CN108811143A

CN108811143A - Uplink based on user activity is exempted to authorize the CTU distribution methods of SCMA

Info

Publication number: CN108811143A
Application number: CN201810481102.8A
Authority: CN
Inventors: 吴玮; 孙俊伟; 吴宣利; 薛彤; 郭珊; 张立强
Original assignee: Harbin Institute of Technology
Current assignee: Harbin Institute of Technology
Priority date: 2018-05-18
Filing date: 2018-05-18
Publication date: 2018-11-13
Anticipated expiration: 2038-05-18
Also published as: CN108811143B

Abstract

Uplink based on user activity is exempted to authorize the CTU distribution methods of SCMA, and the present invention relates to CTU distribution methods.CTU collision probabilities are big between user in being distributed the purpose of the present invention is to solve existing CTU, the high problem of computation complexity.Process is：One：It in exempting to authorize uplink SCMA systems, is calculated in a time slots of T ', the corresponding average CTU collision probabilities of L CTU of N_length user simplify average CTU collision probabilities, obtain NP-hard optimized-types；Two：By user n according to liveness W_nIt sorts from big to small, and gives different priority；The new set that N' expression user's collection N is generated after being arranged from big to small according to liveness, three：Obtain optimal solution set；Four：User and CTU relations of distribution matrixes are obtained according to N, N' and optimal solution set.The present invention CTU for SCMA distribute field.

Description

CTU (computer to utility) distribution method of uplink authorization-free SCMA (Single chip multiple Access) based on user activity

Technical Field

The invention relates to a CTU distribution method in an uplink authorization-free SCMA.

Background

The enormous demand of wireless communication for massive access, low delay and high throughput makes the fifth generation mobile communication (5G) a hot topic of current research. The 5G can satisfy a higher quality of service (QoS) requirement, and satisfy a scenario of Internet of things (IoT).

However, as a key technology of a Long Term Evolution (LTE) system, an orthogonal multiple access (OFDMA) technology divides a frequency band into a plurality of mutually orthogonal subcarriers and distributes the subcarriers to different users, respectively, to achieve multiple access. Limited by the orthogonal division mode, the user gain of the OFDMA does not exceed 1, and the requirement of mass access cannot be met. Meanwhile, the transmission of the uplink LTE follows a request and scheduling mode, which also causes unavoidable delay. Therefore, SCMA is proposed as a new type of non-orthogonal multiple access technique based on a codebook format. The competition-based uplink SCMA transmission mode realizes multiple access through different code books allocated to different users. Because of the sparse codebook, a new dimension is introduced, and the SCMA receiver can decode the overlapped signal information, which also brings user gain that the conventional orthogonal multiple access technique does not have.

In an uplink unlicensed SCMA transmission system, a user communicates data by assigning each user to a CTU, which is a combination of time, frequency, codebook and pilot, before transmitting a message. The receiver detects the users by detecting the pilot sequences in each CTU and performs channel estimation. As long as no pilot collision occurs in the base station, the receiver can decode the information sequence by means of a Message Passing Algorithm (MPA). A traditional CTU allocation strategy adopts a modular division mode, the CTU allocation is completed according to the serial number of the user, and the information characteristics of the user such as activity, position and the like are not considered, so that the CTU collision probability among users in the existing CTU allocation is high, and the calculation complexity is high.

SCMA (Sparse Code Multiple Access) is Sparse Code division Multiple Access; the CTU (ContentionTransmission Unit) is a contention transfer Unit.

Disclosure of Invention

The invention aims to solve the problems of high probability of CTU collision among users and high calculation complexity in the conventional CTU distribution, and provides a CTU distribution method of an uplink authorization-free SCMA (sparse code multiple access) based on user activity.

The CTU distribution method of the uplink authorization-free SCMA based on the user activity comprises the following specific processes:

the method comprises the following steps: in an authorization-free uplink SCMA system, calculating to obtain the average CTU collision probability corresponding to L CTUs of N _ length users in T' time slots, and simplifying the average CTU collision probability to obtain an NP-hard optimized expression;

step two: on the premise of no authorizationIn the SCMA system, users n are arranged according to the activity W_nSorting from big to small and giving different priorities; n — 1,2,. and N _ length;

the activity is high, and the priority is high;

n' represents the users in the user set N according to the activity W_nThe new set generated after the permutation from large to small, N' ═ nxtr,

wherein TR represents a mapping matrix from N to N'; the user set N comprises N _ length users; the set N 'includes N _ length' users;

step three: setting P_co(m) represents CTU collision probability of the first m users in the set N' obtained in step two, and Δ P (m) ═ P_co(m)-P_co(m-1)]Representing CTU collision increment when the mth user is considered to be accessed into the authorization-free uplink SCMA system, and carrying out iterative loop solution optimization on the delta P (m) to obtain the optimal solution g each time_mAccording to the optimal solution g at each time_mObtaining an optimal solution set G' ═ { G ═ G₁,...g_m,...,g_{N_length′}}；

Wherein m is 1, 2., N _ length'; p_co(m-1) representing the CTU collision probability of the first m-1 users in the set N' obtained in the step two;

step four: and obtaining a user and CTU distribution relation matrix G according to N, N 'and the optimal solution set G'.

