CN113194499B - Authorization-free self-adaptive transmission method based on user estimation - Google Patents
Authorization-free self-adaptive transmission method based on user estimation Download PDFInfo
- Publication number
- CN113194499B CN113194499B CN202110362521.1A CN202110362521A CN113194499B CN 113194499 B CN113194499 B CN 113194499B CN 202110362521 A CN202110362521 A CN 202110362521A CN 113194499 B CN113194499 B CN 113194499B
- Authority
- CN
- China
- Prior art keywords
- user
- estimation
- users
- user load
- rbs
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W24/00—Supervisory, monitoring or testing arrangements
- H04W24/08—Testing, supervising or monitoring using real traffic
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W72/00—Local resource management
- H04W72/50—Allocation or scheduling criteria for wireless resources
- H04W72/52—Allocation or scheduling criteria for wireless resources based on load
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention discloses an authorization-free self-adaptive transmission method based on user estimation, which comprises the following steps: 1) The base station observes the occupation condition of the RB after the end of each time slot, namely the number of the conflict-free, conflict-free and idle RBs, and carries out user load estimation through the occupation condition of a channel and a Markov process; 2) The method can monitor the load condition of the user side in real time, adjust the resource allocation strategy, and meet the scene requirement so as to improve the capability of the system for coping with sudden change of the user side.
Description
Technical Field
The invention belongs to the field of information transmission, and relates to an authorization-free self-adaptive transmission method based on user estimation.
Background
Ultra-reliable low-delay communication (URLLC) is one of three application scenes of a fifth generation mobile communication technology (5G), and aims to support ultra-reliable ultra-low-delay transmission service required by industrial Internet and vehicle communication. The 5G URLLC marking performance indexes comprise: the air interface time delay of a user plane between a base station and a terminal is 1ms; meanwhile, for the transmission of a short 32-byte data packet, the reliability should reach 99.999% on the premise that the user plane delay is lower than 1 ms.
In the conventional cellular communication system, authorization transmission is adopted, and a user and a base station need to interact with each other, so that the base station can perform communication after completing authorization of the user. The interaction process obviously introduces a delay of significantly more than 1 ms. The delay mainly comes from two aspects, namely 'waiting delay' from the time when a user sends a data packet to a transmitter until the data packet starts to be transmitted on a channel, and 'transmission delay' from the time when the data packet starts to be transmitted on the channel to the time when the data packet is finally transmitted successfully. The "waiting time delay" is because the transmitter still has to wait for the base station to send the grant after transmitting the preamble, and this handshaking procedure will have a bad influence on the time delay. In order to meet the requirement of the user on time delay, authorization-free transmission is taken as one of key technologies, a user side directly transmits data packets without waiting for the base station to transfer authorization, and the 'waiting time delay' is avoided, so that the total time delay is greatly reduced.
A common unlicensed transmission scheme always allocates a fixed number of resources to a user in advance, but causes serious resource waste when the user side is idle; if the request of the user side is increased rapidly in a short time due to power failure recovery or the access distribution of the user is changed, pre-allocated resources are hard to be sufficient, so that time delay or reliability cannot meet the requirements of the scene. In addition, the base station cannot know the number of the users really active at the user side before, and can only count the number of the users successfully transmitted, so that the base station is difficult to timely feed back a bad transmission result.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provides an authorization-free self-adaptive transmission method based on user estimation, which can monitor the load condition of a user side in real time, adjust a resource allocation strategy and meet the scene requirement so as to improve the capability of a system for coping with sudden change of the user side.
In order to achieve the above object, the method for unlicensed adaptive transmission based on user estimation according to the present invention comprises the following steps:
1) The base station observes the occupation condition of the RB after the end of each time slot, namely the number of the RBs without conflict, with conflict and free, and carries out user load estimation through the occupation condition of a channel and a Markov process;
2) Carrying out self-adaptive resource allocation according to the user load estimated in the step 1), and then carrying out information transmission according to the result of the self-adaptive resource allocation to finish the authorization-free self-adaptive transmission based on the user estimation.
The specific operation of the step 1) is as follows: user load estimation in any time slot, user load estimation of optional K repetition and user load estimation of adjacent K repetition.
The specific operation of the user load estimation in any time slot is as follows:
setting T as the number of time slots divided according to subcarrier intervals in 1ms, W as the number of available RBs, A, B and C respectively representing the number of non-conflicted, conflicted and idle RBs observed in a certain time slot, N representing the user load parameter to be estimated, and setting
Representing the estimation result, calculating a maximum a posteriori with respect to the measurement resultEstimating, then converting into maximum likelihood estimation according to the maximum posterior estimation, and then integrating Markov process to obtain
Where P represents a transition matrix of the markov process, and ((0, w) → (a, B, C)) represents transition probabilities between the (0, w) state and the (a, B, C) state in the P matrix after the N-step transition.
The specific process of the user load estimation of the optional K repetition is as follows:
calculating the load condition of the user side within 1ms
Wherein the content of the first and second substances,
calculating the transition probability from (A, B, C) to any state in the state space S as:
calculating a transition matrix P according to the formula (1-10), and substituting the calculated transition matrix P into the formula (1-6) to obtain P (0,0,WT)→(A,B,C) Maximum size
The process of the user load estimation of the adjacent K repetition is as follows:
determining the number phi of users starting to transmit for the first time in each time slot;
calculating estimated user coincidences
Comprises the following steps:
step 2) during specific operation, firstly, a distribution scene at a user side is judged to be one of a burst-type scene, a stable scene and a Beta distribution scene, wherein the burst-type scene belongs to a delay tolerant scene, the latter two scenes belong to a delay sensitive scene, then, the resources are adaptively allocated according to the user load estimated in the step 1), and then, information transmission is carried out according to the result of the adaptively allocated resources.
The invention has the following beneficial effects:
in the specific operation of the authorization-free self-adaptive transmission method based on user estimation, the base station observes the RB occupation condition after each time slot is finished, then carries out user load estimation based on the channel occupation condition and the Markov process so as to monitor the user side load condition in real time, and then carries out self-adaptive resource allocation according to the user load obtained by estimation so as to meet the scene requirement and improve the capability of the system for dealing with user side sudden change.
Drawings
FIG. 1 is a diagram illustrating simulation results of the first embodiment;
FIG. 2 is a diagram illustrating simulation results of the second embodiment;
fig. 3 is a flow chart of the operation of the present invention.
Detailed Description
The invention is described in further detail below with reference to the accompanying drawings:
referring to fig. 1, the method for unlicensed adaptive transmission based on user estimation according to the present invention includes the following steps:
1) User load estimation through channel occupancy and Markov process
The base station observes the RB (Resource Block) occupancy after the end of each slot, i.e. the number of RBs that are not in conflict, have conflicts, and are free. If no conflict exists, it indicates that there is one and only one user in the RB that can successfully transmit; if there is conflict, it indicates that there are two or more users occupying the RB, and all of them are considered as transmission failure; and in the idle state, the RB is not occupied by users, and the user load in the time slot is estimated by using Markov according to the occupation condition of the RB.
11 Estimate user load in any slot
Let T be the number of slots divided according to the subcarrier spacing within 1ms, W be the number of available RBs, A, B and C respectively represent the number of RBs without collision, with collision and idle observed in a certain slot, N represents the user load parameter to be estimated, the occupancy is represented as a specific state (A, B, C), and then A + B + C = W. Although this is a fixed fact observed, it is not considered as a result of a Markov process with N transitions from a fixed initial state, and each user's choice is considered as a one-step transition, let
Represents an estimate whose maximum a posteriori estimate with respect to the observation is:
and transforming the maximum posterior estimation into maximum likelihood estimation, and combining a Markov process to obtain:
wherein P represents a transition matrix of the markov process, N represents the nth power of P, ((0, w) → (a, B, C)) represents transition probabilities between (0, w) states and (a, B, C) states in the P matrix after N steps of transition, and then the transition matrix P is calculated, specifically, a state space S of the markov process is generated according to equation (1-3), wherein different values of a, B, and C correspond to different state groups, (a, B, C) e S:
for any state (a, B, C) in the state space S, the possible outcome and probability of the next transition are:
the practical meanings of the formulae (1 to 4) are: when the next user selects in the W RBs, the RB which has conflict, the original idle RB or the original successful RB is selected; the selected results are the total state unchanged, the decrease of idle with successful increase and the decrease of successful with conflicting increase, respectively, and the values of the elements in the transition matrix P are calculated according to the probabilities listed in equations (1-4).
