CN112367523A

CN112367523A - Resource management method in SVC multicast based on NOMA in heterogeneous wireless network

Info

Publication number: CN112367523A
Application number: CN202011195061.XA
Authority: CN
Inventors: 沈航; 吉晓祥; 白光伟
Original assignee: Nanjing Tech University
Current assignee: Nanjing Tech University
Priority date: 2020-10-31
Filing date: 2020-10-31
Publication date: 2021-02-12
Anticipated expiration: 2040-10-31
Also published as: CN112367523B

Abstract

A resource management method in SVC multicast based on NOMA in heterogeneous wireless network, the resource management is that the joint resource management includes wireless spectrum allocation and transmission power division, at first, the joint resource management problem is decomposed into sub-problems, then the sub-problems are solved separately, the steps include: 1) the joint resource management problem is converted into a mixed integer nonlinear programming problem, and the mixed integer nonlinear programming problem P0 is decoupled into 2 subproblems: the intra-multicast group transmit power partitioning problem P1; the problem of splitting frequency spectrum resources among base stations and distributing multi-group frequency spectrum resources in the base stations is P2; 2) solving a problem P1, solving all possible power distribution schemes, wherein the group with the highest overall video quality is the optimal solution; 3) the problem P2 is solved by modeling the spectrum resource partitioning between base stations and the resource allocation between multicast groups as a knapsack problem.

Description

Resource management method in SVC multicast based on NOMA in heterogeneous wireless network

Technical Field

The invention belongs to the technical field of computers, and particularly relates to a scalable video multicast method based on non-orthogonal multiple access in a heterogeneous wireless network.

Background

Video services have become the mainstream service of wireless communication, and the increasing video traffic aggravates the shortage of wireless spectrum resources. Multicast technology is one of effective technologies for improving spectrum utilization rate^[1]. Many User Equipments (UEs) will typically request the same video content and therefore may use the same frequency resources^[2]. However, since all multicast ues need to decode the received video correctly, the ue with the worst channel condition becomes the bottleneck restricting the multicast rate increase^[3]. Applying Scalable Video Coding (SVC) to wireless Video multicast can alleviate performance bottlenecks caused by network edge users.

SVC encodes video into one base video layer and several enhancement video layers. The user equipment can flexibly receive the decoded video according to the channel condition. User equipment with poor channel conditions decodes only the base video layer, and user equipment with good channel conditions decodes the base video layer and the enhancement video layer. However, existing SVC video multicasting is mostly based on Orthogonal Multiple Access (OMA)^[4][5]. Document [4 ]]The problem of multicast resource allocation in the WiMAX network is modeled, and a greedy algorithm is designed to maximize the utility of the UEs throughput. For OMA-based multi-carrier systems, document [5 ]]An optimal two-step dynamic programming algorithm is provided for solving optimal resource allocation. Since different video layers use orthogonal frequency resources for transmission, the spectrum utilization is limited to a certain extent. In recent years, Non-Orthogonal Multiple Access (NOMA) as one of the core technology alternatives of the fifth generation mobile communication system (5G) can significantly improve the network capacity and the spectrum utilization rate^[6]。

NOMA achieves multiple access in the power domain, and at the transmitting end, signals are transmitted through Superposition Coding (SC). Actively introducing Interference information, and decoding signals required by users by using Successive Interference Cancellation (SIC) technology at a receiving end^[7]. Decoding of SVC continuous video layer with NOMAThe combination of the SIC technology can allow different video layers to transmit on the same spectrum resource with different powers, further improving the spectrum utilization.

Document [8]The power allocation problem of the downlink NOMA system is researched, and the throughput is maximized under the condition of total power constraint and the minimum service quality of a user. To maximize the weighting of a multi-carrier full-duplex NOMA system and the system throughput [9 ]]An optimal resource allocation algorithm is designed by utilizing a monotone optimization method. Document [10]]A NOMA-based SVC video multicast scheme for cellular networks is proposed. Existing NOMA-based SVC video multicast mostly uses a single layer network architecture. The multilayer heterogeneous network architecture has great potential in improving SVC video multicast performance^[11]. In a heterogeneous network, spectrum slicing (spectrum slicing), on-chip spectrum allocation and multicast group affiliation all affect the overall video quality of a multicast group. Therefore, how to optimize spectrum and power resource management in the NOMA-based SVC video multicast in the heterogeneous network and improve the service quality of multicast users is worth further discussing.

Disclosure of Invention

The scheme provides a scalable video SVC multicast method based on non-orthogonal multiple access NOMA in a heterogeneous wireless network, and the method converts the problem of joint resource management into the problem of mixed integer nonlinear programming.

For ease of solution, the mixed integer nonlinear programming problem is decoupled into 2 sub-problems:

sub-problem 1 — transmit power partitioning problem within multicast group;

sub-problem 2-spectrum resource splitting between base stations and multi-group spectrum resource allocation problem in base station.

