KR20230088591A - Resolution optimization method of cell-free massive mimo and system using same - Google Patents

Abstract

The present invention includes an uplink channel setting step in which an i-th uplink defines a channel with respect to an m-th access point (AP) by [Relational Expression 1]; an uplink spectral efficiency setting step of defining spectral efficiency (SE) of the i-th uplink as a closed-type solution obtained by substituting [Equation 1] into [Equation 2]; and an uplink resolution derivation step of extracting a resolution having maximum energy efficiency using the uplink spectral efficiency defined in the uplink spectral efficiency setting step and energy efficiency (EE) defined by [Equation 3]. It is characterized by including;

Description

Method for optimizing resolution of cell-free massive multiple-input (MIMO) transmission device and system using the same

The present invention relates to wireless communication, and more particularly, to a method for optimizing a resolution of a cell-free massive multiple input (MIMO) transmission device for optimizing transmission performance and a system using the same.

To satisfy the requirements of the 5G system, massive MIMO, HetNet, and full duplex technologies are being evaluated as key elements. Among them, massive MIMO technology is considered the most important technology for 5G communication as it can increase the throughput of the system by simply installing massive MIMO in an existing base station.

When transmitting and receiving data using the MIMO system, if the transmitting end and the receiving end know the state of the transmission channel, performance can be improved by designing a transmit/receive filter suitable for the purpose. In this case, all communication nodes must know information about the MIMO channel, and through this, signal processing can be performed to eliminate interference from neighboring nodes and the effect of multiple channels. In a MIMO communication system, it is easy to estimate whether or not the receiving end is a channel through a pilot signal, but channel recognition at the transmitting end is not only difficult to estimate because the channel information must be fed back, but also can cause complexity due to signal exchange. .

Unlike the conventional half-duplex method, the full-duplex method simultaneously transmits and receives a transmission signal and a reception signal in the same time resource in the same band. As a result, a self-interference signal is inevitably generated, and a self-interference cancellation function to remove it must be implemented in the transmitting/receiving terminal.

1 illustrates a cell-free massive multiple input/output (MIMO) method according to an embodiment of the present invention. Referring to FIG. 1 , in a cell-free MIMO system, a plurality of access points (APs) are distributed over a wide area to jointly service a small number of user equipments (UEs). A cell-free MIMO system has recently attracted attention, and it is advantageous for spectral/energy efficiency (SE/EE) to use a full-duplex method rather than a half-duplex method in a cell-free MIMO system.

However, the full-duplex cell-free MIMO system has problems in that hardware cost and power consumption increase because a high-resolution analog-to-digital converter (ADC) must be used to improve SE/EE. In the full-duplex cell-free MIMO system, power consumption increases in proportion to the sampling frequency (Fs). In a full-duplex cell-free MIMO system, power consumption increases exponentially according to resolution. In addition, the manufacturing cost of an ADC increases with resolution. Designing systems with high-resolution ADCs is not realistic given the huge number of access points (APs) used in cell-free MIMO systems. Therefore, a full-duplex cell-free MIMO system needs to be implemented with a low-resolution ADC.

In a full-duplex cell-free MIMO system, quantization noise (QN) of a base station may deteriorate due to self-interference (SI) when a low-resolution ADC is used. Accordingly, there is a demand for a method for deriving a resolution of a low-resolution ADC capable of maximizing spectral/energy efficiency (SE/EE) in a full-duplex cell-free MIMO system.

Korea Patent No. 1790498

An object of the present invention is to provide a method for optimizing the resolution of a cell-free massive multiple input (MIMO) transmission device using a low resolution ADC and a system using the same.

In order to achieve the above object, the present invention provides an uplink channel setting step in which an i-th uplink defines a channel for an m-th access point (AP) by [Relational Expression 1]; an uplink spectral efficiency setting step of defining spectral efficiency (SE) of the i-th uplink as a closed-type solution obtained by substituting [Equation 1] into [Equation 2]; and an uplink resolution derivation step of extracting a resolution having maximum energy efficiency using the uplink spectral efficiency defined in the uplink spectral efficiency setting step and energy efficiency (EE) defined by [Equation 3]. It is characterized by including;

Preferably, a downlink channel setting step in which the k-th downlink channel is defined by [Relationship 2] with respect to the m-th access point (AP); may be further included.