12 User load estimation of optional K-repeats
Observation of T continuous time slots, user load estimated by each time slot is n t (T is more than or equal to 1 and less than or equal to T), and the load condition of the user side within 1ms is obtained
For a transmission scheme with K repetitions,
comprises the following steps:
however, the estimation result of this method is not very accurate, and it is difficult to be qualified for the scenario that the resource allocation decision depends strictly on the user load, so on the basis of the above markov process analysis, the present invention jointly considers all available resources within 1 ms. Similar formula (1-2) can write a total formula of 1ms joint consideration under the K repeat transmission scheme, where (a, B, C) represents the total RB occupation observed by the base station after the end of the millisecond, and there are:
firstly, a state space S of the Markov process is generated again according to a formula (1-7), wherein different values of a, b and c correspond to different state groups, and (a, b, c) is epsilon to S:
using (A, B, C) to represent the current state,
representing the next state after the transition, in the context of the optional K retransmission scheme, considering in units of a single user, each user is considered to have done K selections at a time, (A, B, C) and
there exists a relationship shown by the formula (1-8):
let (A, B, C) state transition to
If the current RB is the conflict, the user selects x RBs from the A successful RBs, and the number of the conflicted RBs is increased; y are selected from the C free RBs so that successful RBs are increased. After this turnaround, the successful RB was
The conflicted RB is
The remaining free RBs are
Finally, assuming z is selected from the B RBs that would otherwise collide, z = K-x-, i.e.:
is provided with
Partial states satisfying the equations (1-8) are represented, and the transition probability from (a, B, C) to any state in the state space S is calculated as:
calculating a transition matrix P according to the formula (1-10), and substituting the calculated transition matrix P into the formula (1-6) to obtain P (0,0,WT)→(A,B,C) Maximum size
13 User load estimation of neighbor K repetitions
The difference from the optional scheme is that the neighbor scheme specifies that K attempts by the user must occur in consecutive K slots, and therefore the estimation scheme under the optional scheme will no longer work and a new framework must be established. Still jointly consider the occupation situation within 1ms, but it needs to record the occupation situation within each slot in three vectors, let T denote the number of slots divided according to the subcarrier spacing within 1ms, and the a, B, C matrices record the numbers of RBs observed in each slot without collision, with collision and idle, respectively, where:
A＝[A 1 ,2 A ,…,A T ]
B＝[B 1 ,B 2 ,…,B T ]
C＝[C 1 ,C 2 ,…,C T ]
let n t (1 ≦ T ≦ T) represents the estimated user load per slot, n = [ n ] 1 ,n 2 ,…,n T ]，φ r Represents the number of users starting the first transmission in the r-th time slot, phi =[φ 1 ,φ 2 ,…,φ T-K+1 ]. After the number of users phi which start to transmit for the first time in each time slot is determined, the number n of users which are transmitting in each time slot within 1ms can be determined, and the data packet of each user is repeatedly transmitted for K times:
n is the parameter to be estimated, each N determines the sample space of phi,
for the most likely user load, then equation (1-6) can be rewritten as:
the front part of equations (1-12) represents the probability that each set φ produces the observed (A, B, C) combination, i.e., the Markov process goes through n t Calculating the result of step transfer by formula (1-4); the latter part of equations (1-12) indicates that for each possible combination of phi there is a different probability of occurrence, such that
Then there are:
the above calculation method has a drawback in that N users share (T-K + 1) N A different choice, therefore, when the number of users is large, it will result in too large a sample space of phi to be calculated. The invention provides another method: assuming that phi and n respectively represent the number of users in a certain time slot that first try and the corresponding user load estimation in the time slot, the equivalence relation between phi and n can be expressed as a matrix equation:
the left T × (T-K + 1) of the formula (1-14) is represented by 01, and the matrix is simplified for the formula (1-14) and operated as follows:
the estimated user load is then:
2) Adaptive resource allocation based on user load estimation
21 User distribution