Firstly, solving the subproblem 1, solving all possible power distribution schemes, wherein the group with the highest overall video quality is the optimal solution.

And then solving a subproblem 2, modeling the problem of resource allocation between the spectrum resource cutting among the base stations and the multicast groups into a knapsack problem and solving the knapsack problem.

Simulation results show that compared with the existing reference scheme, the scheme provided by the invention can improve the video receiving quality of users to a greater extent and improve the resource utilization rate. For different numbers of subchannels and multicast groups, the system utility increases by an average of 5% -10%, and the average PSNR achieved by the users increases by 1-2 dB.

Drawings

Fig. 1 — network architecture in this embodiment (consider a two-layer heterogeneous wireless network consisting of 1 macro base station and several small base stations);

fig. 2-scalable video multicast schematic;

FIG. 3 is a schematic diagram of the impact of the number of NOMA layers on the proposed scheme performance (system utility and) user average PSNR;

FIG. 4 is a diagram illustrating the effect of the number of subchannels on the utility of a system;

FIGS. 5(a) and 5(b) are schematic diagrams of the effect of multicast group number; wherein:

FIG. 5(a) is a schematic diagram of the impact of multicast group number on system utility;

fig. 5(b) is a graph illustrating the effect of multicast group number on the user average PSNR (peak signal-to-noise ratio).

Detailed Description

The invention will now be illustrated with reference to specific examples, in which:

the first section describes a system model.

The second part describes the SVC video multicast mechanism based on NOMA in the heterogeneous wireless network scene proposed by the scheme and raises the problem.

The third part decouples the problem into two subproblems and designs a loop algorithm and a dynamic programming algorithm for solving.

The fourth section simulates and evaluates the proposed solution.

Finally, the invention is summarized.

1 System model

1.1 resource management framework

This example is shown in fig. 1, and considers a two-layer heterogeneous wireless network composed of 1 macro base station and several small base stations.

The macro base station and the small base station respectively serve multicast groups within the coverage area. Controller (A)controller) grasps all user information and manages resources. The controller divides the total available spectrum resource B into two parts B_mAnd B_s. Wherein B is_mIs allocated for use by the macro base station; b is_sIs allocated to each small cell multiplex.

Representing a set of users of a kth group as a multicast group

The K sets of multicast groups may be represented as

Multicast groups are divided into two types: if a_j，k1, the kth multicast group is served by the jth small base station; if it is

The kth multicast group is served by the macro base station. a is_j，kIs a 0-1 variable used to indicate the slave mode.

1.2 SVC video streaming

Each multicast group requests one SVC video stream. An SVC video can be coded as a base video layer (BL) and several enhancement video layers (ELs) that can enhance the video quality^[12]. For the k-th SVC video, the total number of video layers is marked as L_kThe video layers are arranged in order from the base video layer to the highest enhancement video layer. I.e., the base video layer is referred to as the first video layer, the first enhancement video layer is referred to as the second video layer, and so on. The bit rate of the l video layer in the k SVC video is denoted as λ_l，k。

The receiving end can reconstruct the video by using part of the video layer. SVC video decoding requires the presence of successive video layers, i.e., an upper video layer can be decoded correctly if and only if all lower video layers have been successfully decoded. Ith User Equipment (UE) for kth multicast group_iThe total number of video layers to receive decoding may be expressed as l_i，k. This means that the UE is_iReceiving l_i，kA continuous video layer starting from the base layer, then the UE_iThe received SVC video rate may beIs shown as

UE_iReceived video rate is R_i，kUsing u_i，k(R_i，k) Representing a UE_iThe obtained effect^[5]. Using logarithmic function as utility function, i.e. u_i，k(R_i，k)＝log₂(R_i，k)，u_i，k(R_i，k) With video rate R_i，kThe increase in (c) shows a rising trend.

1.3 NOMA layer

The K multicast groups have B orthogonal sub-channels which can be used, the bandwidth of each sub-channel is W, and the transmission power of the macro base station is P_mThe transmission power of the small base station is P_jNoise power density of N₀. Multiple access of NOMA is achieved in the power domain, and the video transport layer in a NOMA superposition coding scheme is called a NOMA layer.

Assume that the maximum number of layers per video transmission is N. For the k multicast group, the video signal transmitted by the nth NOMA layer on the macro base station (small base station j) is x_m，n，k(x_j，n，k) Then the channel gain is not less than h_m，n，k(h_j，n，k) The device may decode the signal. The NOMA layer is ordered according to the ascending order of the channel gains of the involved user equipments, i.e. for the macro base station, h_m，1，k＜h_m，2，k＜…＜h_m，N，k(for small base stations j, h)_j，1，k＜h_j，2，k＜…＜h_j，N，k)。

According to the principle of superposition coding of NOMA transmitting terminals, the power of an nth NOMA layer allocated to a kth video by a macro base station (small base station j) is P_mα_m，n，k(P_jα_j，n，k). Wherein alpha is_m，n，k(α_j，n，k) Denotes a power allocation ratio of the macro base station m (small base station j).