The invention has the beneficial effects that:

the invention introduces the complexity of a user as a reference factor of CTU distribution on the basis of an original modular division algorithm, re-deduces a formula of average CTU collision probability based on the complexity of the user, and the optimal solution based on the collision formula is a nonlinear mixed integer problem and is also an NP-hard problem. Compared with a violent solving method, the method greatly reduces the complexity of calculation while reducing the CTU collision probability among users.

The performance advantages and features of the invention compared to the original modulo division based allocation are embodied in:

1. the activity characteristic of the user is considered, and the collision probability, namely the packet loss rate, is greatly reduced in the scene of intensive access of the user.

2. The denser the users are, the larger the gap of the activity degrees is, and the more remarkable the performance improvement is.

3. Compared with a brute force solution problem, the method can obtain an optimal solution with extremely low complexity.

As can be seen from fig. 3a, with the algorithm proposed in the present invention, the packet loss rate is reduced by 17%, 20%, and 22% when the number of users is 24,30, and 36, respectively. In fig. 3b, when the number of users is 24 and the number of CTUs is 8, 10, and 12, packet loss rates are reduced by 23%, 19%, and 17%. Therefore, the algorithm provided by the invention has good packet loss rate performance; meanwhile, the more dense scenes of users can be known, the better the packet loss rate performance of the proposed algorithm is, which is very practical for the scenes of the internet of things with 5G mass access in the future.

Fig. 4 shows a graph comparing the complexity of the proposed algorithm with brute force solution when the number of users varies from 24 to 36. It can be seen from fig. 4 that the computational complexity can be reduced by 16 orders of magnitude when the number of users is 30.

Drawings

FIG. 1 is a schematic diagram of uplink unlicensed SCMA transmission according to the present invention, where ACK is an acknowledgement command;

FIG. 2 is a diagram of the relationship between the time-frequency resources, the codebook and the pilot frequency in the user and CTU of the present invention, C₁Is a first codebook, C₂Is a second codebook, C_JIs the J-th codebook, P₁For the first pilot sequence, P_L+1For the L +1 th pilot sequence, P_L(J-1)+1Is the L (J-1) +1 pilot sequence, P_LFor the Lth pilot sequence, P_2LFor the 2L pilot sequence, P_LJIs the LJ-th pilot sequence, f is frequency, t is time, C is codebook, P is pilot sequence;

fig. 3a is a packet loss rate CDF graph corresponding to a CTU allocation algorithm based on user activity and a CTU allocation algorithm based on modulus under the condition that the number of CTUs is not changed, where CDF is cumulative probability density distribution; n is N _ length users;

fig. 3b is a graph of packet loss rate CDF corresponding to a CTU allocation algorithm based on user activity and a CTU allocation algorithm based on modular division under the condition that the number of users is not changed;

FIG. 4 is a comparison graph of the algorithm and the direct optimization solution complexity provided by the present invention.

Detailed Description

The first embodiment is as follows: the CTU allocation method of the uplink authorization-free SCMA based on the user activity of the embodiment specifically comprises the following processes:

therefore, in order to reduce the packet loss rate caused by CTU collision and further improve the sum rate of the whole uplink SCMA system, the invention introduces the activity of the user as an evaluation index to assist the base station to complete the distribution of CTU resources. The invention considers the following scenes of the internet of things: the activity of each user is different and their activity is in a quasi-static state, i.e. the probability that each user is allowed access for a period of time is subject to an allocation regarding its activity. Meanwhile, the access probability of the user in a period of time can also be used for estimating the activity of the user in the next period of time. In addition, the invention deduces the expression of the CTU collision probability, and finds that the collision probability has close relation with the activity of the user and the distribution strategy of the CTU. In this case, a CTU optimization allocation strategy based on user activity is studied to reduce the collision probability and packet loss rate of CTUs and prove that the problem of packet loss rate optimization is an NP-hard problem. By utilizing the CTU distribution algorithm based on the user activity, the NP-hard problem is converted into a low-complexity solvable problem, and the optimization of the system packet loss rate is completed.