According to the method, the distribution scenes of the user side are summarized into a burst type scene, a stable type scene and a Beta distribution type scene, wherein the burst type scene is often caused by wireless link interruption caused by power failure, and all users are accessed simultaneously during recovery, so that serious collision conflict is caused; users who have not successfully accessed within the first millisecond will continue to attempt at the second millisecond and later until all users are successful. The users normally remain in a smooth profile, i.e. all users are in a fixed period of time T a The access is completed internally, and the number of activated users in each millisecond is the same. At some special time, the user distribution will change from stationary type to Beta distribution, and also at a fixed time T a Number of users N activated in each millisecond with access completed internally i Comprises the following steps:
where N denotes the total number of users and B (, β) denotes a Beta function with parameters α and β, i.e.
B(α，β)＝∫ 0 1 t α-1 (1-t) β-1 dt (2-2)
The base station can monitor the number of the activated users in each time slot, so that the distribution situation of the user side at the moment is judged according to the historical records:
2a) When the number of the monitored users is the same for a plurality of milliseconds continuously
The distribution is considered to be smooth.
2b) If the number of users in a certain millisecond is far less than the historical record value
If the user load is the first millisecond after the user is converted into the Beta distribution, the base station calculates alpha and Beta parameters in a Beta distribution formula (2-1) according to the user load monitored in the first two milliseconds, and then predicts the future (T) a -2) user load within milliseconds.
2c) If the user load monitored by the base station is 0 (N) within a certain millisecond i = 0) and the user load is detected to be much greater than the historical value within a certain millisecond at recovery time
It is a burst type scene.
In a uRLLC service scene which is one of three 5G scenes, the requirements on low time delay and high reliability are met, but the emphasis points under different user distributions are different. Burst-type distribution is considered delay tolerant because its ultimate goal is to restore normal traffic, and is not limited to completing transmission in a very short time; in addition, the stationary type and the Beta distribution type aim at transmitting user data packets, need to reach the technical index of 1ms delay as much as possible, and are considered as delay sensitive type.
22 ) adaptive scheme
221 Delay tolerant type)
In a burst-type scenario, although the need can be solved by allocating a greater number of RBs to all users, in the later stage of the recovery process, the number of users still active due to frequent failures is greatly reduced, and at this time, if a greater number of RBs are still occupied, resource waste is caused. In addition, it is also possible to achieve a lower recovery delay if RBs are allocated according to the real-time situation during recovery. Therefore, to implement adaptive allocation, when the base station determines that it is in a burst-type distribution according to the user condition of the first slot, it immediately determines the RB allocation scheme of the next slot. And after each subsequent time slot is ended, the base station allocates a certain number of RBs for the next time slot according to the monitored interval of the user load until the recovery is finished.
In a general case, it is desirable that the number of retransmissions K is taken into account. In the invention, the 120kHZ subcarrier interval is taken as an example, 1ms is divided into 8 time slots, and the maximum number of times of repeated transmission of each user in 1ms is 8. Taking an optional K repetition scheme as an example, after simulation is performed by a monte carlo method, it is found that the selection of the optimal K has a relationship with both the number of users and the number of RBs, and changes of recovery delays corresponding to the change of the users, RBs and retransmission times are shown in table 1:
TABLE 1
Table 1 records recovery delays (ms) corresponding to K different transmission schemes when the number of users and the number of RBs are different. Wherein UE represents the number of users, W represents the number of RBs, K best In order to represent the best retransmission times, the game problem exists between the retransmission times K and the recovery time delay. When K is small, the time delay may be shortened because collision is relatively small, or may be difficult to succeed because the number of user attempts is small, resulting in a long time delay.