Suppose a UE_iAssociated with the nth NOMA layer, then the macro base station and the small base station j are in the UE_iTo decode the signal x_m，t，k(x_j，t，k) (t < n) the achievable video rates are respectively

And

wherein the corresponding signal to interference plus noise ratios are:

due to h_m，t，k＜h_m，n，k(h_j，t，k＜h_j，n，k)，(t＜n)，

If the user equipment associated with the m-th NOMA layer can decode the signal x_m，n，k(x_j，n，k) Then the user equipment associated with the nth NOMA layer must be able to decode signal x_m，n，k(x_j，n，k)。

If the UE is_iCan decode the signal x_m，n，k(x_j，n，k) When decoding higher layer video, the UE_iThe signal x can be deleted_m，n，k(x_j，n，k) The interference of (2). Therefore, the achievable transmission rates of the nth NOMA layer of the macro base station and the small base station are respectively

And

wherein

When N < N, the signals of N +1 to the highest NOMA layer are regarded as interference; the maximum number of layers per video transmission is N,

when N ═ N, there is no interference from other NOMA layers.

For small base station sets

It is shown that the denominator parts of equations (2d) and (3d) take into account the interference between the small base stations due to the spectrum reuse. The multicast group can only be subordinate to one base station, and the achievable video rate of the user equipment of the nth NOMA layer is expressed as

2 federated resource management problem

Joint resource management in non-orthogonal multiple access based SVC video multicast under heterogeneous networks involves: wireless spectrum allocation and transmit power partitioning.

2.1 Wireless Spectrum Allocation

The total spectrum resources available to the system are B subchannels. The controller divides B into two parts B_mAnd B_sFor macro and small base stations, respectively, i.e.

B_m+B_s≤B (5a)

After the spectrum resources are split, the available spectrum resources for each base station are allocated to the associated multicast group. The base station maximizes the utility of all users by allocating appropriate spectrum resources for each multicast group on the basis of meeting the quality of service requirements of all user devices.

The macro base station will B_mThe subchannels are assigned to the associated multicast group. All multicast group allocated subchannels b associated with macro base stations_m,kThe sum of the numbers cannot exceed the number of available subchannels of the macro base station, i.e.

The number of subchannels assigned to the associated multicast group per small base station is B_s. All multicast group assigned sub-channels b associated with small cell_j,kThe sum of the numbers cannot exceed the number of available subchannels of the small base station, i.e.

Multicast groups are divided into two types:

the kth multicast group is served by the macro base station; a is_j，kThe kth multicast group is served by the jth small base station, 1.

2.2 transmit Power splitting

The NOMA-based SVC video multicast scheme is essentially the SVC video layer superposition coding and base station transmission power partitioning problem, and comprises the following steps:

1. sub-grouping: all users in the multicast group are arranged according to the ascending order of the channel gain, and can be divided into N subgroups at most. The indices of the subgroups range from 1 to N. Users of subgroup 1 can only receive the decoded base video layer. The nth subgroup is matched to the nth NOMA layer. That is, the worst channel gain among users of the nth subgroup is the channel gain of the nth NOMA layer. (n, k) denotes the nth subgroup or NOMA layer in the kth multicast group, and defines(n_iK) is comprised of UE_iA subgroup of (a).

2. Power distribution: for the kth multicast group requested video, each NOMA layer is allocated a certain power for transmitting information. For NOMA layer (n, k), if the power allocated to it is proportional to alpha_n，kThe achievable transmission rate at this power is r according to equation (4)_n，k. The corresponding power allocation constraint can be expressed as:

equation (6) ensures that the sum of the transmission power of all NOMA layers for each multicast group cannot exceed the maximum transmission power of the base station.

3. Information transmission: after the sub-grouping and power allocation, the base station multicasts the different video layers through the respective NOMA layers. If epsilon_l，n，kThe ith video layer is transmitted in video transmission layer (n, k) 1, else epsilon _l，n，k0. Accordingly, information transfer constraints may be expressed as

And

where (7a) defines that any video layer is transmitted a maximum of 1 time among all NOMA layers. 7(b) limiting the bit rate of information transmitted in each NOMA layer to be not greater than the achievable transmission rate of the user of the NOMA layer, which ensures that the user can receive the video layer in real time to successfully reconstruct the video. Epsilon_l，n，kIs a 0-1 indicator variable, if ε_l，n，kThe ith video layer is transmitted in video transmission layer (n, k) 1, else epsilon_l，n，k＝0。

