Wireless communication multi-queue rotation priority scheduling method for deterministic time delay

Abstract

The invention discloses a wireless communication multi-queue rotation priority scheduling method for deterministic time delay, and belongs to the field of communication. Comprising the following main characteristics: establishing a queue system of rotation priority scheduling, enabling a data packet to enter a corresponding queue according to the maximum tolerance time delay from the deadline, and rotating to the next queue after the scheduling at each moment is finished; considering the influence that the data packets in each queue are not scheduled to violate the deterministic time delay demand, and obtaining the weighted queue length of all the data packets of the user; scheduling oriented to deterministic time delay is carried out on the weighted queue, deterministic time delay constraint is guaranteed by establishing an interrupt virtual queue, and meanwhile wireless resource consumption is minimized. The method can be used for allocating different scheduling time according to the channel quality of different devices in an environment with a dynamic wireless channel change, and the beneficial effect of dynamic wireless network deterministic time delay guarantee is achieved.

Description

Technical Field

[0001] The present invention relates to the field of wireless communications, and in particular to a wireless communications multi-queue round-robin priority scheduling method oriented to deterministic delay. Background Art

[0002] With the development of industry, smart manufacturing and wireless factories have become the development direction of modern manufacturing. In this environment, traditional wired connection methods are gradually being replaced by wireless technology due to their installation complexity, maintenance costs and lack of flexibility. With the increase in the number of equipment on the production line and the integration of multiple types of equipment, the demand for real-time, stability and efficiency of wireless networks has also increased dramatically.

[0003] Priority Scheduling (PS) is a network scheduling strategy that sorts and transmits packets according to their priority. Deadline is a concept related to real-time systems that defines the last moment a packet or task must be completed or transmitted. If a packet is not successfully transmitted before its deadline, it will cause system failure or performance degradation.

[0004] In a factory environment, different devices and tasks have different requirements for the network. Some critical tasks, such as precise control of robots or safety monitoring, require a higher network priority than other tasks. Therefore, an effective priority scheduling mechanism is crucial to ensure that critical tasks can still obtain sufficient scheduling resources when network resources are tight.

[0005] There are two challenges in real-time scheduling on dynamic wireless networks. First, the channel quality of devices in different environments changes in real time, and real-time scheduling needs to be able to quickly adapt to changes in channel quality and make corresponding adjustments. Second, to ensure the real-time and high efficiency of wireless networks, scheduling algorithms need to consider multiple factors, such as data packet priority, deadline, current network status, etc. This increases the complexity of scheduling algorithms and may cause delays in calculations and decisions. Summary of the invention

[0006] In order to overcome the shortcomings of the prior art, the purpose of the present invention is to provide a multi-queue round-robin priority scheduling method for wireless communication facing deterministic delay. Following the basic idea of ​​scheduling data packets to optimize network performance, the scheduling strategy is optimized: on the one hand, a sequential and orderly queue is established, and the packet is assigned to the designated queue according to the deadline to distinguish the urgency of the scheduling, and then a weighted queue and an interrupt virtual queue are proposed to accurately measure the scheduling time; on the other hand, the soft threshold function is improved to solve the water injection and proximal accelerated descent algorithms with low complexity to obtain the optimal scheduling time of the strategy. The scheduling strategy can obtain the best scheduling time for each device without violating the delay constraint.

[0007] The technical scheme for achieving the purpose of the present invention is as follows:

[0008] A wireless communication multi-queue round-robin priority scheduling method for deterministic delay, comprising the following steps: establishing a queue system for round-robin priority scheduling, wherein data packets enter corresponding queues according to the maximum tolerable delay from the deadline, and rotate to the next queue after the scheduling is completed at each moment; considering the impact of data packets in each queue not being scheduled on the violation of the deterministic delay requirement, obtaining the weighted queue length of all user data packets; scheduling the weighted queue for deterministic delay, ensuring the deterministic delay constraint by establishing an interrupt virtual queue, and minimizing the wireless resource consumption.