Preferably, a downlink spectral efficiency setting step in which the spectral efficiency (SE) of the k-th downlink is defined as a closed solution obtained by substituting [Equation 4] into [Equation 5]; can include

Preferably, a downlink that extracts a resolution having maximum energy efficiency using the downlink spectrum efficiency defined in the downlink spectral efficiency setting step and energy efficiency (EE) defined by [Equation 3] Resolution derivation step; may further include.

In addition, the present invention includes an uplink channel setting unit for defining a channel for an i-th uplink with respect to an m-th access point (AP) by [Relationship Equation 1]; an uplink spectral efficiency setting unit defining spectral efficiency (SE) of the i-th uplink as a closed solution obtained by substituting [Equation 1] into [Equation 2]; and an uplink resolution derivation unit for extracting a resolution having maximum energy efficiency using the uplink spectral efficiency defined by the uplink spectral efficiency setting unit and energy efficiency (EE) defined by [Equation 3]; It is another feature that includes.

According to the present invention, it is possible to extract the optimal resolution of a low-resolution ADC through a closed solution of spectral / energy efficiency (SE / EE) without directly implementing and simulating a cell-free massive multiple-input (MIMO) system. There is an advantage.

1 illustrates a cell-free massive multiple input/output (MIMO) method according to an embodiment of the present invention.
2 shows a flowchart for deriving an optimal resolution of an ADC used to receive a signal transmitted in an uplink in an access point (AP) according to an embodiment of the present invention.
FIG. 3 shows a flowchart for deriving an optimal resolution of an ADC used to receive a signal transmitted from an access point (AP) in a downlink according to an embodiment of the present invention.
4 is a graph illustrating a change in sum spectral efficiency (SE) with respect to the number of access points (APs) according to an embodiment of the present invention.
5 is a graph showing a change in sum spectral efficiency (SE) with respect to resolution (b) of an ADC according to an embodiment of the present invention.
6 is a graph illustrating a change in sum spectral efficiency (SE) with respect to residual self-interference (SI) according to an embodiment of the present invention.
7 is a graph of energy efficiency (EE) of an uplink, and FIG. 7 (a) shows a graph showing a change in energy efficiency (EE) with respect to resolution (b) of an ADC. 7(b) shows a graph showing the change in energy efficiency (EE) for sum spectral efficiency (SE).
8 is a graph of downlink energy efficiency (EE), and FIG. 8 (a) shows a graph showing a change in energy efficiency (EE) with respect to ADC resolution (b). 8(b) shows a graph showing the change in energy efficiency (EE) for sum spectral efficiency (SE).

Hereinafter, the present invention will be described in detail with reference to the contents described in the accompanying drawings. However, the present invention is not limited or limited by exemplary embodiments. The same reference numerals in each figure indicate members performing substantially the same function.

The objects and effects of the present invention can be naturally understood or more clearly understood by the following description, and the objects and effects of the present invention are not limited only by the following description. In addition, in describing the present invention, if it is determined that a detailed description of a known technology related to the present invention may unnecessarily obscure the subject matter of the present invention, the detailed description will be omitted.

개의 업링크와

의 다운링크를 포함할 수 있다. 각각의 엑세스 포인트(AP)는

수신안테나와

송신안테나를 포함할 수 있다. 엑세스 포인트(AP)는 수신된 업링크 신호와 다운링크에 대한 프리코딩 데이터를 감지하기 위해 업링크 파일럿 시그널링을 통해 CSI(channel state information)를 획득할 수 있다. 엑세스 포인트(AP)는 채널 상호성을 활용하여 다운링크에 데이터를 프리코드할 수 있다. 엑세스 포인트(AP)는 coherence 구간(

) 중

구간에는 CSI를 획득하고, 남은

구간에는 데이터를 전송할 수 있다. In the cell-free massive multiple input (MIMO) transmission device considered in the present invention, M full-duplex access points (APs) may be connected to a central processing unit (CPU) through an error-free backhaul link. In addition, a cell-free massive multiple input multiple output (MIMO) transmission device is composed of a single antenna providing the same phase in the same time-frequency resource.

dog uplink

Can include downlinks from. Each access point (AP)

receiving antenna

It may include a transmitting antenna. An access point (AP) may obtain channel state information (CSI) through uplink pilot signaling in order to detect a received uplink signal and downlink precoded data. An access point (AP) can utilize channel reciprocity to precode data on the downlink. The access point (AP) is a coherence section (

) middle

Acquire CSI in the section, and

Data can be transmitted in the interval.