The uplink authorization-free SCMA system model is as follows:

suppose that there are N users, L CTUs, J codebooks, and K subcarriers in an uplink SCMA system. And there is a correspondence: n ≧ J > K, the information received in the base station can be expressed as:

where y ═ y₁y₂...y_K]^TIs the information received in the sub-carriers; x is the number of_n＝[x_n1x_n2...x_nK]^TIs a UE_nThe information sent; f. of_n＝[f_n1f_n2...f_nK]^TThe corresponding relation between the user and the subcarrier can also be understood as codebook information; h is_n＝[h_n1h_n2...h_nK]^TIs a user UE_nThe channel state parameter of (a); n-CN (0,1) is Additive White Gaussian Noise (AWGN).

A transmission diagram of an uplink unlicensed SCMA is shown in fig. 1. Unlike the request-based scheduling mode of transmission, when a user wants to access, it transmits information in a follow-up and go (arrival and go) manner. I.e. neither a transmission request nor an access grant feedback from the base station to the user is required. For the uplink SCMA system, the codebook in the CTU corresponds to f in equation (1)_nThe pilot sequence in the CTU is used to assist in estimating the channel information h_n. Therefore, before the user sends information, the uplink unlicensed SCMA system allocates CTUs to the user. And finally, after blind detection, the base station feeds back an ACK to the user to confirm transmission.

Each CTU has its unique pilot sequence but may share one and the same SCMA codebook. The pilot sequence can help the receiver estimate the channel information. User information aliased to each other on one subcarrier can be demodulated at the cost of less than 0.5dB performance loss as long as they are assigned different CTUs. This loss of performance is quite acceptable for a real system. So in general in an uplink SCMA system, multiple users may share one CTU, and multiple CTUs may share one pilot sequence, but each CTU has its own pilot sequence. The relation between users and time-frequency resources, codebooks and pilots in CTUs is shown in fig. 2.

However, if users select the same CTU, they share the same pilot sequence to estimate their respective channel information. The receiver cannot decode the user information and this process is called pilot collision or CTU collision.

Assuming that there are N users in an uplink SCMA system, where N users are in an active state, the probability of CTU collision is:

the average arrival rate λ of users is defined as the number of users expected to need to transmit data per Transmission Time Interval (TTI), and can be expressed as,

wherein W_nIs the activity of the user UEn, the probability of n active users being subjected to a Poisson distribution

P_n＝e^-λ(λ)ⁿ/n！ (10)

The packet loss rate is defined as the probability of the first transmission failure of a user to the base station. Its upper bound can be expressed as:

the distribution relationship between the user and the CTU plays a decisive role in packet loss rate. An optimized allocation can effectively reduce the collision probability of the CTUs, thereby reducing the error rate. Let g be_nl1, if g, the assignment relationship between the nth user and the l CTU is expressed_nl1 means that the nth user is allocated the lth CTU, whereas g_nl0 means no pairing, and g_n＝[g_n1g_n2... g_nL]^TAn allocation relation vector, g, representing the nth user_l＝[g_1lg_2l... g_Nl]^TIs the assignment of the ith CTU. As can be seen from the above, one user can be allocated to only one CTU, but one CTU may correspond to a plurality of users. Therefore, the following constraints can be obtained:

on the premise of having a mapping relationship, the average CTU collision probability corresponding to L CTUs of N users in T time slots can be obtained:

wherein W_n(t) represents the probability of access of the nth user in the tth time slot. From equations (3-6), it can be known that the collision probability of the pilots is closely related to the activity of the users and the allocation relationship of the CTUs. The mapping relationship between users and CTUs in the original CTU allocation algorithm based on the modular division is as follows:

CTU_index＝UE_index mod CTU_num (14)

that is to say:

this allocation strategy of CTUs ignores the liveness characteristics of the user. So in the following, how to achieve the purpose of minimizing CTU collision probability by allocating CTU resources using the activity information of the user is discussed. Let N and L be used to represent the set of users and CTUs, respectively. The optimization formula can be expressed as:

min P_co

the optimized solution is an N × L matrix G representing the entire assignment correspondence of N users to L CTUs. The optimization problem is a nonlinear mixed integer problem, which is an NP-hard problem, that is, the problem has extremely high complexity, and it is impossible for the base station to directly solve and redistribute.

step two: in an authorization-free uplink SCMA system, a user n is set according to an activity W_nSorting from big to small and giving different priorities; n — 1,2,. and N _ length;

the activity is high, and the priority is high;

n' represents the users in the user set N according to the activity W_nArranged from large to smallThe new set generated later, N' ═ nxtr,

And step two, step three and step four are further optimization of the NP-hard optimization formula of the step one, and G obtained in the step four is the optimization result of the NP-hard optimization formula of the step one.