Let the ratio of the number of users to the number of RBs available
Characterize how busy, record
In particular, user ratio
When the mobile station is in an idle state, the retransmission times can be properly increased, and the collision is not severe;in the congested state, the retransmission times should be reduced as much as possible, and at this time, the delay difference corresponding to different retransmission times is obvious and far greater than the difference in the idle state. The busy degree is classified according to the ratio, and corresponding K values are selected according to different grades, as shown in Table 2:
TABLE 2 user ratio vs. proposed number of retransmissions
K in Table 2 av Indicating the number of repeated transmissions suggested by the ratio of the number of users to the number of RBs during the recovery process. However, in a burst-type scenario, the number of general users may far exceed the number of RBs to form extreme congestion, default K =2, and table 2 may also be expressed by an approximate formula:
specifically, the RB allocation scheme is obtained in advance by simulation. Since the recovery process is finally expected to end within 6ms, if 3ms is set as the delay criterion and the required number of RBs is sought for each quantity interval, the final recovery process will be slightly higher than 3ms, and the 6ms criterion can be satisfied, the optional delay-tolerant resource allocation scheme is shown in table 3, and the adjacent delay-tolerant resource allocation scheme is shown in table 4:
TABLE 3
TABLE 4
Tables 3 and 4 show the user interval and the number of RBs recommended to be allocated under the optional K repetition scheme and the sensitive K repetition scheme, respectively, where N represents the interval where the number of users is located, and W represents the interval where the number of users is located av RepresentIn the recovery process, the number of RBs recommended to be allocated when the remaining number of users is in the corresponding user interval is represented by an approximate formula:
222 Time delay sensitive type
In the above discussion of delay tolerant schemes, the delay is more loosely constrained in view of the large number of users flooding and their intent to recover. But in the stationary and Beta distributions, it is desirable to be as close as possible to the criteria for a uRLLC scene for 1ms delay and 99.999% reliability. The number of newly activated users in each millisecond in the stable distribution is the same, a fixed number of RBs can be allocated to the newly activated users, and the user conditions of retransmission due to previous failure can be slightly adjusted; the number of newly activated users in each millisecond of Beta distribution changes, and the users retransmitted due to failure need to be allocated with RBs in real time.
Specifically, first, the influence of the retransmission times K on the success probability within 1ms under the condition of different user ratios is discussed, and taking an optional K repetition scheme as an example, the monte carlo simulation results are listed as follows:
TABLE 5
In Table 5, UE denotes the number of active users per millisecond, W denotes the number of available RBs, K best Indicating the number of retransmissions with the lowest total probability of failure for the user.
Observation of K best It can be seen that K, which performs best when the average number of occupied RBs is large, is concentrated around its maximum value, and K =8 is not the best choice if the number of available RBs is equal to the number of users. However, the reliability of 99.999 percent needs to be satisfied, namely the probability of transmission failure is controlled to be 10 -5 However, if the number of available RBs and the number of users are kept constant, the failure probability is much higher than that, so that this case is not realistic. To sum up, default K =8. Adaptation under delay-sensitive distributionThe scheme should be described as:
judging whether the current user side distribution is in a stable type or a Beta type according to the historical record and the user load monitored by the base station in the last time slot, and if the current user side distribution is in the Beta type, calculating key parameters of the Beta distribution after continuously observing for two milliseconds, and predicting the future T in advance a-2 Load trend within milliseconds; if the load is smooth, the last millisecond detected data is used as the future load prediction.