And the receiving end decodes the received signal through the SIC receiver. UE of k-th multicast group_iCan decode not higher than (n)_iK) a NOMA layer. UE of k-th multicast group_iThe number of available video layers that can be received is indicated as

Fig. 2 shows the SVC video overlay coding and decoding process. When the base station transmits signals, an SVC video layer l_1，kBy x_1，kTransport, SVC video layer l_2，kAnd l_3，kBy x_2，kTransport, SVC video layer l_4，kBy x_3，kTransmission, signal { x_1，k，x_2，k，x_3，kThe allocated power ratios are { alpha } respectively_1，k，α_2，k，α_3，k}. These signals are transmitted simultaneously by superposition coding, and SIC decoding is used at the receiving end. Subgroup S_1，kThe user channel in (1) has poor gain and can only decode the SVC video layer l_1，kSubgroup S with better user channel gain_2，kA user in (1) can decode the SVC video layer l_1，k，l_2，kAnd l_3，kAnd the subgroup S with the best channel gain_3，kA user in (1) can decode the SVC video layer l_1，k，l_2，k，l_3，kAnd l_4，k。

2.3 problem modeling

SVC video multicast is carried out based on NOMA under a heterogeneous network, the problem of joint resource management is solved, and the overall video quality is maximized on the basis of meeting the service quality requirements of all users. Under the constraint of the total available sub-channel number of the system and the power of the base station, the joint resource management problem can be expressed as

UE_iReceived SVC video rate

3 Algorithm design

For convenience of processing, the optimization problem P0 is decoupled into two sub-problems P1 and P2, and an algorithm is designed to solve.

3.1 problem resolution

The area covered by the small base station is a hot spot area gathered by the user equipment. When a multicast group requesting a certain video is all covered by a certain small cell, then the multicast group is subordinate to this small cell. There is an opportunity for spectrum reuse when there are more than 2 small base stations with multicast group affiliations. By spectrum splitting, the multicast group subordinate to the small base station can multiplex the spectrum, and the spectrum efficiency is further improved. Assuming that the coverage areas of the small base stations are all Rs, the dependency relationship is

d_i,jIndicating a request multicast group

Is from the small base station j.

If the spectrum resources of a given multicast group are b_kThe optimal power allocation for the multicast group may be found, thereby obtaining the maximum system utility for the multicast group. On the basis, an algorithm is further designed to solve B_m、B_sAnd all b_kThe value of (c). The P0 is divided into two sub-problems, P1 and P2, to solve.

Given b_kThen the intra-group scalable video power allocation problem P1 can be described as

If the expression (10) as the dependency relationship is substituted into the expression (11), the macro base station has

For small base station j there is

Based on P1, the problem of spectrum partitioning among base stations and allocation of multiple groups of spectrum resources in base stations P2 can be described as

3.2 Intra-group Power Allocation Algorithm

For P1, spectrum resource b of the kth multicast group is given_kBase station to a group of users

The multicast SVC video stream is transmitted as a video stream,

are arranged in ascending order of user channel gain, I denotes the number of users of this multicast group. P1 requires solving two sets of variables: 1) SVC video layer transmitted by each NOMA layer; 2) power distribution coefficient of each NOMA layer.

The first is to determine the SVC video layers transmitted within each NOMA layer, assuming that the SVC video is encoded into 4 layers and transmitted over 3 NOMA layers. The base video layer is denoted as BL and the enhancement video layers are denoted as EL1, EL2 and EL3 in that order, all 4 schemes are shown in table 1

Table 1N ═ 3, L_kInformation transmission when 4

Serial number	1	2	3	4
					x_1，k	BL	BL	BL+EL1	BL
x_2，k	EL1	EL1+EL2	EL2	EL1
					x_3，k	EL2+EL3	EL3	EL3	EL2

The power distribution coefficient of each NOMA layer is then determined. Given an information transmission scheme, the information transmission scheme will be

In ascending order of channel gain. And solving the power distribution coefficient according to the sequence of the first high NOMA layer and the second low NOMA layer. First, the calculation of alpha is performed circularly from 1 to I +1_3，kFrom 1 to I represent

When the loop is to I +1, the I users in (1) indicate that none of the users can receive the NOMA layer (3, k), i.e., alpha _3，k0; similarly, alpha can be calculated circularly_2，k. User-based minimum serviceQuality requirement, all user equipments can receive the decoded NOMA layer (1, k), according to

Alpha can be obtained by corresponding to the user with the worst channel gain_1，k. For each set of power distribution coefficients { alpha ] that complies with the constraint (6)_1，k，α_2，k，α_3，kCompute the sum utility of the multicast groups.

The maximum utility, that is, the maximum utility of the multicast group under a given spectrum resource, can be obtained by calculating the utility of all information transmission schemes. Compared with the method of the document [10], the complexity of the loop solution is high, but the base station power is not used completely, so the scheme has higher energy efficiency.