The second embodiment is as follows: the first difference between the present embodiment and the specific embodiment is: in the first step, in an authorization-free uplink SCMA system, calculating to obtain the average CTU collision probability corresponding to L CTUs of N _ length users in T' time slots, and simplifying the average CTU collision probability to obtain an NP-hard optimized formula; the specific process is as follows:

wherein,

g_nl{0,1} represents the nth user and the th userAssignment of l CTUs if g_nl1 means that the nth user is assigned to the lth CTU, whereas g_nl0 means that the nth user and the lth CTU are not paired;

g_n＝[g_n1g_n2... g_nL]^Tan allocation relation vector representing the nth user; t is transposition;

g_l＝[g_1lg_2l... g_{N_lengthl}]^Tan allocation relation vector representing the ith CTU;

W_n(t) the activity is used for representing the probability of accessing the nth user in the tth time slot; n ≠ N, g, N ═ 1,2_n′lThe {0,1} represents the allocation relationship between the nth' user and the lth CTU;

W_n'(t) the activity is used for representing the probability of accessing the nth user in the tth time slot; t is time;

the formula (1) is simplified

Wherein, W_nThe activity of the nth user; e (W)_n(t)) is to obtain W_n(t) expectation;

and the NP-hard optimized formula at this time is obtained:

min P_co

represents arbitrary; s.t. denotes such that.

Other steps and parameters are the same as those in the first embodiment.

The third concrete implementation mode: the present embodiment differs from the first or second embodiment in that: setting P in the third step_co(m) represents CTU collision probability of the first m users in the set N' obtained in step two, and Δ P (m) ═ P_co(m)-P_co(m-1)]Representing CTU collision increment when the mth user is considered to be accessed into the authorization-free uplink SCMA system, and carrying out iterative loop solution optimization on the delta P (m) to obtain the optimal solution g each time_mAccording to the optimal solution g at each time_mObtaining an optimal solution set G' ═ { G ═ G₁,...g_m,...,g_{N_length′}}; the specific process is as follows:

when the number m of the access users is less than or equal to L, the CTU allocation is not required to be completed through optimized calculation at the moment, and the method is as followsAllocating different CTUs to each user;

mod is the remainder;

when the number of access users m > L, the CTU collision increment occurring when the mth user is considered to access the authorization-free uplink SCMA system is expressed as:

wherein, g_mlThe assignment relationship between the mth access user and the lth CTU is expressed as {0,1}, if g_ml1, it means that the mth user is allocated to the lth CTU; otherwise if g_ml0 means that the mth access user and the lth CTU do not match each otherCarrying out pairing;

W_mrepresenting the activity of the mth user; p_co(l, m-1) represents the probability of collision of the l-th CTU in the (m-1) -th iteration;

by storing the collision value obtained from the last iteration as known information, the amount of optimization calculation can be greatly reduced at the cost of a small amount of space complexity.

For the mth iteration, Δ P (m) optimization is simplified to:

minΔP(m)

optimized result g_m＝[g_m1g_m2... g_mL]^TRepresenting the CTU assignment relationship vector for the mth user in the new set N';

judging whether m is equal to N _ length', if yes, executing step four; and if not, the step three is executed again until m is equal to N _ length'.

CTU collisions may be considered from another perspective, which may also be understood as one CTU being assigned to multiple accessed users. Therefore, for the ith CTU in the L set, compared with the result obtained in (5), there is no lower collision probability obtained by other user allocation modes on the premise that the collision rate of other CTUs is not increased, so that the result obtained by the dynamic planning can meet the condition of the Pareto optimal solution.

Other steps and parameters are the same as those in the first or second embodiment.

The fourth concrete implementation mode: the difference between this embodiment mode and one of the first to third embodiment modes is: in the fourth step, a user and CTU distribution relation matrix G is obtained according to N, N 'and the optimal solution set G'; the specific process is as follows:

other steps and parameters are the same as those in one of the first to third embodiments.

The summary of the proposed uplink SCMA CTU allocation strategy based on user activity is shown in algorithm 1: .