Let P i Representing the number of users activated in the ith millisecond predicted in the previous step, F i-1 Representing the number of users that have failed the last millisecond and need to be retransmitted, n i Representing the total number of users estimated by the ith millisecond base station according to the Markov process, S i Representing the number of successful users, W, counted by the ith millisecond base station i Representing the number of RBs to be allocated for the ith millisecond, then
Example one
A certain base station is set to serve 200 users in total, and the subcarrier interval is 120KHz. When power supply is recovered in a certain power failure accident, all users can simultaneously initiate a recovery access request in a certain same and fixed transmission scheme; if the recovery is not successful, the request is continuously initiated in the next millisecond, and the user who successfully recovers does not initiate any more. In the first millisecond, the base station matches the corresponding amount of resources according to the historical record value of the user; after each next millisecond is finished, the base station estimates the total number of the active users in the previous millisecond by using the invention, and subtracts the counted number of the successful users in transmission to finally obtain the total number of the possible active users in the next millisecond, and then the resources are distributed according to the invention.
In order to verify the superiority of the scheme, monte Carlo simulation is carried out on the scene in Matlab software, four schemes of optional type self-adaption K repetition, adjacent type self-adaption K repetition, optional type non-self-adaption K repetition and adjacent type non-self-adaption K repetition are compared, wherein the latter two schemes are used as a comparison group, and the simulation result is shown in figure 1.
K =2/K =3 in fig. 1 indicates that each user repeatedly tries 2/3 times of access within 1ms, and it is worth noting that K =2 is the best choice in this scenario, so K =3 functions as the control group at this time. In addition, in order to control the resource utilization rate variables to be consistent, the resource allocation of the two non-adaptive schemes is the average value of the adaptive schemes, and the average value is 15RB/ms. The trend in the figure shows that the recovery efficiency under the adaptive scheme is much higher than that of the non-adaptive scheme, the recovery can be basically completed in 6ms, and the non-adaptive scheme needs more than 9 ms. When K =2, the recovery efficiency of the optional type self-adaptive scheme is slightly higher than that of the adjacent type self-adaptive scheme; the advantage of the optional type adaptation scheme is more pronounced at K = 3.
In conclusion, the method is obviously superior to the traditional non-adaptive transmission scheme in the power failure recovery type scene; in addition, the invention also determines a transmission scheme with relatively best performance under the scene, namely an optional type self-adaptive transmission scheme with the retransmission times of 2, and the transmission scheme can finish the recovery access for 200 users within 6ms on the premise that the average resource utilization rate is 15RB/ms.
Example two
Let 200 users served by a certain base station have 120KHz subcarrier spacing. The total time T of the access process is distributed and accessed in a stable way in daily life a If the length is not less than 10ms, the base station allocates resources for the transmission process according to the historical value of the stable distribution. If the user-side distribution bursts to be Beta-type (taking α =3, β =4 as an example) at the beginning of the 6ms, the number of users estimated by the base station after the end of the first millisecond after the transition will be much smaller than the history value. At this time, the base station will perceive the change of the user side, and after the 7 th ms, according to the two estimated user numbers, calculate the parameter of Beta distribution to predict the number of new accessed users in each millisecond in the future. Meanwhile, in the whole process, the base station still keeps monitoring the condition of each millisecond, and the total load of the users in the last millisecond is estimated by the invention, and the counted number of the users with successful transmission is subtracted, so that the number of the users needing to be accessed again in the next millisecond due to failure is obtained. Adding the predicted new access and the number of the failed retransmission users to finally obtainBy the total number of possible active users in the next millisecond, resources are then allocated in accordance with the present invention.
In order to verify the superiority of the scheme, monte Carlo simulation is carried out on the scene in Matlab software, four schemes of optional type self-adaption K repetition, adjacent type self-adaption K repetition, optional type non-self-adaption K repetition and adjacent type non-self-adaption K repetition are compared, wherein the latter two schemes are used as a comparison group, and the simulation result is shown in figure 2.
K =8/K =7 in fig. 2 indicates that each user repeatedly tries 8/7 accesses within 1ms, respectively, and it is worth mentioning that K =8 is the best choice in this scenario, so K =7 functions as the control group here. In addition, the resource allocation amount of the non-adaptive scheme under stable distribution is the same as that of the adaptive scheme, and when the distribution on the user side is suddenly changed, the non-adaptive scheme still maintains the amount. In this embodiment, the resource utilization of the optional non-adaptive and neighboring non-adaptive are both 66RB/ms. Observing the trend in fig. 2, it can be found that the probability of transmission failure of the non-adaptive scheme is increased along with the rapid increase of the number of newly accessed users in Beta distribution in 9 th and 10 th ms, which will cause adverse effects on user services; while the failure probability of the adaptive scheme still remains at 1 × 10 -5 Left and right. And the resource utilization rate of the two self-adaptive schemes is 56RB/ms, so that the resource is greatly saved compared with the non-self-adaptive scheme.
In conclusion, the performance of the method is obviously superior to that of the traditional non-adaptive transmission scheme when the sudden change recovery type scene and the user side distribution change; in addition, the invention realizes higher transmission success probability with lower resource occupancy rate; the method is superior to the traditional scheme in both the transmission success probability and the resource occupancy rate.
Claims (1)
1. An authorization-free adaptive transmission method based on user estimation is characterized by comprising the following steps:
1) The base station observes the occupation condition of the RB after the end of each time slot, namely the number of the conflict-free, conflict-free and idle RBs, and carries out user load estimation through the occupation condition of a channel and a Markov process;
2) Performing adaptive resource allocation according to the user load estimated in the step 1), and then performing information transmission according to the result of the adaptive resource allocation to finish the authorization-free adaptive transmission based on the user estimation;
the specific operation of the step 1) is as follows: estimating user load in any time slot, estimating user load of optional K repetition and estimating user load of adjacent K repetition;
the specific operation of the user load estimation in any time slot is as follows:
setting T as the number of time slots divided according to subcarrier intervals in 1ms, W as the number of available RBs, A, B and C respectively representing the number of non-conflicted, conflicted and idle RBs observed in a certain time slot, N representing the user load parameter to be estimated, and setting
Representing the estimation result, calculating a maximum a posteriori estimate for the measurement result, then transforming to a maximum likelihood estimate based on the maximum a posteriori estimate, and then aggregating Markov processes to obtain
Wherein P represents a transition matrix of the markov process, and ((0, w) → (a, B, C)) represents transition probabilities between (0, w) states and (a, B, C) states in the P matrix after the N-step transition;
the specific process of the user load estimation of the optional K repetition is as follows:
calculating the load condition of the user side within 1ms
Wherein the content of the first and second substances,
calculating the transition probability from (A, B, C) to any state in the state space S as follows:
calculating a transition matrix P according to the formula (1-10), and substituting the calculated transition matrix P into the formula (1-6) to obtain P (0,0,WT→(A,B,C) Maximum size
The process of the user load estimation of the adjacent K repetition is as follows:
determining the number phi of users starting to transmit for the first time in each time slot;
calculating estimated user coincidences
Comprises the following steps:
step 2) during specific operation, firstly, judging a distribution scene at a user side as any one of a burst type scene, a stable type scene and a Beta distribution type scene, wherein the burst type scene belongs to a delay tolerant type scene, the latter two scenes belong to a delay sensitive type scene, then, carrying out self-adaptive resource allocation according to the user load estimated in the step 1), and then, carrying out information transmission according to the result of the self-adaptive resource allocation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110362521.1A CN113194499B (en) | 2021-04-02 | 2021-04-02 | Authorization-free self-adaptive transmission method based on user estimation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110362521.1A CN113194499B (en) | 2021-04-02 | 2021-04-02 | Authorization-free self-adaptive transmission method based on user estimation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113194499A CN113194499A (en) | 2021-07-30 |