3.3 Spectrum management Algorithm

P2 is solved in two steps: 1) given B_mAnd B_sCalculating the sub-channel distribution between the corresponding multicast groups when the utility of the macro base station and each small base station is maximum; 2) and solving the spectrum segmentation among the base stations so as to maximize the system utility.

First, assuming that a total of b subchannels are available for the first k multicast groups, the maximum sum of the utility of the first k multicast groups is represented using the function S (k, b). According to the idea of dynamic programming, S (k, b) can be expressed as a recursive formula, namely:

U_k(b) it can be obtained by equation (11), and S (k, b) is solved by a polynomial selection knapsack algorithm. The K multicast groups are treated as K types of items and are packed in a backpack with a capacity of B. Each class having B items, class kth B_kProfit U for each project_k(b_k) And weight b_k. Then the essence of S (k, b) is to select a portion from each category that does not exceed the total capacity of the backpack, maximizing the sum of profits.

Algorithm 1 Spectrum Allocation Algorithm among multicast groups within base stations

Algorithm 1 uses a dynamic programming method to solve the optimal solution of the knapsack problem, namely the optimal solution of S (k, b), when the weights of all items are nonnegative integers^[13]. First, initialization S (0, b) ═ 0, and k ═ 0 indicates that no multicast group exists. Then, all the U's are circularly calculated_k(b) The value of (c). Next, S (k, b) is recursively calculated starting from k equal to 1, and in each iteration of the recursion, S (k, b) is calculated using a recursion equation (15), where S (k-1, b-b') has been calculated in the previous iteration of the recursion, and U is calculated_k(b') has been calculated prior to the recursive process. After KB iterations, all values of S (k, b) can be obtained. Finally, return S (K, B) as optimal system utility and use B_kAs the number of subchannels allocated to the kth multicast group.

Given the spectrum resources of the base station, the spectrum allocation in the base station can be obtained through the algorithm 1, and the spectrum segmentation between the base stations is calculated next. The macro base station comprises

A multicast group, a small base station j comprising

A multicast group. Respectively calculating the maximum and utility of a macro base station and a small base station j by an algorithm 1 to obtain the maximum and utility of the system

Here, B_mAnd B_sThe value of (d) can be found by algorithm 2:

and 2, algorithm: spectrum segmentation algorithm for macro base station and small base station

Algorithm 2 first divides the available spectrum resources into two parts B_tAnd B_iAnd respectively allocating the macro base station and the small base station for use. And respectively obtaining the total utility of the macro base station and the small base station through the algorithm 1 to obtain the total utility of the system. Finally, B capable of maximizing total utility of the system is found through searching_tAnd B_iThe value of (c) is the spectrum segmentation of the macro base station and the small base station.

4 Experimental design and results analysis

The validity of the proposed scheme is verified by means of a simulation method. Considering 1 heterogeneous wireless network with 4 small base stations in the coverage area of the macro base station, the macro base station is located at the origin of a coordinate system, and the 4 small base stations are respectively located at the origin of the coordinate system

Here, the distance between the macro base station and each small base station is 500 m. The downlink transmitting power of the macro base station is 40dBm, the communication coverage radius is 800m, and each small base station has the same transmitting power of 33dBm and the coverage radius of 300m^[14]. For the channel transmission model, L is used respectively_m(z)＝-30-35log₁₀(z) and L_s(z)＝-30-35log₁₀(z) to describe the downlink channel gain of the macro base station and each small base station^[15]Where z is the distance between the base station and the user equipment. Other parameters are shown in table 2. Use document [16 ]]Is coded as 10 standard video test sequences of an SVC video stream and uses the average bit rate and PSNR values as evaluation indexes.

TABLE 2 simulation parameters

Firstly, evaluating the SVC video multicast scheme based on NOMA under the heterogeneous wireless network proposed by the scheme. In this scheme, there are 10 multicast groups requesting different SVC video streams, where each of the 4 small base stations covers 1 multicast group, and each multicast group has 15 user equipments. Figure 3 shows the effect of the number of NOMA layers on the performance of the proposed scheme.

As can be seen from fig. 3, the system utility and user average PSNR increase with increasing number of NOMA layers. While there was no significant improvement in performance when the number of NOMA layers was increased from 3 to 4. This is because the power of the base station is limited and as the number of NOMA layers increases, the power allocated to each layer decreases. In order to reduce the computational complexity, the number of the following simulation NOMA layers is set to be 3, and all frequency resource allocation methods adopt a knapsack algorithm to solve.

To objectively assess performance, two benchmark scenarios were implemented for comparison, including:

OMA-based scalable video multicast (OMA)^[6]In orthogonal multiple access, different video layers are transmitted at different transmission rates, which is acceptable for users with different channel gains. In this scheme, the SVC video layer is transmitted over orthogonal channel resources.

Fixed power scalable video multicast (F-NOMA)^[17]The power ratio of each NOMA layer is predetermined. Extending fixed power allocation into scalable video multicast is another benchmark scheme.

4.1 Effect of the number of System subchannels

The number of the given multicast groups is 10, the number of the user equipment in each multicast group is 15, and the number of the system subchannels is 10-30. Figure 4 shows the trend of the system utility as the number of subchannels increases.

As shown in fig. 4, as the number of subchannels increases, both the 3-scheme system utility and the average PSNR value increase. When the number of subchannels is sufficiently large, the system utility and the average PSNR value tend to be constant. This is because more subchannels means more system capacity and more video information can be transmitted. However, the system utility and average PSNR value also do not grow infinitely with increasing number of subchannels, since eventually all SVC video layers can be received and decoded by all users when the number of subchannels is large enough. Furthermore, the proposed scheme has the maximum system utility and average PSNR value for all subchannel numbers, followed by SOM and F-NOMA. The reason is because the power allocation scheme of the proposed scheme performs optimally when the multicast group frequency resources have been determined.

4.2 Effect of multicast group number

The number of given subchannels is 10, the number of user equipments in each multicast group is 15, and the number of multicast groups is 4-10. Fig. 5(a) and 5(b) show the trend of the system utility and the user average PSNR as the number of multicast groups increases, respectively.

As can be seen from fig. 5(a), the system utility increases in an upward trend as the number of multicast groups increases from 4 to 10. The system utility will not increase indefinitely due to the limited number of sub-channels available. In fig. 5(b), when the number of user groups increases from 4 to 6, the subchannels allocated to each group in the proposed scheme and the OMA scheme do not decrease due to the spectral reuse among the cell base stations, and thus the user average PSNR shows a tendency to increase. When the number of multicast groups exceeds 6, the number of subchannels allocated to each group is gradually reduced as the number of multicast groups increases, and thus the average PSNR per user equipment is also reduced. The performance of the proposed scheme is not much different from OMA when the number of multicast groups is small. As the number of multicast groups increases, the proposed scheme maintains the highest bit for both system utility and user average PSNR.

Concluding sentence

The scheme provides an SVC video multicast mechanism based on NOMA in a heterogeneous wireless network. In a heterogeneous network, NOMA and SVC video multicast are combined, and the problems of multicast resource management and power distribution in a group are researched, so that the overall video quality of a multicast group is improved to the maximum extent. Simulation results show that the proposed NOMA-based SVC video multicast scheme is superior to the two other schemes in system utility and user equipment average PSNR.

The references in the present invention are:

[1]ARANITI G,CONDOLUCI M,SCOPELLITI P,et al.Multicasting over emerging 5G networks:Challenges and perspectives[J].IEEE Network,2017,31(2):80-89.

[2]CHEN J,CHIANG M,ERMAN J,et al.Fair and optimal resource allocation for LTE multicast(eMBMS):Group partitioning and dynamics[C]//2015IEEE Conference on Computer Communications(INFOCOM).IEEE,2015:1266-1274.

[3]AFOLABI R O,DADLANI A,KIM K.Multicast scheduling and resource allocation algorithms for OFDMA-based systems:A survey[J].IEEE Communications Surveys&Tutorials,2012,15(1):240-254.

[4]DEB S,JAISWAL S,NAGARAJ K.Real-time video multicast in WiMAX networks[C]//IEEE INFOCOM 2008-The 27th Conference on Computer Communications.IEEE,2008:1579-1587.

[5]LI P,ZHANG H,ZHAO B,et al.Scalable video multicast with adaptive modulation and coding in broadband wireless data systems[J].IEEE/ACM Transactions on Networking,2011,20(1):57-68.

[6]DING Z,LIU Y,CHOI J,et al.Application of non-orthogonal multiple access in LTE and 5G networks[J].IEEE Communications Magazine,2017,55(2):185-191.

[7]DAI L,WANG B,YUAN Y,et al.Non-orthogonal multiple access for 5G:solutions,challenges,opportunities,and future research trends[J].IEEE Communications Magazine,2015,53(9):74-81.

[8]WANG C L,CHEN J Y,CHEN Y J.Power allocation for a downlink non-orthogonal multiple access system[J].IEEE wireless communications letters,2016,5(5):532-535.

[9]SUN Y,NG D W K,DING Z,et al.Optimal joint power and subcarrier allocation for full-duplex multicarrier non-orthogonal multiple access systems[J].IEEE Transactions on Communications,2017,65(3):1077-1091.

[10]ZHU H,CAO Y,JIANG T,et al.Scalable NOMA multicast for SVC streams in cellular networks[J].IEEE Transactions on Communications,2018,66(12):6339-6352.

[11]WANG C X,HAIDER F,GAO X,et al.Cellular architecture and key technologies for 5G wireless communication networks[J].IEEE communications magazine,2014,52(2):122-130.

[12]SCHWARZ H,MARPE D,WIEGAND T.Overview of the scalable video coding extension of the H.264/AVC standard[J].IEEE Transactions on circuits and systems for video technology,2007,17(9):1103-1120.

[13]HANS KELLERER,ULRICH PFERSCHY,DAVID PISINGER.Knapsack Problems[M].Springer Berlin Heidelberg,2004.

[14]QIAN L P,WU Y,ZHOU H,et al.Non-orthogonal multiple access vehicular small cell networks:Architecture and solution[J].IEEE Network,2017,31(4):15-21.

[15]YE Q,ZHUANG W,ZHANG S,et al.Dynamic radio resource slicing for a two-tier heterogeneous wireless network[J].IEEE Transactions on Vehicular Technology,2018,67(10):9896-9910.

[16]SHARANGI S,KRISHNAMURTI R,HEFEEDA M.Energy-efficient multicasting of scalable video streams over WiMAX networks[J].IEEE transactions on multimedia,2010,13(1):102-115.

[17]BENJEBBOVU A,LI A,SAITO Y,et al.System-level performance of downlink NOMA for future LTE enhancements[C]//2013 IEEE Globecom Workshops(GC Wkshops).IEEE,2013:66-70.

Claims

1. a resource management method in SVC multicast based on NOMA in heterogeneous wireless network, the resource management is joint resource management, it includes wireless frequency spectrum allocation and transmission power division, its characteristic is to decompose the joint resource management problem into subproblems at first, then solve the subproblem separately, the step includes:

1) the joint resource management problem is converted into a mixed integer nonlinear programming problem, and the mixed integer nonlinear programming problem P0 is decoupled into 2 subproblems: the intra-multicast group transmit power partitioning problem P1; the problem of splitting frequency spectrum resources among base stations and distributing multi-group frequency spectrum resources in the base stations is P2;

2) solving a problem P1, solving all possible power distribution schemes, wherein the group with the highest overall video quality is the optimal solution;

3) the problem P2 is solved by modeling the spectrum resource partitioning between base stations and the resource allocation between multicast groups as a knapsack problem.

2. The method of claim 1, wherein the resource management method is performed in accordance with a protocol of a mobile communication system

The wireless spectrum allocation method comprises the following steps:

dividing B sub-channels of total spectrum resource into two parts B_mAnd B_sFor macro and small base stations, respectively, i.e.

B_m+B_s≤B (5a)

After the spectrum resources are segmented, the available spectrum resources of each base station are allocated to the associated multicast group; the base station distributes proper spectrum resources for each multicast group, and maximizes the utility of all users on the basis of meeting the service quality requirements of all user equipment;

the macro base station will B_mThe sub-channels are allocated to the associated multicast group, and all the sub-channels b allocated to the multicast group associated with the macro base station_m，kThe sum of the numbers cannot exceed the number of available subchannels of the macro base station, i.e.

The number of subchannels assigned to the associated multicast group per small base station is B_sAll multicast groups associated with the small cell assigned subchannel b_j,kThe sum of the numbers cannot exceed the number of available subchannels of the small base station, i.e.

Multicast group in accordance with affiliation a_j，kIt is divided into two types:

the kth multicast group is served by the macro base station; a is_j，k1, the kth multicast group is served by the jth small base station; j is the total number of small base stations;

the method for dividing the transmitting power comprises the following steps:

sub-grouping: all users in the multicast group are divided into N subgroups according to ascending sequence of channel gain; the index of the subgroup ranges from 1 to N; the users of the 1 st subgroup can only receive the decoded base video layer, and the nth subgroup is matched with the nth NOMA layer, namely, the worst channel gain in the users of the nth subgroup is the channel gain of the nth NOMA layer; (n, k) denotes the nth subgroup or NOMA layer in the kth multicast group, and defines (n)_iK) is comprised of UE_iA subgroup of (a);

power distribution: for the video requested by the kth multicast group, each NOMA layer is allocated a certain power for transmitting information; for NOMA layer (n, k), if the power allocated to it is proportional to alpha_n，kThen the achievable transmission rate at that power is r_n，kThe corresponding constraint on power allocation is expressed as:

information transmission: after the sub-grouping and the power distribution, the base station multicasts different video layers through corresponding NOMA layers; if epsilon_l，n，kThe ith video layer is transmitted in video transmission layer (n, k) 1, else epsilon_l，n，k0, according to which the information transmission constraint is expressed as

And

wherein (7a) it is defined that any video layer is transmitted at most 1 time in all NOMA layers, and 7(b) it is defined that the information bit rate transmitted in each NOMA layer must not be greater than the achievable transmission rate of a user of the NOMA layer;