[0009] Preferably, a wireless communication multi-queue round-robin priority scheduling method for deterministic delay comprises the following steps:

[0010] 1) Obtain the following information: total measurement time T. The total number of devices N at the current time t (t∈T); device i ( is the channel quality C of a set of N devices i (t), the average number of packets arriving λ i ; Maximum tolerable delay D of data packets, arrival rate of data packets a i (t); Get network delay requirement∈ i ; Get the time slot length T slot ;

[0011] 2) Establish a queue system with round-robin priority scheduling: establish 2M queues according to twice the number of the maximum tolerable delay that may occur, and the queue subscripts are 0, 1, 1 in sequence + , ..., (M-1) + , M; where queue 0 is used to store discarded data packets that have not been scheduled out before the deadline.

[0012] The data packet enters the corresponding queue according to the maximum tolerable delay D from the deadline. If m-1≤D<m, it enters queue m. Newly arrived data packets will not enter m + Queue, the data packets in queue m+1 that were not sent at the last moment are transferred to queue m + , and then merged with the newly arrived data packets in queue m.

[0013] At time t, the number of packets of device i in queue m is The number of packets dropped is Q i，0 (t); Calculate the time T required to schedule all the data packets of device i in queue m at time t i，m (t):

[0014]

[0015] 3) Use β m Indicates the number of packets lost in the future due to unscheduled packets in each queue Q i，0 (t), according to the current packet loss benefit of -1, the benefit of continuing to wait is -β m , the benefit of successful sending is 1, and the Q-learning method is used to learn the β corresponding to the minimum number of packet loss m According to the β of different cohorts m The weighted queue length of all packets of user i is designed to be:

[0016]

[0017] Time mapping is performed based on weighted time:

[0018]

[0019] 4) Schedule weighted queues with deterministic delay to minimize wireless resource consumption.

[0020]

[0021] The constraints are as follows:

[0022]

[0023] In order to ensure long-term deterministic delay constraints, an interrupt virtual queue H is used. i (t) The dynamic changes of the interrupt virtual queue are:

[0024] H i (t+1)=max{H i (t)-∈ i λ i +Q i，0 (t), 0}

[0025] Among them, ∈ i is the network delay constraint, λ i is the average number of packets arriving.

[0026] Using Lyapunov's method, we add the long-term delay constraint to the current moment and optimize the following problems:

[0027]

[0028] The parameter V represents the trade-off between queue backlog and resource occupancy.

[0029] Beneficial effects of the present invention:

[0030] The present invention optimizes the scheduling of resources in dynamic wireless networks, follows the basic idea of ​​scheduling data packets to optimize network performance, optimizes the scheduling strategy, and allocates different scheduling times according to the channel quality of different devices with low complexity, minimizing the allocated time resources without violating the delay constraint. The scheduling strategy ensures network performance and meets industrial network requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] FIG1( a ) is a time resource allocation tendency of the algorithm proposed by the present invention after comprehensively considering the deadline and the device channel status;

[0032] Figure 1(b) is a comparison of the time resource allocation rate of the algorithm proposed by the present invention and the other three algorithms under the delay constraint guarantee. Figure 2(a) is a comparison of the algorithm proposed by the present invention and the other three algorithms under the delay constraint guarantee at different data packet arrival rates;

[0033] FIG2( b ) is a comparison of the algorithm proposed in the present invention with other three algorithms under the guarantee of delay constraint in terms of different numbers of devices;

[0034] FIG2(c) is a comparison of the packet loss rate of the algorithm proposed in the present invention and the other three algorithms under different signal-to-noise ratios of different data packets under delay constraint guarantee.

[0035] FIG3( a ) is a comparison of the time complexity between the algorithm proposed in the present invention and the lead time normalization-largest deficit first (L-LDF) algorithm under delay constraint guarantee.

[0036] FIG3( b ) is a comparison of the number of iterations between the algorithm proposed in the present invention and the lead time normalization-largest deficit first (L-LDF) algorithm under delay constraint guarantee. DETAILED DESCRIPTION

[0037] The present invention is further described below in conjunction with the accompanying drawings and examples. It is worth noting that the following examples are only used to illustrate the present invention and are not used to limit the present invention.

[0038] A wireless communication multi-queue round-robin priority scheduling method for deterministic delay, the steps are as follows:

[0039] Step 1: Obtain the following information: total measurement time T. The total number of devices N at the current time t (t∈T); device i ( is the channel quality C of a set of N devices i (t), the average number of packets arriving λ i ; Maximum tolerable delay D of data packets, arrival rate of data packets a i (t); Get network delay requirement∈ i ; Get the time slot length T slot ;

[0040] Step 2: Establish a queue system with round-robin priority scheduling: Establish 2M queues according to twice the maximum tolerated delay that may occur, and the queue subscripts are 0, 1, 1 in sequence + , ..., (M-1) + , M; where queue 0 is used to store discarded data packets that have not been scheduled out before the deadline.

[0041] Step 3: The data packet enters the corresponding queue according to the maximum tolerable delay D from the deadline. If m-1≤D<m, it enters queue m. Newly arrived data packets will not enter m. + Queue, the data packets in queue m+1 that were not sent at the last moment are transferred to queue m + , and then merged with the newly arrived data packets in queue m.

[0042] At time t, the number of packets of device i in queue m is The number of packets dropped is Q i，0 (t); Calculate the time T required to schedule all the data packets of device i in queue m at time t i，m (t):

[0043]

[0044] Step 4: Calculate the delay constraint and time slot constraint, perform mathematical modeling and construct the optimization problem:

[0045] The optimization problem is constructed as:

[0046]

[0047] The constraints are as follows:

[0048]

[0049]

[0050] Step 5: Use β m Indicates the number of packets lost in the future due to unscheduled packets in each queue Q i，0 (t), according to the current packet loss benefit of -1, the benefit of continuing to wait is -β m , the benefit of successful sending is 1, and the Q-learning method is used to learn the β corresponding to the minimum number of packet loss m According to the β of different cohorts m The weighted queue length of all packets of user i is designed to be:

[0051]

[0052] Time mapping is performed based on weighted time:

[0053]

[0054] Step 6: Schedule the weighted queues with deterministic delay to minimize wireless resource consumption.

[0055]

[0056] The constraints are as follows:

[0057]

[0058] In order to ensure long-term deterministic delay constraints, an interrupt virtual queue H is used. i (t) The dynamic changes of the interrupt virtual queue are:

[0059] H i (t+1)=max{H i (t)-∈ i λ i +Q i，0 (t), 0}

[0060] Among them, ∈ i is the network delay constraint, λ i is the average number of packets arriving.

[0061] Using Lyapunov's method, we add the long-term delay constraint to the current moment and optimize the following problems:

[0062]

[0063] The parameter V represents the trade-off between queue backlog and resource occupancy.

[0064] As a better implementation example, we set the following simulation parameters: We use a wireless network for data transmission and deploy a scheduler at its central location. The arrival of data packets is Poisson distributed. The average arrival rate varies randomly between 8Kbps and 16Kbps. In addition, the number of devices N is between 2 and 20. The channel capacity varies according to the Rayleigh distribution model. The deadline of the data packet is randomly generated in the range of 1 to 10 time slots, and the duration of each time slot is 0.05 seconds. The scheduler needs to send these data packets to the device before their respective deadlines. The computer simulation is run 1000 times and the average value is taken as the final result. Three comparison methods are introduced, namely the polling algorithm (RR), the deadline priority algorithm (EDF) and the lead time normalization-maximum deficit priority (L-LDF). The advantages of the algorithm proposed in the present invention are illustrated by performance comparison.

[0065] Time resource allocation tendency, as shown in Figure 1(a), the present invention comprehensively considers the deadline and channel quality. When the deadline is less than four, about 84.5% of the time will be allocated. When the data packet is not urgent, the allocated time resources are proportional to the channel quality. The better the channel quality, the more time is allocated.

[0066] As shown in FIG1( b ), compared with the three comparison algorithms, the algorithm proposed in the present invention achieves the least resource occupancy while satisfying the delay deadline constraint, and provides greater flexibility for the time resource scheduling of non-urgent data packets.

[0067] The packet loss rate comparison, as shown in Figure 2(a), Figure 2(b), and Figure 2(c), shows that the algorithm proposed in the present invention has obvious advantages over the three comparison algorithms in different packet arrival rates, different numbers of users, and different signal-to-noise ratios. The performance advantage of the algorithm of the present invention is more significant in the three cases of high packet arrival rate, large number of users, and low signal-to-noise ratio.

[0068] The comparison of time complexity and number of iterations is shown in Figure 3(a) and Figure 3(b). Compared with the L-LDF algorithm, the algorithm proposed in the present invention has obvious advantages in time complexity and number of iterations.

[0069] Based on the above performance comparison, the present invention follows the basic idea of ​​scheduling data packets to optimize network performance for tasks in a dynamic wireless environment, comprehensively considers the data packet deadline and device channel quality, obtains the optimal scheduling time for each device with low computing overhead and low time resource overhead, and achieves the beneficial effect of transmission performance of dynamic wireless networks.

[0070] The above-mentioned embodiments only express several implementation modes of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the invention. It should be pointed out that, for a person of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the attached claims.

Claims

1. A wireless communication multi-queue round-robin priority scheduling method for deterministic delay, characterized in that the steps include: A queue system with round-robin priority scheduling is established. Data packets enter the corresponding queue according to the maximum tolerable delay from the deadline, and rotate to the next queue after each scheduling is completed. The impact of unscheduled data packets in each queue on the violation of deterministic delay requirements is considered to obtain the weighted queue length of all user data packets. The weighted queue is scheduled for deterministic delay, and the deterministic delay constraint is guaranteed by establishing an interrupt virtual queue, while minimizing wireless resource consumption.

2. According to claim 1, a wireless communication multi-queue round-robin priority scheduling method for deterministic delay is characterized in that: The steps include: 1) Obtain the following information: total measurement time T; total number of devices N at the current time t, where t∈T; channel quality C of device i i (t), where is a set of N devices; the average number of packets arriving is λ i ; Maximum tolerable delay D of data packets, arrival rate of data packets a i (t); Get network delay requirement∈ i ; Get the time slot length T slot ; 2) Establish a queue system with round-robin priority scheduling: establish 2M queues according to twice the maximum tolerable delay that may occur, and the queue subscripts are 0, 1, 1 in sequence + ,…,(M-1) + ,M; where queue 0 is used to store the discarded data packets that have not been scheduled out before the deadline; Data packets enter the corresponding queues according to the maximum tolerable delay D from the deadline; if m - 1 ≤ D < m, they enter queue m; newly arrived data packets do not enter queue m + Queue. Packets in queue m + 1 that were not sent in the previous moment are transferred to queue m + , and then merged with newly arrived data packets in queue m; At time t, the number of packets of device i in queue m is The number of packets dropped is Q i,o (t); Calculate the time T required to schedule all the data packets of device i in queue m at time t i,m (t): 3) Use β m Indicates the number of packets lost in the future due to unscheduled packets in each queue Q i,0 (t), according to the current packet loss benefit of -1, the benefit of continuing to wait is -β m , the benefit of successful sending is 1, and the Q-learning method is used to learn the β corresponding to the minimum number of packet loss m value; according to the β of different cohorts m The weighted queue length of all packets of user i is designed to be: Time mapping is performed based on weighted time: 4) Schedule weighted queues with deterministic delay to minimize wireless resource consumption; The constraints are as follows: In order to ensure long-term deterministic delay constraints, an interrupt virtual queue H is used. i (t) The dynamic changes of the interrupt virtual queue are: H i (t+1)=max{H i (t)-∈ i λ i +Q i,0 (t),0}; Among them, ∈ i is the network delay constraint, λ i is the average number of packets arriving; Using Lyapunov's method, we add the long-term delay constraint to the current moment and optimize the following problems: The parameter V represents the trade-off between queue backlog and resource occupancy.

Images