The following examples were used to demonstrate the beneficial effects of the present invention:

the first embodiment is as follows:

the simulation scenario of the figure is that 4 mutually orthogonal subcarriers, 6 codebooks and 12 CTUs are provided. The activity of the user is known to the base station in advance. The channel conditions are ideal channel estimates, i.e. the receiver of the base station can demodulate the user information as long as the users are assigned to different CTUs. The cumulative probability density distribution (CDF) of packet loss ratios corresponding to the CTU allocation algorithm based on user liveness and the CTU allocation algorithm based on modular division, which are proposed by us, is shown in fig. 3a and 3 b. Fig. 4 shows a comparison graph of algorithm complexity and brute force solution when the number of users varies from 24 to 36.

The following points can be seen from the simulation result:

1. in fig. 3a, in the case that the number of CTUs is 12, light-colored dotted lines represent a packet loss rate CDF map of a previous CTU allocation algorithm based on the modulo division, and dark-colored dotted lines represent a packet loss rate CDF map of a proposed CTU allocation algorithm based on the user activity; the solid line indicates the case where the number of users is 24, the broken line indicates the case where the number of users is 30, and the dotted line indicates the case where the number of users is 36. As can be seen from the images, with the proposed algorithm, the packet loss rates are reduced by 17%, 20%, and 22% for the number of users 24,30, and 36, respectively. This is because the proposed algorithm takes into account the liveness information of the users, providing more CTU resources for more active users, thereby reducing the collision probability. Similarly, in fig. 3b, when the number of users is 24 and the number of CTUs is 8, 10, and 12, the packet loss rate decreases by 23%, 19%, and 17%. Therefore, it can be seen that the proposed algorithm has very good packet loss rate performance; meanwhile, the more dense scenes of users can be known, the better the packet loss rate performance of the proposed algorithm is, which is very practical for the scenes of the internet of things with 5G mass access in the future.

2. Fig. 4 shows a graph comparing the complexity of the proposed algorithm with brute force solution when the number of users varies from 24 to 36. It can be seen from fig. 4 that the computational complexity can be reduced by 16 orders of magnitude when the number of users is 30. This is because the algorithm stores and applies the local optima obtained from each iteration to the next calculation by means of the idea of dynamic programming.

The present invention is capable of other embodiments and its several details are capable of modifications in various obvious respects, all without departing from the spirit and scope of the present invention.

Claims

1. The CTU distribution method of the uplink authorization-free SCMA based on the user activity is characterized in that: the method comprises the following specific processes:

the activity is high, and the priority is high;

2. The method for allocating CTUs of an uplink unlicensed SCMA based on user activity as claimed in claim 1, wherein: in the first step, in an authorization-free uplink SCMA system, calculating to obtain the average CTU collision probability corresponding to L CTUs of N _ length users in T' time slots, and simplifying the average CTU collision probability to obtain an NP-hard optimized formula; the specific process is as follows:

wherein,

g_nl{0,1} represents the assignment of the nth user to the lth CTU, if g_nl1 means that the nth user is assigned to the lth CTU, whereas g_nl0 means that the nth user and the lth CTU are not paired;

g_n＝[g_n1g_n2...g_nL]^Tan allocation relation vector representing the nth user; t is transposition;

g_l＝[g_1lg_2l...g_{N_lengthl}]^Tan allocation relation vector representing the ith CTU;

the formula (1) is simplified

and the NP-hard optimized formula at this time is obtained:

min P_co

3. the method for CTU distribution of an uplink unlicensed SCMA based on user activity as claimed in claim 2, wherein: setting P in the third step_co(m) represents CTU collision probability of the first m users in the set N' obtained in step two, and Δ P (m) ═ P_co(m)-P_co(m-1)]Representing CTU collision increment occurring when the mth user is considered to be accessed into the authorization-free uplink SCMA system, and carrying out iterative loop solution optimization on the delta P (m) to obtainOptimal solution g at a time_mAccording to the optimal solution g at each time_mObtaining an optimal solution set G' ═ { G ═ G₁,...g_m,...,g_{N_length′}}; the specific process is as follows:

m is 0, when the number of access users m is less than or equal to LAllocating different CTUs to each user;

wherein, g_mlThe assignment relationship between the mth access user and the lth CTU is expressed as {0,1}, if g_ml1, it means that the mth user is allocated to the lth CTU; otherwise if g_mlIf the number is 0, the mth access user and the lth CTU are not paired;

for the mth iteration, Δ P (m) optimization is simplified to:

minΔP(m)

optimized result g_m＝[g_m1g_m2...g_mL]^TRepresenting the CTU assignment relationship vector for the mth user in the new set N';

4. The method for CTU distribution of an uplink unlicensed SCMA based on user activity as claimed in claim 3, wherein: in the fourth step, a user and CTU distribution relation matrix G is obtained according to N, N 'and the optimal solution set G'; the specific process is as follows: