein the storage medium so configured causes the computer to operate in a specific and predefined manner - according to the methods and drawings described in the specific embodiments. Each program may be implemented in a high-level procedural or object-oriented programming language to communicate with a computer system. However, if desired, the program may be implemented in an assembly or machine language. In any case, the language may be a compiled or interpreted language. In addition, the program may be run on a programmed ASIC for this purpose.
[0101] In addition, the operations of the process described in this embodiment may be performed in any suitable order, unless otherwise indicated in this embodiment or otherwise clearly contradicted by the context. The process described in this embodiment (or variations and / or combinations thereof) may be performed under the control of one or more computer systems configured with executable instructions, and may be implemented as a code (e.g., executable instructions, one or more computer programs, or one or more applications) executed jointly on one or more processors, by hardware or a combination thereof. The computer program includes a plurality of instructions that may be executed by one or more processors.
[0102] Further, the method can be implemented in any type of computing platform that is operably connected to a suitable computer, including but not limited to a personal computer, a minicomputer, a mainframe, a workstation, a network or distributed computing environment, a separate or integrated computer platform, or in communication with a charged particle tool or other imaging device, etc. Various aspects of the present invention can be implemented in machine-readable code stored on a non-transitory storage medium or device, whether removable or integrated into a computing platform, such as a hard disk, an optical read and / or write storage medium, a RAM, a ROM, etc., so that it can be read by a programmable computer, and when the storage medium or device is read by the computer, it can be used to configure and operate the computer to perform the process described herein. In addition, the machine-readable code, or a portion thereof, can be transmitted via a wired or wireless network. When such media includes instructions or programs that implement the steps described above in conjunction with a microprocessor or other data processor, the invention described in this embodiment includes these and other different types of non-transitory computer-readable storage media. When programmed according to the methods and techniques of the present invention, the present invention also includes the computer itself.
[0103] The computer program can be applied to input data to perform the functions described in the present embodiment, thereby converting the input data to generate output data stored in a non-volatile memory. The output information can also be applied to one or more output devices such as a display. In a preferred embodiment of the present invention, the converted data represents a physical and tangible object, including a specific visual depiction of the physical and tangible object produced on the display.
[0104] The above is only a preferred embodiment of the present invention. The present invention is not limited to the above implementation. As long as the technical effect of the present invention is achieved by the same means, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the scope of protection of the present invention. Within the scope of protection of the present invention, its technical scheme and / or implementation method may have various modifications and changes.
Claims
1. An optimal path routing method for backscatter wireless networks based on online learning, It is characterized in that The optimal path routing method for backscatter wireless network based on online learning includes: Acquire network parameters and working parameters of each routing node in the backscatter wireless network; the routing node can choose to work in active transmission mode or backscatter transmission mode; the network parameters represent the topology information of the routing node in the backscatter wireless network, and the working parameters represent the working state of the routing node; Using the KL-UCB algorithm, determining the link index value of the link between each two of the routing nodes according to the network parameters and the operating parameters of each two of the routing nodes; Determine a target link; the target link is a link with the smallest link indicator value among all links; transmitting data through the target link to determine an actual path gain of the target link; Determine the empirical path gain of the target link; the empirical path gain is the total path gain of transmitting data through the target link when the data transmission success rate of the target link is known; Determining a performance indicator value corresponding to the target link according to a difference between the actual path gain and the empirical path gain; When the performance indicator value reaches a performance threshold, determining the target link as an optimal path; The method of using the KL-UCB algorithm to determine the link index value of the link between each two routing nodes according to the network parameters and the working parameters of each two routing nodes includes: By formula Determine the link indicator value; wherein, Indicates a number The routing node is associated with a The links between the routing nodes, Indicates link The corresponding link indicator value, Indicates that the link The number of the packet sent, represents a vector composed of the corresponding data transmission success rates when data is transmitted through all the links in the backscatter wireless network in the active transmission mode, represents a vector composed of the corresponding data transmission success rates when data is transmitted through each link in the backscattering wireless network in the backscattering transmission mode, Indicates that the link Transmission Before the packet, on the link The sum of the number of transfers attempted on Represents a routing node a set of neighbor nodes in said backscatter wireless network, Represents a routing node a set of back-hop nodes in the backscatter wireless network, Represents a routing node A set of ready back-hop nodes in the backscatter wireless network, each node in the set of ready back-hop nodes has completed backscatter transmission condition preparation, represents the number of bits contained in a data packet, and represents the parameters that satisfy the corresponding conditions in the formula, Represents a routing node The probability of choosing to use active transmission mode, Represents a routing node The probability of choosing to use the backscatter transmission mode, , and Indicates the sum number, Indicates that in active transmission mode, through the link The empirical data transmission success rate corresponding to data transmission, In backscatter transmission mode, the link The empirical data transmission success rate corresponding to data transmission, Represents a routing node The energy consumed in active transmission mode, Represents a routing node The energy consumed in the backscatter transmission mode, Indicates that in active transmission mode, the link The routing node as the receiving node The received power, Represents a routing node From the routing node The efficiency of absorbing energy, represents the KL divergence obtained by processing using the KL-UCB algorithm, Representation and Related functions; ; The transmitting data through the target link and determining the actual path benefit of the target link includes: By formula Determine the actual path benefit; wherein, It represents the average operation. represents the total number of packets transmitted through the target link, Indicates the transmission The actual path benefit brought by each data packet; The determining the empirical path benefit of the target link includes: By formula Determining the experience path benefit; in, Indicates a link, Indicates the target link.
2. The optimal path routing method for backscatter wireless network based on online learning according to claim 1, It is characterized in that The determining of the target link comprises: Determining a link set; the link set initially includes all links of the backscatter wireless network; The target link is determined from all links included in the link set.
3. The optimal path routing method for backscatter wireless network based on online learning according to claim 2, It is characterized in that The backscatter wireless network optimal path routing method based on online learning also includes: When the performance indicator value does not reach the performance threshold, deleting the target link from the link set; The process returns to the step of determining the target link and starts execution.
4. The optimal path routing method for backscatter wireless network based on online learning according to claim 1, It is characterized in that The determining, according to the difference between the actual path gain and the empirical path gain, a performance indicator value corresponding to the target link includes: By formula Determine the performance indicator value.
5. A computer device, It is characterized in that It comprises a memory and a processor, wherein the memory is used to store at least one program, and the processor is used to load the at least one program to execute the backscatter wireless network optimal path routing method based on online learning according to any one of claims 1 to 4.
6. A computer-readable storage medium having a program executable by a processor stored therein, It is characterized in that The program executable by the processor is used to execute the optimal path routing method for backscatter wireless network based on online learning as described in any one of claims 1 to 4 when executed by the processor.
Citation Information
Patent Citations
Backscatter devices and network systems incorporating backscatter devices
CN108141646A
Robust opportunistic routing protocol applied to environment backscatter wireless sensor network
CN109618383A
Load balancing regulation and control method applied to multi-hop environment backscattering wireless network
CN112738849A
Multi-hop BackCom network energy optimization routing scheme
CN113438706A
LPRF device wake up using wireless tag
US20090181625A1