| CN113194499B true CN113194499B (en) | 2022-12-09 |
Family
ID=76974862
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110362521.1A Active CN113194499B (en) | 2021-04-02 | 2021-04-02 | Authorization-free self-adaptive transmission method based on user estimation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113194499B (en) |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107852722A (en) * | 2015-07-27 | 2018-03-27 | 华为技术有限公司 | Exempt to authorize the link circuit self-adapting in multiple access system |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9301161B2 (en) * | 2012-07-27 | 2016-03-29 | Qualcomm Incorporated | Method and apparatus for available bandwidth estimation by a user equipment in idle and/or connected mode |
| CN109121166A (en) * | 2017-06-22 | 2019-01-01 | 维沃移动通信有限公司 | A kind of data transmission method, base station and user terminal |
| CN107453857B (en) * | 2017-08-14 | 2020-07-03 | 北京邮电大学 | Link self-adapting method for non-scheduling uplink transmission |
| CN110011775B (en) * | 2019-03-22 | 2022-06-07 | 安徽师范大学 | Method and system for jointly realizing active user detection and channel estimation |
| CN110071881B (en) * | 2019-04-26 | 2020-12-15 | 北京理工大学 | Active user detection and channel estimation method with adaptive overhead |
| CN111542002B (en) * | 2020-04-23 | 2021-09-03 | 浙江大学城市学院 | Data transmission method, base station and terminal for ultra-high-reliability low-delay communication |
-
2021
- 2021-04-02 CN CN202110362521.1A patent/CN113194499B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107852722A (en) * | 2015-07-27 | 2018-03-27 | 华为技术有限公司 | Exempt to authorize the link circuit self-adapting in multiple access system |
Non-Patent Citations (1)
| Title |
|---|
| Adaptive generalized space shift keying and downlink multi-user transmission in MIMO fading channels;Dongl Liang;《2014 IEEE International Conference on Communication Systems》;20150129;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113194499A (en) | 2021-07-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP1145501B1 (en) | System for allocating resources in a communication system | |
| Barcelo et al. | Learning-BEB: avoiding collisions in WLANs | |
| CN106982467B (en) | Access congestion control method based on dynamic allocation of PRACH resources | |
| CN102625466B (en) | Distributed media access method used in wireless local area network | |
| EP1626594A2 (en) | Method and apparatus for communicating in a wireless communication system | |
| EP2429232B1 (en) | Method and equipment for selecting terminal during congestion process | |
| CN108811001B (en) | CSMA competition and TDMA reservation-based discretization channel access method | |
| CN111245541B (en) | Channel multiple access method based on reinforcement learning | |
| Assarut et al. | Region division assignment of orthogonal variable-spreading-factor codes in W-CDMA | |
| CN111556528A (en) | Channel state judgment method based on frequency hopping time hopping pattern under self-organizing network | |
| CN100369524C (en) | CDMA system up-bag dispatching method | |
| CN113194499B (en) | Authorization-free self-adaptive transmission method based on user estimation | |
| CN113784444A (en) | TDMA time slot allocation method with priority on efficiency | |
| Petras et al. | MAC protocol for Wireless ATM: contention free versus contention based transmission of reservation requests | |
| CN115278919A (en) | Method for improving broadcasting performance of safety messages of Internet of vehicles | |
| CN108668379B (en) | Method for reserving lead code in low-delay and high-reliability scene of 5G system | |
| CN109905922B (en) | Large-scale terminal access method based on time slot Aloha and adaptive ACB mixing | |
| CN110958717A (en) | Contention-free random medium access control method and system for centralized wireless local area network | |
| Iyengar et al. | Analysis of contention-based multi-channel wireless MAC for point-to-multipoint networks | |
| Nagul | Channel scheduling by spectrum channel white space filling in cognitive radio networks | |
| Zhang et al. | Throughput and delay performance Analysis of Packet Aggregation Scheme for PRMA | |
| Bianchi et al. | Channel‐dependent load balancing in wireless packet networks | |
| CN116419410B (en) | Self-adaptive adjustment method for scheduling request period | |
| Foh et al. | Improving the Efficiency of CSMA using Reservations by Interruptions | |
| Buot | Channel allocation strategy for voice/data TDMA systems |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |