KR102275426B1 - Efficient resource allocation method and apparatus for satisfying rate-guarantee constraint in device-to-device underlaid cellular network - Google Patents

Abstract

An efficient resource allocation method and apparatus satisfying a transmission rate guarantee constraint in terminal to terminal communication coexisting with a cellular network are provided. A resource allocation method in terminal-to-terminal communication coexisting with a cellular network according to an embodiment of the present invention considers terminal-to-terminal (D2D) sharing cellular resources, and all user equipment (UE) and calculating a minimum channel bandwidth required for any match of a cellular terminal (CU) and a D2D terminal (DU) of a channel while achieving the determined target data rate. Here, the minimum channel bandwidth is not fixed and may be adaptively determined based on a triple match of the cellular terminal (CU), the D2D terminal (DU), and the channel.

Description

Efficient resource allocation method and apparatus satisfying transmission rate guarantee constraint conditions in terminal-to-terminal communication coexisting with a cellular network

Embodiments of the present invention relate to an efficient resource allocation method and apparatus satisfying a transmission rate guarantee constraint in terminal-to-terminal communication coexisting with a cellular network.

5G is expected to create new services in various industries. One of the newly activated services in 5G is proximity-based services related to vehicle-to-everything (V2X) networks, public safety networks, social networks, and the like. Recently, the demand for services requiring high speed and low latency is increasing. For example, in V2X communication, lane merging, bird's-eye view, and remote driving for parking are representative examples. In a public safety network, in order to smoothly perform remote monitoring through voice/video transmission and mission-critical tasks in emergency situations, data transmission rate of 1 Mbps or more and short delay conditions must be met.

Device-to-device (D2D) communication is a technology that directly communicates between mobile devices in a short distance for a mobile device. It has been recognized as an important technology to support. Here, the D2D terminal (D2D UE, DU) includes a D2D transmitter (DT) and a D2D receiver (D2D receiver, DR). Because direct communication between devices is possible, D2D communication is advantageous in terms of latency compared to conventional communication through network infrastructure such as 3GPP's Uu-interface. In addition, since the distance between the devices is relatively shorter than the distance between the device and the corresponding base station (BS), better link quality can be obtained, which is advantageous in terms of high data rate and reliability. In addition, when a D2D UE (DU) and a cellular UE (CU) share a spectrum, spectrum efficiency and network capacity may be further improved.

However, in the case of D2D communication sharing cellular resources, there is a possibility that serious interference between the CU and the DU may occur. Therefore, effective interference management measures must be applied to ensure a certain level of quality of service (QoS). Conventionally, various power control schemes and resource allocation schemes have been proposed to guarantee QoS requirements in terms of signal-to-interference noise ratio (SINR). In (Non-Patent Document 1), a power control technique for maximizing the sum of the achievable data rates of a pair of CUs and DUs was studied. However, the above technique is guaranteed to ensure that only the SINR of the CU is above the target SINR. In addition, only 1 CU and 1 DU were considered, so resource allocation or channel allocation issues were not addressed. On the other hand, when a plurality of CUs and a plurality of DUs exist, a power control and resource allocation scheme for maximizing the sum of data rates of terminals while guaranteeing an SINR threshold of all CUs and DUs has been proposed. In addition to power control and resource allocation, optimal mode selection for the DU was reviewed, and energy efficiency was used as a performance measure. However, conventional studies assume that each channel has a fixed bandwidth. In addition, it is assumed that each channel is already allocated to the CU before the resource allocation step. Therefore, the resource allocation problem dealt with in the prior studies is only to select the best DU for each CU.

The main goal of most of the prior studies was to maximize the sum of data rates, so the data rates of individual UEs were not considered in the resource allocation process. Although some prior studies have considered the lowest target SINR per UE, nevertheless, the data rate of each UE may not be guaranteed due to the fixed channel bandwidth. That is, the SINR threshold should be set high enough to satisfy the target data rate with a fixed channel bandwidth. In a D2D communication system sharing cellular resources, the UE experiences various interference situations according to CU-DU match and power control, and the fixed channel bandwidth assumption may not be beneficial. When the received SINR is low due to high interference, the fixed channel bandwidth may not be wide enough to guarantee the target data rate. Therefore, it is optimal to jointly manage transmit power, channel bandwidth, and channel allocation. However, there are currently few studies on minimizing the total channel bandwidth while ensuring that all UEs can achieve the target data rate.

C. Yu, K. Doppler, C. Ribeiro, and O. Tirkkonen, “Resource sharing optimization for device-to-device communication underlaying cellular networks,” IEEE Trans. Wireless Commun., vol. 10, no. 8, pp. 2752-2763, 2011 N. Z. Shor, Minimization Methods for Non-differentiable Functions, Springer Series in Computational Mathematics. Springer, 1985

Embodiments of the present invention describe an efficient resource allocation method and apparatus satisfying a transmission rate guarantee constraint in terminal-to-terminal communication coexisting with a cellular network, and more specifically, considering a D2D communication system sharing cellular resources, It provides a technique for minimizing the total channel bandwidth used at this time while ensuring that all user equipments (UEs) achieve a predetermined target data rate.

A resource allocation method in terminal-to-terminal communication coexisting with a cellular network according to an embodiment of the present invention considers a terminal-to-terminal (D2D) sharing a cellular resource, and a cellular terminal (CU) and a D2D terminal (DU) In this arbitrarily matched case, two user equipments (UEs) calculate a minimum channel bandwidth required to achieve a predetermined target data rate. Here, the minimum channel bandwidth is not fixed and may be adaptively determined based on a triple match of the cellular terminal (CU), the D2D terminal (DU), and the channel.

The step of calculating the minimum channel bandwidth required for a random match between a cellular terminal (CU) and a D2D terminal (DU) of the channel can be controlled through power control based on the signal-to-interference noise ratio (SINR) of the shared channel. have.

After obtaining the minimum channel bandwidth for all the cellular terminals (CUs), the D2D terminals (DUs) and channels, the cellular terminal (CU) to which each channel is allocated and the D2D terminal (DU) The method may further include allocating a channel to minimize bandwidth.

The step of allocating a channel to minimize the overall bandwidth includes applying Lagrangean relaxation, then decomposing a 3-D integer programming problem (IPP) into a 2-D linear programming problem (LPP), A channel suitable for a match between the UE (CU) and the D2D UE (DU) may be allocated.

An apparatus for allocating resources in terminal-to-terminal communication coexisting with a cellular network according to another embodiment of the present invention considers terminal-to-terminal (D2D) sharing cellular resources, and all user equipment (UE) in advance and a minimum channel bandwidth calculator for calculating a minimum channel bandwidth required for a random match between a cellular terminal (CU) and a D2D terminal (DU) while achieving the determined target data rate. Here, the minimum channel bandwidth is not fixed and may be adaptively determined based on a triple match of the cellular terminal (CU), the D2D terminal (DU), and the channel.

According to embodiments of the present invention, considering D2D sharing cellular resources, and minimizing the total channel bandwidth while all user terminals (UEs) must achieve a predetermined target data rate, the terminal to coexist with the cellular network It is possible to provide an efficient resource allocation method and apparatus satisfying a transmission rate guarantee constraint condition in terminal communication.

1 is a diagram for explaining a system model according to an embodiment of the present invention.
2 is a diagram for explaining an existing method of allocating channel resources.
3 is a diagram illustrating an example of a 3D channel resource allocation method according to an embodiment of the present invention.
4 is a flowchart illustrating a resource allocation method according to an embodiment of the present invention.
5 is a diagram for explaining an apparatus for allocating a resource according to an embodiment of the present invention.
6 is a diagram illustrating a visualization of an optimization problem according to an embodiment.

Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. However, the described embodiments may be modified in various other forms, and the scope of the present invention is not limited by the embodiments described below. In addition, various embodiments are provided in order to more completely explain the present invention to those of ordinary skill in the art. The shapes and sizes of elements in the drawings may be exaggerated for clearer description.

In the present embodiments, 'Cellular User Equipment (CU)' refers to user terminals belonging to a base station, for example, a smartphone, a tablet, and a wearable device located in the coverage of the base station. (wearable device), notebook computers, 2G/3G terminals, and the like.

In the present embodiments, 'D2D User Equipment (DU)' refers to an electronic device capable of direct communication, and may represent, for example, an electronic device forming an Internet of Things (IoT) network. That is, it may represent an electronic device capable of directly transmitting and receiving data between terminals without going through a base station.

Terminal-to-terminal (D2D) communication is an important technique for various proximity services. In addition to high data rates and high spectral efficiency, various applications, such as gaming and real-time audio/video transmission, increasingly require minimum data rates and tighter latency.

The following embodiments of the present invention consider D2D sharing cellular resources and aim to minimize the total channel bandwidth while all user terminals (UEs) must achieve a predetermined target data rate. The optimization problem relates to the match between cellular UE (CU) and D2D UE (DU), channel allocation and power control. The optimization problem is split into two sub-optimization problems, solving the power control and channel allocation problems separately. For any match of CU, DU, and channel, the minimum channel bandwidth of the shared channel is derived based on signal-to-interference noise ratio (SINR) based power control. Channel assignment (CU, DU, channel) is a triple three-dimensional (3-D) integer programming problem (IPP). After applying Lagrangean relaxation, the 3-D IPP is decomposed into a 2-D linear programming problem (LPP).

First, the system model and problem formulation are described.

system model

1 is a diagram for explaining a system model according to an embodiment of the present invention.

Referring to FIG. 1 , it shows D2D communication in which cellular resources are shared during uplink.

The cellular network 100 may include a user terminal belonging to the base station 110 (ie, a cellular terminal (CU) 120 ) and D2D terminal pairs, and the D2D terminal pairs are a D2D transmitter (DT) 131 . and a D2D receiver (D2D receiver, DR) 132 .

A frequency resource allocated for communication between the base station 110 and the cellular terminal 120 may be shared with the D2D terminal pairs 131 and 132 . In this case, the D2D transmitter DT 131 and the D2D receiver DR 132 belonging to the D2D terminal pair 131 and 132 may be located close to the base station 110 and the cellular terminal 120 . In addition, the D2D transmitter DT 131 and the D2D receiver DR 132 belonging to the D2D terminal pair 131 and 132 may be located far from the base station 110 and the cellular terminal 120 .

Here, consider a single cell in which N cellular terminals (CUs) 120 and M (NM) D2D transmitters (DTs, 131) are arbitrarily located in the cell. A D2D transmitter (DT, 131) and one D2D receiver (DR, 132) form a D2D terminal (DU), and all DRs 132 are located in the vicinity of the corresponding DT 131 .

Here, it is assumed that all CUs 120 and DTs 131 are equipped with a single transmit antenna. BS 110 and all DRs 132 are equipped with a single receive antenna. It is assumed that the network is provided by N orthogonal channels with variable channel bandwidth. All UEs are assigned only one channel. In addition, all channels are allocated to only one CU 120 or to a match composed of one CU 120 and one DU 131 and 132 . All DUs 131 and 132 must be matched with CU 120 to share a transport channel.

When channel n is shared by CU i (or CUi) and DU j (or DUj), CU i and DT j simultaneously transmit signals to BS and DR j (or DRj), respectively. Here, to indicate the channel gain of the link from the CU, the letter h is used here. Also, B is used as the BS index. Then, the channel gain of the link from CUi of channel n to BS is denoted by hi,B[n]. The letter g is used here to indicate the channel gain for the link of the DT. For example, gj,j[n] represents the channel gain of the link from DTj to DRj of channel n. The signal transmitted from DTj to DRj causes interference (CCI) in the BS, and the channel gain of the interference link is expressed as gj,B[n]. Similarly, hi,j[n] represents the channel gain of the interfering link from CUi to DRj.

2 is a diagram for explaining an existing method of allocating channel resources.

Referring to (a) of FIG. 2 , a channel allocation scheme in which each channel having a fixed channel bandwidth is pre-allocated to 1 CU before channel allocation is shown. In each channel, there are M boxes implying the channel bandwidth, and the intensity of the box color represents the sum of the data rates of the CU and DU matches.

For example, channel 1 is pre-allocated to CU1, and the color intensity varies depending on which DU shares channel 1. For channel 1, since the match between CU1 and DU2 is the best among M possible matches, the box linked to DU2 is marked with the darkest green. Similarly, DU1 and DUM are selected as the best match of CU2 and CUN, respectively. 2B shows the channel allocation result.

3 is a diagram illustrating an example of a 3D channel resource allocation method according to an embodiment of the present invention.

Referring to FIG. 3A , an example of a 3D channel resource allocation method according to an embodiment of the present invention is shown. Since the channel bandwidth corresponding to the size of each box is adaptively determined according to the channel capacity of each match, the CU and the DU can achieve a predetermined target data rate. When a channel is shared, the CCI and thus the channel capacity may vary depending on which CU and DU match. Thus, there are NM possible matches for each channel with different bandwidths, which are described as NM boxes.

When channels are allocated to minimize the total channel bandwidth, the best match per channel is determined by the channel bandwidth. For example, a match between CU2 and DU2 corresponding to the smallest box in channel 1 may be the best match in channel 1. Likewise, CU1 and DUM may be the best match on channel 2, and CUN and DU1 may be the best match on channel N.

3B shows the channel allocation result according to the channel bandwidth. Here, the DU selected as the best match in the previous selection step may also be the best match with the later selection step. For example, DUM may be the best match for CU3. Therefore, in order to find the best channel assignment, it is necessary to search the total possible N ² M (N x N x M) matches.

4 is a flowchart illustrating a resource allocation method according to an embodiment of the present invention.

Referring to FIG. 4 , a resource allocation method in terminal-to-terminal communication coexisting with a cellular network according to an embodiment of the present invention considers terminal-to-terminal (D2D) sharing cellular resources, and all user terminals (UEs) ) while achieving a predetermined target data rate, calculating the minimum channel bandwidth required for any match between the cellular terminal (CU) and the D2D terminal (DU) ( S110 ).

In addition, after obtaining the minimum channel bandwidth for all cellular terminals (CUs), D2D terminals (DUs) and channels, the total bandwidth is minimized by matching the cellular terminals (CUs) and D2D terminals (DUs) allocated to each channel The step of allocating a channel to do so (S120) may be further included.

The resource allocation method according to an embodiment of the present invention may be described in more detail with the example of the resource allocation apparatus according to an embodiment of the present invention.

5 is a diagram for explaining an apparatus for allocating a resource according to an embodiment of the present invention.

Referring to FIG. 5 , the resource allocation apparatus 500 in terminal-to-terminal communication coexisting with a cellular network according to an embodiment of the present invention may include a minimum channel bandwidth calculating unit 510, and in an embodiment Accordingly, the 3D channel allocator 520 may be further included.

In step S110, the minimum channel bandwidth calculating unit 510 considers the terminal to terminal (D2D) sharing the cellular resource, and while all the user terminals (UE) achieve a predetermined target data rate, the cellular terminal (CU) ) and a minimum channel bandwidth required for a random match between the D2D UE (DU) can be calculated. Here, the minimum channel bandwidth is not fixed, and may be adaptively determined based on a triple match of a cellular terminal (CU), a D2D terminal (DU), and a channel.

The minimum channel bandwidth of the shared channel can be controlled through power control based on the signal-to-interference noise ratio (SINR). Conventionally, since the UE does not consider interference by controlling the power with SNR, the power of the UE is controlled regardless of how the cellular UE (CU) and the D2D UE (DU) are paired. By controlling the cellular terminal (CU) and the D2D terminal (DU), the optimal power control can be performed whenever a pair is established. Such power control is a complex problem to be solved iteratively, and it can be simply obtained in a closed-form form through two approximations. At this time, the approximation used is based on the fact that the noise power is sufficiently small and the SINR values of the terminals are mainly larger than 1. In addition, the channel allocation problem is a complex three-dimensional problem of finding a 3-tuple combination of a D2D terminal, a cellular terminal, and a channel, which can be solved by changing it to Lagrangean relaxation and two two-dimensional problems.

In step S120, the 3D channel allocator 520 obtains the minimum channel bandwidth for all cellular terminals (CUs), D2D terminals (DUs), and channels, and then cellular terminals (CUs) to which each channel is allocated. A channel may be allocated to minimize the overall bandwidth by matching the D2D terminal (DU) with the D2D terminal (DU).

In order to minimize the total channel bandwidth, when a channel is allocated, a match of a CU and a DU (D2D UE) per channel may be determined according to the channel bandwidth. And, as the channel bandwidth is adaptively determined according to the channel capacity of the match of each CU and DU, the CU and the DU may achieve a predetermined target data rate.

To reduce computational complexity, we also propose an optimal transmit power control for deriving the minimum channel bandwidth for a random match and a method for deriving sub-optimal solutions to the 3-D channel allocation problem, respectively. The optimal transmit power control problem proposes a sub-optimal solution in a closed-form using approximation. The 3-D channel allocation problem can be calculated by decomposing it into two sub-optimization problems. The three-dimensional channel allocator 520 applies Lagrangean relaxation, and then decomposes the 3-D integer programming problem (IPP) into a 2-D linear programming problem (LPP), each cellular terminal (CU) and A channel suitable for the match of the D2D terminal (DU) may be allocated.

The present invention aims to ensure that all UEs can achieve a target data rate with a minimum total channel bandwidth with an appropriate power control scheme. Since most existing tasks assume that each CU has a pre-allocated channel, the channel allocation problem is to match the highest DU for each CU to maximize the sum of data rates. In this case, one of M DUs is selected for each CU. However, according to the present invention, at the moment of determining each channel allocation, channels can be reallocated not only for DUs but also for CUs, which increases the possibility of selecting a better channel. This 3-D channel assignment can improve performance at the expense of increased computational complexity. To alleviate the complexity, Lagrangean relaxation and sub-step iterative algorithms can be used. We propose a sub-optimization algorithm below.

Existing work focuses on maximizing the sum of data rates, so the transmit power of the UE is determined to ensure SINR. However, since the channel bandwidth is fixed, this method is not efficient to secure the target data rate. As the SINR changes, the channel capacity changes, so it is necessary to adjust the channel bandwidth required to ensure the target data rate. At the same time, since the channel bandwidth of the shared channel is determined by the user with the worse SINR, it is necessary to control the power to maximize the worse SINR. A precise solution to power control involves iterative computation, for example, through a route finding algorithm. To reduce the complexity, a sub-optimization algorithm that can compute the power in a closed-form is proposed.

In an embodiment of the present invention, a system in which the channel bandwidth is not fixed, and is adaptively determined based on a triple match of CU, DU, and channel, unlike many existing studies, is considered. In more detail, the minimum channel bandwidth for guaranteeing the target data rate is derived while satisfying the maximum transmission power constraint. The minimum channel bandwidth is different for each match of CU and DU. Therefore, the remaining problem is a three-dimensional (3-D) allocation problem that guarantees a target data rate while minimizing the overall channel bandwidth. Since it is a mixed integer nonlinear programming (MINLP) problem, Lagrangean relaxation is applied, and then the relaxed problem is decomposed into two suboptimal problems. Finally, this two-dimensional (2-D) problem is iteratively solved based on the subgradient algorithm.

problem formulation

와 R

로 표시된다. CUi는 P

의 전력으로 신호를 BS에 전송하고, DTj는

의 전력으로 신호를 DRj에 전송한다. 그런 다음, Cui가 대역폭 b

[n]을 할당받은 경우, BS의 신호 대 간섭 잡음비(SINR)는 다음과 같이 표현할 수 있다.Calculate bi,j[n], which is the minimum channel bandwidth required for any match of CUi and DUj of channel n to ensure the target data rate. The target data rates of CU and DU are R

and R

is displayed as CUi is P

transmit a signal to the BS with the power of , DTj is

transmit a signal to DRj with a power of Then, Cui is the bandwidth b

When [n] is assigned, the signal-to-interference noise ratio (SINR) of the BS can be expressed as follows.

는 잡음 전력 밀도이다. 그런 다음, 달성 가능한 데이터 전송률 R

[n]이 R

를 보장해야 한다.where σ

is the noise power density. Then, the achievable data rate R

[n] is R

should ensure

[n]을 할당받은 경우 DRj의 SINR은 다음과 같이 나타낼 수 있다.Similarly, DTj is the bandwidth b

When [n] is allocated, the SINR of DRj can be expressed as follows.

And the following conditions must be satisfied:

와 R

를 동시에 보장하려면, bi,j[n]을 다음과 같이 결정해야 한다.Then, R

and R

To guarantee at the same time, bi,j[n] must be determined as follows.

와 R

를 보장한다. 총 채널 대역폭을 최소화하려면 가능한 총 N²M 일치 항목을 조사해야 한다. 또한, CU와 DT의 전송 전력은 최적으로 결정되어야 한다.Then, allocate channels to minimize the total channel bandwidth while R for CU and DU, respectively.

and R

to ensure To minimize the total channel bandwidth, the possible total N ² M matches should be investigated. In addition, the transmit power of the CU and the DT should be optimally determined.

Ωc={1,...,N} and Ωd={1,...,M} are introduced respectively to represent a set of CUs and DUs. In addition, xi,j[n], which is a channel matching index, is introduced. If CUi and DUj share channel n, xi,j[n] is set to 1, otherwise 0.

For simplicity, it is assumed here that N=M, that is, the number of CUs is equal to that DU. If N>M, (N-M) CUs do not match any DUs. If a CU does not match, the CU uses the channel independently. Which CUs independently decide to use the channel is also a resource allocation issue, but this is a trivial and simple matter.

Equation (7a) represents a condition in which all CUs of Ωc must match one DU in one channel. Equation (7b) indicates a condition that all DUs must match one CU in one channel. Equation (7c) represents a condition that all channels should be allocated only once. Meanwhile, Equations (7d) and (7e) are conditions for guaranteeing a target data rate. Equation (7f) satisfies the binary condition. Equations (7g) and (7h) are power limiting conditions having a maximum transmit power of Pmax for CU and DU, respectively.

The data rate guarantee resource allocation scheme will be described below.

Equation (7) is a mixed integer nonlinear programming (MINLP) problem, and its solution is very complex. To reduce computational complexity, it can be decomposed into two sub-optimization problems: 1) optimal transmit power control to derive the minimum channel bandwidth for a random match, and 2) 3-D channel allocation problem.

Minimum Channel Bandwidth for Random Match

와 R

의 목표 데이터 전송률을 보장하도록 채널 대역폭을 조정할 필요가 있다. 필요한 채널 대역폭을 최소화하기 위한 최적의 전력 할당 문제를 다음과 같이 공식화할 수 있다.First, when determining the minimum channel bandwidth for a random match of (CUi, DUj, channel n), since the channel capacity is different according to the received SINR, R for CU and DU, respectively

and R

It is necessary to adjust the channel bandwidth to ensure the target data rate of The optimal power allocation problem to minimize the required channel bandwidth can be formulated as

및

는 다음과 같이 주어진 전송 전력 밀도이다.When solving the above optimization problem, conditions (2), (4), (7g), and (7h) must be satisfied.

and

is the transmit power density given by

가 주어질 때,

가 증가할수록

는 반드시 감소한다. 또한,

가 주어진 경우

는

에 비례하여 반드시 증가한다. 흥미롭게도,

가 증가하고

가 동시에 증가한다면,

는 반드시 증가하거나 아니면 반드시 감소한다. 즉, 정의역 내 어떤 두 지점을 연결하는 선 세그먼트를 고려할 때, 상기 선 세그먼트에 상응하는 값들 중 최대값은 항상 세그먼트의 끝점(end point) 중 하나에서 찾아진다. 그렇지 않으면

과 같은 지점이 존재해야 하는데, 이는 사실이 아니다. 이는

과

이 모두

및

에 관한 quasi-convex 함수임을 의미한다. 두 개의 quasi-convex 함수 중 최대값을 취하는 함수도 또한 quasi-convex라는 것은 잘 알려져 있는 사실이다. 즉, 수학식 (8)은 quasi-convex 최적화 문제이다. 일반적으로 이 문제는 root-finding 알고리즘을 통해 해결할 수 있다.

When is given,

as it increases

is necessarily reduced. Also,

if given

is

necessarily increases in proportion to Interestingly,

is increasing

If at the same time increases,

must either increase or must decrease. That is, when considering a line segment connecting two points in a domain, the maximum value among the values corresponding to the line segment is always found at one of the end points of the segment. Otherwise

There should be points such as , which is not true. this is

and

all of this

and

It means that it is a quasi-convex function for Of the two quasi-convex functions, it is well known that the function that takes the maximum is also quasi-convex. That is, Equation (8) is a quasi-convex optimization problem. In general, this problem can be solved through the root-finding algorithm.

와

의 함수로 달성 가능한 SINR을 조사한다. 채널 대역폭은 수학식 (2) 및 수학식 (4)와 같이 SINR에 반비례한다. R

=R

인 경우

는 아래 (10)을 풀어서 얻을 수 있다.To reduce the computational complexity, first

Wow

Examine the achievable SINR as a function of The channel bandwidth is inversely proportional to the SINR as shown in Equations (2) and (4). R

=R

if

can be obtained by solving (10) below.

It should be noted that Equation (10) is quasi-concave, and Equation (8) is quasi-convex.

6 is a diagram illustrating a visualization of an optimization problem according to an embodiment.

=-174dB/mHz로 가정한다. 여기에서는 R

=R

=1Mbps라고 가정한다. 정의역은 수학식 (9)를 수학식 (7g) 및 수학식 (7h)로 대체하여 결정되며, 이는 최대 전송 전력 조건에 해당한다. 특히, CUi 또는 DTj (또는 둘 다)가 Pmax를 사용하는 경우, (

,

)는 검은 굵은 선으로 도해되는 경계선에 위치한다.As shown in FIG. 6 , the optimization problem of Equation (10) can be visualized. where hi,B[n]=-100dB, gj,j[n]=-108dB, hi,j[n]=139dB, gj,B[n]=115dB, Pmax=23dBm,

=-174dB/mHz is assumed. R here

=R

= 1Mbps. The domain is determined by replacing Equation (9) with Equations (7g) and (7h), which corresponds to the maximum transmit power condition. In particular, if CUi or DTj (or both) use Pmax, (

,

) is located on the boundary line illustrated by the thick black line.

,

,

,

, bi,j[n] 및 이 세 지점의

와

를 나타낸다. In FIG. 6 , three representative points p1, p2, and p3 may be considered. Table 1 is

,

,

,

, bi,j[n] and of these three points

Wow

indicates

[Table 1]

<

인 지점에서 bi,j[n] 채널 대역폭 bi,j[n]는 CUi(예: bi,j=

=304.9kHz에 의해 결정된다. 따라서 bi,j[n]을 최소화하기 위해서는

를 증가시켜

을 개선해야 한다.

가 증가하면 p1 지점이 화살표 A 방향으로 이동한다.

이 개선되는 반면, CUi에서 DRj까지의 간섭이 증가하여

이 감소한다. 따라서 p1은

과

이 같은 지점, 즉 p2 지점에 도달한다.

=

지점에서 nCUi와 DUj의 채널 대역폭은 동일하다. 예: bi,j=

=

=135.4kHz. 화살표 B 방향으로

가 더 증가해도

는 여전히 개선되지만,

가 감소한다. 결과적으로, min[

,

]는 감소한다. 그러면 bi,j[n]이 증가하게 된다. 따라서

을 저하시키지 않고

를 개선하려면, min[

,

]이 감소되는 것이 방지되도록,

=

이 유지되어야 한다.

<

At the point bi,j[n] the channel bandwidth bi,j[n] is CUi (e.g. bi,j=

=304.9 kHz. Therefore, in order to minimize bi,j[n],

to increase

should be improved

When is increased, point p1 moves in the direction of arrow A.

While this is improved, the interference from CUi to DRj increases

this decreases So p1 is

and

This same point, ie, point p2, is reached.

=

At a point, the channel bandwidths of nCUi and DUj are the same. Example: bi,j=

=

=135.4kHz. in the direction of arrow B

even if it increases further

is still improved, but

is decreased As a result, min[

,

] decreases. Then bi,j[n] increases. therefore

without degrading

To improve on min[

,

] is prevented from being reduced,

=

this should be maintained

=

을 유지하도록

와

를 동시에 증가시키면 p2 지점이 곡선 화살표 C 방향으로 이동한다. p2 지점이 화살표 C를 따라 계속 움직이면 마침내 p3 지점의 경계선과 만난다. 경계선에 도달한다는 것의 물리적인 의미는 CUi 또는 DTj가 최대 전송 전력인 Pmax를 사용한다는 것을 의미한다. 그러면

과

을 동시에 개선하는 것은 더 이상 불가능하게 된다. 따라서 p3 지점은 bi,j[n]을 최소화하는 최적 전력 지점이다. 일 실시예에서, 표 1과 같이 최대 전송 전력을 활용하는것은 CUi이고 이때 전송 전력은

=

bi,j[n]=23dBm 이다.

=

to keep

Wow

Simultaneously increasing , moves the point p2 in the direction of the curved arrow C. If point p2 continues moving along arrow C, it finally meets the boundary of point p3. The physical meaning of reaching the boundary line means that CUi or DTj uses the maximum transmit power, Pmax. then

and

It is no longer possible to improve both at the same time. Therefore, point p3 is the optimal power point that minimizes bi,j[n]. In one embodiment, it is CUi that utilizes the maximum transmit power as shown in Table 1, where the transmit power is

=

bi,j[n]=23dBm.

=R

의 경우, 최적 전력 설정은

=

조건을 만족해야 하며, 따라서 다음 식과 같이 표현할 수 있다.R

=R

In this case, the optimal power setting is

=

The condition must be satisfied, so it can be expressed as the following equation.

≠R

에도 유지됨을 쉽게 알 수 있다. 이제 남은 문제는 CU 또는 DT 중 하나가 최대 송신 전력에 도달할 때까지 전송 전력을 최대로 증가시키는 동시에 수학식 (11)을 만족시키는 것이다. 채널 대역폭은 CU와 DT의 전송 강도뿐만 아니라 채널 대역폭에 따라 달라지기 때문에 최적의 전력은 예를 들어 루트 찾기 알고리즘을 통해 반복적으로 얻을 수 있다.The state of Equation (11) is R

≠R

It can be easily seen that the The remaining problem now is to satisfy Equation (11) while maximally increasing the transmit power until either the CU or the DT reaches the maximum transmit power. Since the channel bandwidth depends on the channel bandwidth as well as the transmission strength of the CU and DT, the optimal power can be iteratively obtained, for example, through a route finding algorithm.

≫1과

≫1을 가정하여 차선의 솔루션을 제공한다. 그런 다음 수학식 (11)을 다음과 같이 근사하게 추정할 수 있다.In order to reduce computational complexity, at the optimal

≫ Lesson 1

» Assume 1 to provide a suboptimal solution. Then, Equation (11) can be approximated as follows.

인 경우

=Pmax가 되는 것을 쉽게 도출할 수 있다. 이 후,

는는 수학식 (12), 수학식 (11)에 의해 얻을 수 있다.

일 때,

=Pmax 가 된다.after that

if

=Pmax can be easily derived. after,

can be obtained by Equations (12) and (11).

when,

=Pmax.

3D Channel Assignment

을 획득한 후, 채널 할당 문제는 1) CU-DU 매칭 문제와2) CU-DU 매치에 어떤 채널을 할당할 건지 결정하는 문제이며 다음과 같이 정수 프로그래밍 문제(IPP)로 공식화할 수 있다.For all i∈Ωc, j∈Ωd, and N channels

After obtaining , the channel allocation problem is 1) a CU-DU matching problem and 2) a problem of deciding which channel to assign to a CU-DU match, which can be formulated as an integer programming problem (IPP) as follows.

When solving the above optimization problem, Equations (7a), (7b), (7c), and (7f) must be satisfied. Equation (13) is 3-D for the triplet (i,j,n) of x(i,j)[n] for all i∈Ωc, j∈Ωd and n=1,...,N It's a channel allocation problem. The optimal solution can be found with a branch-and-bound algorithm. However, Equation (13) is an NP-hard problem and requires enormous computational complexity.

In the present specification, an algorithm for finding the suboptimal solution of Equation (13) with lower complexity is proposed. To express the problem in normal form, two column vectors b=[b1,...,b _N ² _M ] ^T and x=[x1,...,x _N ² _M ) ] ^T are introduced. The {N ² (j-1)+N(i-1)+n}th element of the vector is the same as the (i,j,n)th element of the 3D matrix. Then, Equation (13) can be expressed again as follows.

Equation (7a) can be expressed as follows.

는 Kronecker 곱을 나타낸다. 마찬가지로 수학식 (7b)는 다음과 같이 다시 나타낼 수 있다.Here, IN,K=[IN,...,IN] is a matrix composed of K INs, IN is an identity matrix of size NХN, and 1N is a vector of size NХ1 and all elements are 1.

denotes the Kronecker product. Similarly, Equation (7b) can be expressed again as follows.

Here, Em represents an MХN matrix in which all elements of the mth row are 1 and all other elements are 0. By introducing ϕn(x)=∑(i∈Ωc)∑(j∈Ωd)xi,j[n], Equation (7c) can be expressed again as follows.

Here, (14c) is approximated by introducing an element in which the Lagrange multiplier, u=[u1,...,uN] ^{T , that is, un≥0,∀n is both positive.} Then we can formulate the double problem of Equation (14).

When solving the above optimization problem, conditions (14a), (14b), and (7f) must be satisfied.

NХN ² M matrix of Equation (14a) and ² M NХN matrix of Equation (14b) is totally unimodular (TU). Then, by converting the binary variable into a continuous variable, the optimization problem of Equation (15) can be changed to a linear programming problem (LPP) having polynomial time complexity.

를 수학식 (15)의 해라고 하자. 라그랑지안 릴랙션(Lagrangean relaxation)로 인해, 모든 채널을 한 번 이상 사용할 수 없다는 제약은 엄격히 충족되지 않는다. 따라서

에서 CUi와 DUj의 매칭 정보인 yi,j를 추출한다. 매칭이 된 경우 yi,j=1, 그렇지 않으면 yi,j=0이다;when u is given

Let be a solution of Equation (15). Due to Lagrangean relaxation, the constraint that all channels cannot be used more than once is not strictly met. therefore

Extracts yi,j, which is matching information between CUi and DUj, from If there is a match, yi,j=1, otherwise yi,j=0;

yi,j does not convey information about the channel assigned to the match. Therefore, the entire channel allocation can be decomposed as follows.

j[n]는 채널 n의 DUj에 할당된 채널 대역폭을 표시한다. 그러면,

j[n]는 다음에 의해 얻을 수 있다.Here, ηj[n] is a channel allocation index. If channel n is assigned to DUj then ηj[n]=1, otherwise ηj[n]=0. Given yi,j according to equation (17), the original problem in equation (14) can be expressed again with respect to ηj[n]. for teeth

j[n] indicates the channel bandwidth allocated to DUj of channel n. then,

j[n] can be obtained by

와 η을 각각

j[n]와 ηj[n]에서 파생된 크기 NM의 열 벡터라 한다. 이제 수학식 (14)의 문제를 LPP로 재구성하여 각 매치에 맞는 최적의 채널을 찾을 수 있게 되었다.

and η respectively

Let it be a column vector of magnitude NM derived from j[n] and ηj[n]. Now, by reconstructing the problem of Equation (14) with LPP, it is possible to find the optimal channel for each match.

Here, Equation (19a) is a constraint that only one channel is allocated to a single match of a CU-DU, and Equation (19b) is a constraint that one channel cannot be used more than once. Since the matrices of Equations (19a) and (19b) are TUs, the binary constraint corresponding to Equation (19c) is relaxed as a positive real condition as in Equation (19c).

)이므로, 다음과 같이 업데이트할 수 있다.Given u, L(u) is the lower bound of ^{b * due to a positive Lagrangean multiplier.} That is, L(u)≤b ^* . In contrast, l2 is not always less than ^b*. Here we use the subgradient algorithm to update u . A detailed description of the subgradient algorithm can be found in (Non-Patent Document 2). The subgradient of L(u) in un is φn(

), so it can be updated as follows.

는 스텝 크기이다. 제안된 채널 할당 체계에 대한 상세한 절차는 알고리즘 1에 제시되어 있다. 마지막으로,

[n]에 따라 완전한 채널 할당 정보를 얻는다.here,

is the step size. The detailed procedure for the proposed channel allocation scheme is presented in Algorithm 1. Finally,

Complete channel allocation information is obtained according to [n].

[Algorithm 1]

The device described above may be implemented as a hardware component, a software component, and/or a combination of the hardware component and the software component. For example, devices and components described in the embodiments may include, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable array (FPA), It may be implemented using one or more general purpose or special purpose computers, such as a programmable logic unit (PLU), microprocessor, or any other device capable of executing and responding to instructions. The processing device may execute an operating system (OS) and one or more software applications running on the operating system. The processing device may also access, store, manipulate, process, and generate data in response to execution of the software. For convenience of understanding, although one processing device is sometimes described as being used, one of ordinary skill in the art will recognize that the processing device includes a plurality of processing elements and/or a plurality of types of processing elements. It can be seen that can include For example, the processing device may include a plurality of processors or one processor and one controller. Other processing configurations are also possible, such as parallel processors.

The software may comprise a computer program, code, instructions, or a combination of one or more thereof, which configures a processing device to operate as desired or is independently or collectively processed You can command the device. The software and/or data may be any kind of machine, component, physical device, virtual equipment, computer storage medium or device, to be interpreted by or to provide instructions or data to the processing device. may be embodied in The software may be distributed over networked computer systems, and stored or executed in a distributed manner. Software and data may be stored in one or more computer-readable recording media.

The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions recorded on the medium may be specially designed and configured for the embodiment, or may be known and available to those skilled in the art of computer software. Examples of the computer-readable recording medium include magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic such as floppy disks. - includes magneto-optical media, and hardware devices specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine language codes such as those generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like.

As described above, although the embodiments have been described with reference to the limited embodiments and drawings, various modifications and variations are possible for those skilled in the art from the above description. For example, the described techniques are performed in a different order than the described method, and/or the described components of the system, structure, apparatus, circuit, etc. are combined or combined in a different form than the described method, or other components Or substituted or substituted by equivalents may achieve an appropriate result.

Therefore, other implementations, other embodiments, and equivalents to the claims are also within the scope of the following claims.

Claims (5)

In a resource allocation method in terminal-to-terminal communication coexisting with a cellular network,
Considering terminal-to-terminal (D2D) sharing cellular resources, while all user equipment (UE) achieve a predetermined target data rate, for any match of a cellular terminal (CU) and a D2D terminal (DU) calculating a minimum required channel bandwidth; and
After obtaining the minimum channel bandwidth for all the cellular terminals (CUs), the D2D terminals (DUs) and channels, the cellular terminal (CU) to which each channel is allocated and the D2D terminal (DU) Allocating channels to minimize bandwidth
including,
The minimum channel bandwidth is not fixed, and is adaptively determined based on a triple match of the cellular terminal (CU), the D2D terminal (DU), and a channel;
Allocating a channel to minimize the overall bandwidth comprises:
After applying Lagrangean relaxation, a 3-D integer programming problem (IPP) is decomposed into a 2-D linear programming problem (LPP), and the cellular terminal (CU) and the D2D terminal (DU) are matched allocating the right channel for
characterized in that, resource allocation method. According to claim 1,
Calculating the minimum channel bandwidth required for a random match between a cellular terminal (CU) and a D2D terminal (DU) of the channel comprises:
Controlling the minimum channel bandwidth of the shared channel through power control based on a signal-to-interference noise ratio (SINR)
characterized in that, resource allocation method. delete delete An apparatus for allocating resources in terminal-to-terminal communication coexisting with a cellular network, the apparatus comprising:
A random match of a cellular terminal (CU) and a D2D terminal (DU) of a channel, while considering the terminal to terminal (D2D) sharing cellular resource, and all user equipment (UE) achieve a predetermined target data rate a minimum channel bandwidth calculator for calculating a minimum channel bandwidth required for ; and
After obtaining the minimum channel bandwidth for all the cellular terminals (CUs), the D2D terminals (DUs) and channels, the cellular terminal (CU) to which each channel is allocated and the D2D terminal (DU) 3D channel allocator to allocate channels to minimize bandwidth
including,
The minimum channel bandwidth is not fixed, and is adaptively determined based on a triple match of the cellular terminal (CU), the D2D terminal (DU), and a channel;
The three-dimensional channel allocator,
After applying Lagrangean relaxation, a 3-D integer programming problem (IPP) is decomposed into a 2-D linear programming problem (LPP), and the cellular terminal (CU) and the D2D terminal (DU) are matched allocating the right channel for
characterized in that, the resource allocation device.

Images