ε_l，n，kis a 0-1 indicator variable, if ε_l，n，kThe ith video layer is transmitted in video transmission layer (n, k) 1, else epsilon_l，n，k＝0；

The receiving end decodes the received signal through the SIC receiver, and the user equipment UE of the k multicast group_iDecoding is not higher than (n)_iNOMA layer of k), UE of k-th multicast group_iThe number of available video layers that can be received is indicated as

Problem modeling of the federated resource management problem

Under the constraints of the total available sub-channel number and the base station power, the joint resource management problem P0 is expressed as

3. The resource management method according to claim 2, characterized in that in step 1), the optimization problem P0 is decoupled into two sub-problems P1 and P2:

the area covered by the small base station is a hot spot area gathered by the user equipment; when the multicast group requesting a certain video is covered by a certain small base station, the multicast group belongs to the small base station; when more than 2 small base stations have multicast group affiliation, the opportunity of spectrum multiplexing exists, and the multicast group affiliated to the small base stations multiplexes the spectrum through spectrum segmentation:

suppose the coverage areas of the small base stations are all R_sIf the dependency is

d_i,jIndicating a request multicast group

The distance between the user of (1) and the small base station j;

if the spectrum resources of a given multicast group are b_kThen, the optimal power distribution of the multicast group is obtained, so that the maximum system utility of the multicast group is obtained; on the basis, an algorithm is further designed to solve B_m、B_sAnd all b_kA value of (d);

the P0 is divided into two sub-problems, P1 and P2, to solve.

Further solving B on the basis_m、B_sAnd all b_kA value of (d);

spectral resource b for a given kth multicast group_kThen problem P1 is described as:

will depend on the relation a_k,jSubstitution formula (11), then:

for macro base stations have

For small base station j there is

On the basis of P1, problem P2 is described as:

4. the resource management method according to claim 3, wherein in the step 2), the intra-group power allocation step:

spectral resource b of a given k-th multicast group_kBase station to a group of users

The multicast SVC video stream is transmitted as a video stream,

is arranged according to the ascending order of the user channel gain, I represents the number of users of the multicast group;

then 2.1) first solve two sets of variables: SVC video layer transmitted by each NOMA layer; power distribution coefficient of each NOMA layer;

2.1.1) first determine the SVC video layers transmitted within each NOMA layer:

assume that SVC video is coded as L_kThe layers are transmitted through N NOMA layers, a base video layer is marked as BL, an enhancement video layer is marked as EL1, EL2, EL3 and … … ELn in sequence, and all transmission schemes are obtained;

2.1.2) the power distribution coefficient of each NOMA layer is then determined:

given an information transmission scheme, the information transmission scheme will be

The power distribution coefficient alpha is solved according to the ascending sequence of the channel gain and the sequence of the high NOMA layer and the low NOMA layer_n，k；

First, the calculation of alpha is performed circularly from 1 to I +1_n，kFrom 1 to I represent

When the loop is to I +1, the I users in (1) indicate that none of the users can receive the NOMA layer (n, k), i.e., alpha_n，k0; calculating alpha circularly by the same theory_n-1，k；

Based on the minimum user quality of service requirement, all user equipments can receive the decoded NOMA layer (1, k) according to

The user with the worst channel gain correspondingly obtains alpha_1，k(ii) a For each set of power distribution coefficients { alpha ] that complies with the constraint (6)_1，k，α_2，k，α_3，k......，α_n，kCalculating the sum utility of the multicast groups;

2.2) calculating the utility of all information transmission schemes to obtain a maximum utility, namely the maximum utility of the multicast group under the given spectrum resource.

5. The resource management method according to claim 4, wherein in the step 3), the step of solving the problem P2 comprises:

3.1) given B_mAnd B_sCalculating the sub-channel distribution between the corresponding multicast groups when the utility of the macro base station and each small base station is maximum;

3.2) solving the spectrum segmentation among the base stations to obtain the maximized system utility;

first, assuming that b subchannels are available for the first k multicast groups, the maximum sum of the utility of the first k multicast groups is represented using a function S (k, b);

according to the dynamic programming idea, S (k, b) is expressed as a recursive formula, namely:

U_k(b) obtained by formula (11);

s (k, b) is solved through a multinomial selection knapsack algorithm;

the method for calculating the spectrum segmentation between the base stations comprises the following steps:

let the macro base station contain

A multicast group, a small base station j comprising

A plurality of multicast groups; respectively calculating the maximum and utility of the macro base station and the small base station j by the formula 15, and obtaining the maximum and utility of the system

Dividing the available spectrum resources into two parts B_tAnd B_iRespectively allocating the macro base station and the small base station for use; respectively obtaining the total utility of the macro base station and the small base station to obtain the total utility of the system, and finally finding out B capable of maximizing the total utility of the system through searching_tAnd B_iThe value of (2) is obtained, namely the spectrum segmentation of the macro base station and the small base station.