Each access point (AP) is self-interference (SI) in which the received uplink signal is damaged by the downlink signal transmitted by another access point (AP) due to the full-duplex method, and interference between access points (inter-access point) -AP interference, IAI) may be affected. Downlink may be affected by uplink-downlink interference (UL-to-DL interference, UDI) due to uplink signal transmission. Access points (APs) and downlinks can be affected by quantization noise (QN) in the process of using low-resolution ADCs for quantization. A method for optimizing the resolution of a cell-free massive multiple input (MIMO) transmission device according to the present invention includes modeling the interference and noise as variables for the resolution (b) in order to derive the resolution of a low resolution ADC having high efficiency. can do.

2 shows a flowchart for deriving an optimal resolution of an ADC used to receive a signal transmitted in an uplink in an access point (AP) according to an embodiment of the present invention. Referring to FIG. 2, the resolution optimization method of the cell-free massive multiple input/output transmission device may include an uplink channel setting step (S210), an uplink spectral efficiency setting step (S220), and an uplink resolution derivation step (S230). can

In the uplink channel setting step (S210), the i-th uplink may define a channel for the m-th access point (AP) according to [Relationship Equation 1].

[Relationship 1]

(여기서, β는 large-scale의 레일리 페이딩 계수이고,

는 small-scale 페이딩 벡터이다.)

(Where β is the large-scale Rayleigh fading coefficient,

is a small-scale fading vector.)

을 만족하고 각 원소들은

로써 모델링되는

를 정의할 수 있다. 업링크 채널 설정 단계(S210)는 small-scale 페이딩이 coherence 구간(

) 동안 변경 없이 유지되며, 각 coherence 구간(

)은 독립적으로 변경되는 것으로 채널을 정의할 수 있다. The uplink channel setting step (S210) is

and each element is

modeled as

can define In the uplink channel setting step (S210), small-scale fading is a coherence section (

) and remains unchanged during each coherence interval (

) can define a channel as being changed independently.

In the uplink spectral efficiency setting step (S220), the spectral efficiency (SE) of the i-th uplink may be defined as a closed solution obtained by substituting [Equation 1] into [Equation 2]. [Equation 1] and [Equation 2] used in the uplink spectral efficiency setting step (S220) assume that the CPU detects a signal using long-term channel statistics.

[Equation 1]

(여기서,

는 coherence 구간,

는 남은 구간,

는 업링크 전송 전력, M은 엑세스 포인트의 수,

(

는 ADC 해상도),

는 빔포밍 불확실성 이득,

는 다중 사용자 간섭,

는 SI/IAI(self-interference/inter-AP interference),

는 노이즈,

는 양자화잡음(quantization noise, QN)을 의미한다.)

(here,

is the coherence interval,

is the remaining interval,

is the uplink transmit power, M is the number of access points,

(

is the ADC resolution),

is the beamforming uncertainty gain,

is multi-user interference,

Is SI/IAI (self-interference/inter-AP interference),

is the noise,

means quantization noise (QN).)

[Equation 2]

(여기서,

는 수신 안테나 수,

는 업링크가 전송한 파일럿의 전력,

는 파일럿 길이,

는 large-scale의 레일리 페이딩 계수,

는 채널추정 관련 상수,

는 다중 사용자 간섭과 파일럿 오염(PC)의 효과를 반영한 변수,

는 잔여 SI와 IAI의 영향을 반영한 변수,

는 양자화 잡음(QN)을 의미한다.)

(here,

is the number of receiving antennas,

is the power of the pilot transmitted by the uplink,

is the pilot length,

is the large-scale Rayleigh fading coefficient,

is a channel estimation related constant,

is a variable reflecting the effects of multiuser interference and pilot contamination (PC),

is a variable reflecting the influence of residual SI and IAI,

stands for quantization noise (QN).)

는 i번째의 업링크의 스펙트럼 효율(SE) 하한을 의미한다. [수학식 1]에서 빔포밍 불확실성 이득인

는

로 계산될 수 있다. [수학식 1]에서 다중 사용자 간섭인

는

로 계산될 수 있다. [수학식 1]에서 SI/IAI인

는

로 계산될 수 있다. [수학식 1]에서 노이즈인

는

로 계산될 수 있다. [수학식 1]에서 양자화 잡음인

은

로 계산될 수 있다. [Equation 1] used in the uplink spectral efficiency setting step (S220) considers that the effective noise has no correlation with the desired signal and the added Gaussian noise is the worst uncorrelated noise. [Equation 1]

Means the lower limit of the spectral efficiency (SE) of the i-th uplink. In [Equation 1], the beamforming uncertainty gain

Is

can be calculated as In [Equation 1], the multi-user interferer

Is

can be calculated as SI / IAI in [Equation 1]

Is

can be calculated as Noise in [Equation 1]

Is

can be calculated as In [Equation 1], the quantization noise

silver

can be calculated as

는 저해상도 ADC를 구비한 전이중 셀-프리 MIMO의 i번째 업링크의 SINR(signal-to-interference plus noise ratio)을 의미한다. [수학식 2]에서 다중 사용자 간섭과 파일럿 오염의 효과를 반영한 변수인

는

로 계산될 수 있다. [수학식 2]에서 잔여 SI와 IAI의 영향을 반영한 변수인

는

로 계산될 수 있다. [수학식 2]에서 양자화 잡음인

은

로 계산될 수 있다. [Equation 2] used in the uplink spectral efficiency setting step (S220)

Means signal-to-interference plus noise ratio (SINR) of the ith uplink of full-duplex cell-free MIMO with low-resolution ADC. In [Equation 2], a variable that reflects the effects of multi-user interference and pilot contamination

Is

can be calculated as In [Equation 2], a variable that reflects the influence of residual SI and IAI

Is

can be calculated as In [Equation 2], the quantization noise

silver

can be calculated as

In the uplink resolution derivation step (S230), the maximum energy efficiency is determined by using the uplink spectral efficiency defined in the uplink spectral efficiency setting step (S220) and the energy efficiency (EE) defined by [Equation 3]. resolution can be extracted.

[Equation 3]

(여기서,

는 대역폭,

는 링크의 수,

는 스펙트럼효율,

는 소모전력을 의미한다.)

(here,

is the bandwidth,

is the number of links,

is the spectral efficiency,

means power consumption.)

In the uplink resolution derivation step (S230), the resolution having the maximum energy efficiency may be extracted by using the resolution (b) as a variable in [Equation 3]. In the uplink resolution derivation step ( S230 ), loss due to quantization noise may be compensated for by increasing the number of receiving arrays. However, as the number of receiving arrays increases, more ADCs are required and power consumption deteriorates. Therefore, it is preferable to use only resolution as a variable.

를 의미한다. 따라서, [수학식 3]은 다운링크 해상도 도출에도 사용될 수 있다. [수학식 3]에서 업링크에서의 소모전력인

(단, a=u)는

로 계산될 수 있다. 여기서,

는 저잡음 증폭기의 소모전력,

는 믹서의 증폭기의 소모전력,

는 중간 주파수 증폭기의 소모전력,

는 능동 필터의 소모전력,

는 주파수 합성기의 소모전력,

는 자동 이득 제어기의 소모전력,

는 ADC의 소모전력을 의미한다. a in [Equation 3] used in the uplink resolution derivation step (S230) is

means Therefore, [Equation 3] can also be used to derive downlink resolution. In [Equation 3], the power consumption in the uplink is

(However, a = u)

can be calculated as here,

is the power consumption of the low-noise amplifier,

is the power consumption of the amplifier of the mixer,

is the power consumption of the intermediate frequency amplifier,

is the power consumption of the active filter,

is the power consumption of the frequency synthesizer,

is the power consumption of the automatic gain controller,

is the power consumption of the ADC.

는

로 계산될 수 있다(단,

은 1일 때, c=0 또는

일 때, c=1). 여기서,

는 컨버터의 공급 전원,

는 코너 주파수,

는 주어진 CMOS에 대한 최소 채널 길이를 의미한다.

Is

It can be calculated as (however,

is 1, c=0 or

When c=1). here,

is the supply power of the converter,

is the corner frequency,

means the minimum channel length for a given CMOS.

FIG. 3 shows a flowchart for deriving an optimal resolution of an ADC used to receive a signal transmitted from an access point (AP) in a downlink according to an embodiment of the present invention. Referring to FIG. 3, the resolution optimization method of the cell-free massive multiple input/output transmission device may include a downlink channel setting step (S310), a downlink spectral efficiency setting step (S320), and a downlink resolution derivation step (S330). can

In the downlink channel setting step (S310), the k-th downlink channel may be defined by [Relationship 2] for the m-th access point (AP).

[Relationship 2]

(여기서, ζ는 large-scale의 레일리 페이딩 계수이고,

는 small-scale 페이딩 벡터이다.)

(Where ζ is the large-scale Rayleigh fading coefficient,

is a small-scale fading vector.)

을 만족하고 각 원소들은

로써 모델링되는

를 정의할 수 있다. 다운링크 채널 설정 단계(S310)는 small-scale 페이딩이 coherence 구간(

) 동안 변경 없이 유지되며, 각 coherence 구간(

)은 독립적으로 변경되는 것으로 채널을 정의할 수 있다. Downlink channel setting step (S310)

and each element is

modeled as

can define In the downlink channel setting step (S310), small-scale fading is a coherence section (

) and remains unchanged during each coherence interval (

) can define a channel as being changed independently.

In the downlink spectral efficiency setting step (S320), the spectral efficiency (SE) of the k-th downlink may be defined as a closed solution obtained by substituting [Equation 4] into [Equation 5]. [Equation 1] and [Equation 2] used in the downlink spectral efficiency setting step (S320) assume that instantaneous CSI is not used to decode a signal in the downlink.

[Equation 4]

(여기서,

는 ADC 해상도,

는 다운링크 전송 전력, M은 업링크의 수,

는 정규화 요소,

는 빔포밍 불확실성 이득,

는 다중 사용자 간섭,

는 업링크-다운링크 간섭,

는 양자화잡음(QN)을 의미한다. 여기서

는 수신신호이다.)

(here,

is the ADC resolution,

is the downlink transmit power, M is the number of uplinks,

is the regularization factor,

is the beamforming uncertainty gain,

is multi-user interference,

is the uplink-downlink interference,

denotes quantization noise (QN). here

is the received signal.)

[Equation 5]

(여기서,

는 송신 안테나 수,

(

는 ADC 해상도),

은 다중 사용자 간섭과 파일럿 오염의 효과를 반영한 변수,

는 양자화 잡음,

(

는 n번째 AP에서 m번째 AP로의 large-scale 페이딩,

는 잔여간섭)을 의미한다.)

(here,

is the number of transmit antennas,

(

is the ADC resolution),

is a variable reflecting the effects of multi-user interference and pilot contamination,

is the quantization noise,

(

is large-scale fading from the n-th AP to the m-th AP,

is the residual interference).)

는 k번째의 다운링크의 성취 가능한 스펙트럼 효율(SE)을 의미한다. [수학식 4]에서 빔포밍 불확실성 이득인

는

로 계산된다. [수학식 4]에서 빔포밍 불확실성 이득인

는

로 계산된다. [수학식 4]에서 다중 사용자 간섭인

는

로 계산된다. [수학식 4]에서 업링크-다운링크 간섭인

는

로 계산된다. [Equation 4] used in the downlink spectral efficiency setting step (S320) considers that the effective noise has no correlation with the desired signal and the added Gaussian noise is the worst uncorrelated noise. [Equation 4]

denotes the achievable spectral efficiency (SE) of the k-th downlink. In [Equation 4], the beamforming uncertainty gain

Is

is calculated as In [Equation 4], the beamforming uncertainty gain

Is

is calculated as In [Equation 4], the multi-user interferer

Is

is calculated as In [Equation 4], the uplink-downlink interferer

Is

is calculated as

는 MRT 프리코딩 및 저해상도 ADC를 구비한 전이중 셀-프리 MIMO의 k번째 다운링크의 SINR(signal-to-interference plus noise ratio)을 의미한다. [수학식 5]에서 다중 사용자 간섭인

는

로 계산될 수 있다. [수학식 5]에서 양자화 잡음인

는

로 계산될 수 있다. [Equation 5] used in the downlink spectral efficiency setting step (S320)

Means signal-to-interference plus noise ratio (SINR) of the k-th downlink of full-duplex cell-free MIMO with MRT precoding and low-resolution ADC. In [Equation 5], the multi-user interferer

Is

can be calculated as In [Equation 5], the quantization noise

Is

can be calculated as

In the downlink resolution derivation step (S330), the maximum energy efficiency is determined by using the downlink spectral efficiency defined in the downlink spectral efficiency setting step (S320) and energy efficiency (EE) defined by [Equation 3]. resolution can be extracted.

In the downlink resolution derivation step (S330), as shown in [Equation 5], since the achievable signal and quantization noise increase at the same rate as the number of access points (M), the number of access points is increased to compensate for the quantization noise Doing so is not efficient. Therefore, in the downlink resolution deriving step (S330), it is preferable to extract the maximum energy efficiency using the resolution of the ADC as a variable.

(단, a=d)는

로 계산될 수 있다.In [Equation 3], the power consumption in the downlink is

(However, a = d)

can be calculated as

In another embodiment of the present invention, a resolution optimization system of a cell-free massive multiple input/output transmission device may include an uplink channel setting unit, an uplink spectrum efficiency setting unit, and an uplink resolution derivation unit.

The uplink channel setting unit may define a channel for the i-th uplink to the m-th access point (AP) according to [Relationship Equation 1]. The uplink channel setting unit performs the function of the above-described uplink channel setting step.

The uplink spectral efficiency setting unit may define spectral efficiency (SE) of the i-th uplink as a closed solution obtained by substituting [Equation 1] into [Equation 2]. The uplink spectral efficiency setting unit performs the function of the above-described uplink spectral efficiency setting step.

The uplink resolution derivation unit may extract a resolution having maximum energy efficiency using the uplink spectrum efficiency defined in the uplink SE setting step and energy efficiency (EE) defined by [Equation 3]. The uplink resolution derivation unit performs the function of the above-described uplink resolution derivation step.

Hereinafter, an exemplary embodiment of a method for optimizing the resolution of a cell-free Massive Multiple Input Output (MIMO) transmission device according to the present invention will be described in detail with reference to the accompanying drawings.

개 채널 구현에 대한 Monte Carlo 시뮬레이션에 의한 분석 결과이다. 본 실시예에서는 파라미터값으로

,

,

,

,

를 사용하였다. 파일럿 시퀀스는 각 사용자에게 무작위로 할당되는 것으로 설정하였다. ADC는 단순화를 위해 모든 AP 및 DL 사용자에게 동일한 해상도를 사용하였다.

,

,

이고, 여기서

는 업링크 전력,

는 다운링크 전력,

는 파일럿 전력을 의미한다.

,

,

,

,

의 값을 사용하였다. 엑세스 포인트와 사용자는

의 영역에 균일하게 분포되어 있는 것으로 설정하였다. 임의의 두 노드 사이의 large-scale 페이딩은

으로 설정된다. 여기서

는 표준편차가

,

인 섀도우 페이딩을 의미한다.

는 경로손실을 의미한다. 또한,

의 값을 사용하였다. In this embodiment

This is the analysis result by Monte Carlo simulation for the implementation of two channels. In this embodiment, as a parameter value

,

,

,

,

was used. The pilot sequence was set to be randomly assigned to each user. The ADC uses the same resolution for all AP and DL users for simplicity.

,

,

and here

is the uplink power,

is the downlink power,

denotes the pilot power.

,

,

,

,

The value of was used. access points and users

It was set to be uniformly distributed in the area of . The large-scale fading between any two nodes is

is set to here

is the standard deviation

,

Indicates in-shadow fading.

is the path loss. also,

The value of was used.

4 is a graph illustrating a change in sum spectral efficiency (SE) with respect to the number (M) of access points (APs) according to an embodiment of the present invention. Referring to FIG. 4 , in the case of uplink, spectrum efficiency (SE) increases monotonically as the number (M) of access points (APs) increases and decreases as the resolution b of the ADC decreases. Therefore, speed loss due to quantization noise in the uplink can be compensated for by increasing the number of access points. For example, to achieve an uplink spectral efficiency of 30 bps/Hz, approximately 23 access points are required for infinite resolution. On the other hand, for resolutions of 3, 2, or 1, 25, 28, or 42 access points are required to obtain a spectral efficiency similar to 30 bps/Hz. It can be seen that the sum spectral efficiency of the downlink rapidly saturates as the number of access points increases at a low resolution, unlike the uplink.

5 is a graph showing a change in sum spectral efficiency (SE) with respect to resolution (b) of an ADC according to an embodiment of the present invention. Referring to Figure 5, the spectral efficiency of both downlink and uplink increases monotonically with increasing resolution (b) until convergence. For the uplink, about 5 bits of ADC resolution are required to converge. For the downlink, about 7 bits of ADC resolution are required to converge. The uplink shows superior growth compared to the downlink in the lower resolution region (i.e. b = 1, 2), but the superiority decreases in the resolution region b>2. In the downlink, quantization noise in the low resolution region is more dominant than in the uplink. The downlink has a higher spectral efficiency than the uplink because the effect of quantization noise is reduced when b>2.

6 is a graph illustrating a change in sum spectral efficiency (SE) with respect to residual self-interference (SI) according to an embodiment of the present invention. Referring to FIG. 6 , it can be seen that the spectral efficiency of the full-duplex method decreases as residual self-interference (SI) increases. The spectral efficiency of the half-duplex scheme remains fixed because it is not affected by self-interference (SI), inter-access point interference (IAI), and uplink-downlink interference (UDI).

영역에서 반이중 방식을 능가한다. 반면, b=2일 때, 전이중 방식은

영역에서 반이중 방식을 능가한다. 이는 해상도(b)가 감소함에 따라

가 감소하여 양자화 잡음에 대한 잔여 자기간섭(SI), 엑세스 포인트간 간섭(IAI), 및 업링크-다운링크 간섭(UDI)의 영향을 증가시키는 [수학식 2]과 [수학식 4]로부터 알 수 있다. 또한, 전이중 방식은

에서 반이중 방식 보다 약 98.82% 더 높은 스펙트럼 효율을 달성한다. The sum spectral efficiency (SE) generally improves as the resolution (b) increases. In full-duplex cell-free MIMO, as the resolution (b) increases, the full-duplex resistance to residual self-interference (SI) increases, so the operating range increases. In one embodiment, when b = 1, the full-duplex scheme is

It outperforms half-duplex in the area. On the other hand, when b = 2, full-duplex

It outperforms half-duplex in the area. As the resolution (b) decreases

It can be seen from [Equation 2] and [Equation 4] that increases the effects of residual self-interference (SI), inter-access point interference (IAI), and uplink-downlink interference (UDI) on quantization noise by decreasing can Also, full-duplex

achieves about 98.82% higher spectral efficiency than half-duplex in

7 is a graph of energy efficiency (EE) of an uplink, and FIG. 7 (a) shows a graph showing a change in energy efficiency (EE) with respect to resolution (b) of an ADC. 7(b) shows a graph showing the change in energy efficiency (EE) for sum spectral efficiency (SE).

Referring to (a) of FIG. 7 , while the resolution increases from 1 to 2, the energy efficiency (EE) increases linearly with the resolution. On the other hand, energy efficiency (EE) converges to zero when the resolution is greater than 2, as the power consumption of the ADC increases exponentially. Using a simple line search algorithm such as the bisection method, an optimal resolution (b) that maximizes energy efficiency (EE) can be derived. In the uplink resolution derivation step (S230), a function of deriving an optimal resolution maximizing energy efficiency (EE) can be performed.

에 따라 선형적으로 증가한다. 반면에, 스펙트럼 효율(SE)은 M이 증가함에 따라 대수 비율로 증가한다. 또한, 에너지 효율(EE)은

가 40dB에서 100dB로 증가함에 따라 감소한다. 전이중 방식은 반이중 방식보다 훨씬 더 높은 에너지 효율(EE)을 제공함 역시 알 수 있다. Energy efficiency (EE) typically decreases as M increases, as more ADCs are required as the RF chain increases. Thus, the total power consumption depends on the number of access points (M) and the receiving array

increases linearly with On the other hand, the spectral efficiency (SE) increases in a logarithmic ratio as M increases. Also, energy efficiency (EE) is

decreases as β increases from 40 dB to 100 dB. It can also be seen that full-duplex provides much higher energy efficiency (EE) than half-duplex.

Referring to (b) of FIG. 7, the best sum spectral efficiency (SE) is displayed at the rightmost point. On the other hand, the best energy efficiency (EE) is displayed at the topmost point. Thus, it can be seen that the best EE/SE compromise is achieved at the upper right corner. As shown, energy efficiency (EE) and spectral efficiency (SE) improve as the resolution (b) increases from 1 to 2 bits. However, when b>2, the energy efficiency (EE) decreases while the spectral efficiency (SE) is saturated. Also, as M increases, the spectral efficiency (SE) improves overall and the energy efficiency (EE) decreases.

가 증가함에 따라 감소한다. 따라서, 전이중 방식의 시스템의 유연성이 감소한다. 전이중 방식의 작동 영역은 1/2 pre-log factor로 인해 반이중 방식 보다 훨씬 넓다. 따라서, 반이중 방식 보다 전이중 셀-프리 MIMO에서 저해상도 ADC를 사용하는 것이 에너지 효율 측면에서 더 유리하다. The envelope of the full-duplex EE/SE operating range is

decreases as increases. Accordingly, the flexibility of the full-duplex system is reduced. The operating range of full-duplex is much wider than that of half-duplex due to the 1/2 pre-log factor. Therefore, it is more advantageous in terms of energy efficiency to use a low-resolution ADC in full-duplex cell-free MIMO than in half-duplex mode.

8 is a graph of downlink energy efficiency (EE), and FIG. 8 (a) shows a graph showing a change in energy efficiency (EE) with respect to ADC resolution (b). 8(b) shows a graph showing the change in energy efficiency (EE) for sum spectral efficiency (SE).

의 증가에 따른 영향이 거의 없다. Referring to (a) of FIG. 8 , energy efficiency (EE) increases monotonically as the resolution (b) increases from 1 to 4. Energy efficiency (EE) decreases at b>4. Energy efficiency (EE) of downlink increases as M increases because, in contrast to uplink, spectrum efficiency (SE) improves when the number of access points (M) increases. The power consumption of a downlink user's ADC is independent of the number of access points (M). Downlink energy efficiency (EE) is doubled in full-duplex compared to half-duplex. In addition, the energy efficiency (EE) of the downlink is

has little effect on the increase in

Referring to (b) of FIG. 8 , it can be seen that energy efficiency (EE) and spectral efficiency (SE) increase as the resolution (b) increases from 1 to 4 arcs. On the other hand, when b>4, energy efficiency (EE) decreases as spectral efficiency (SE) remains stagnant.

Although the present invention has been described in detail through representative embodiments, those skilled in the art will understand that various modifications are possible to the above-described embodiments without departing from the scope of the present invention. will be. Therefore, the scope of the present invention should not be limited to the described embodiments and should not be defined, and should be defined by all changes or modifications derived from the claims and equivalent concepts as well as the claims to be described later.

Claims (5)

an uplink channel setting step in which the i-th uplink defines a channel with respect to the m-th access point (AP) by [Relationship Equation 1];
an uplink spectral efficiency setting step of defining spectral efficiency (SE) of the i-th uplink as a closed-type solution obtained by substituting [Equation 1] into [Equation 2]; and
an uplink resolution derivation step of extracting a resolution having maximum energy efficiency using the uplink spectral efficiency defined in the uplink spectral efficiency setting step and energy efficiency (EE) defined by [Equation 3]; Resolution optimization method of a cell-free massive multiple input/output transmission device comprising a.
[Relationship 1]

(Where β is the large-scale Rayleigh fading coefficient,

is a small-scale fading vector.)

[Equation 1]

(here,

is the coherence interval,

is the remaining interval,

is the uplink transmit power, M is the number of access points,

(

is the ADC resolution),

is the beamforming uncertainty gain,

is multi-user interference,

Is SI/IAI (self-interference/inter-AP interference),

is the noise,

means quantization noise (QN).)

[Equation 2]

(here,

is the number of receiving antennas,

is the power of the pilot transmitted by the uplink,

is the pilot length,

is the large-scale Rayleigh fading coefficient,

is a channel estimation related constant,

is a variable reflecting the effects of multiuser interference and pilot contamination (PC),

is a variable reflecting the influence of residual SI and IAI,

stands for quantization noise (QN).)
According to claim 1,
Resolution optimization of a cell-free massive multiple input/output transmission device further comprising; a downlink channel setting step in which the k-th downlink defines a channel for the m-th access point (AP) by [Relationship 2] method.
[Relationship 2]

(Where ζ is the large-scale Rayleigh fading coefficient,

is a small-scale fading vector.)
According to claim 2,
Further comprising a downlink spectral efficiency setting step in which the spectral efficiency (SE) of the k-th downlink is defined as a closed solution obtained by substituting [Equation 4] into [Equation 5]. Method for optimizing the resolution of a cell-free massive multiple input/output transmission device.
[Equation 4]

(here,

is the ADC resolution,

is the downlink transmit power, M is the number of uplinks,

is the regularization factor,

is the beamforming uncertainty gain,

is multi-user interference,

is the uplink-downlink interference,

denotes quantization noise (QN). here

is the received signal.)

[Equation 5]

(here,

is the number of transmit antennas,

(

is the ADC resolution),

is a variable reflecting the effects of multi-user interference and pilot contamination,

is the quantization noise,

(

is large-scale fading from the n-th AP to the m-th AP,

is the residual interference).)
According to claim 3,
a downlink resolution derivation step of extracting a resolution having maximum energy efficiency using the downlink spectral efficiency defined in the downlink spectral efficiency setting step and energy efficiency (EE) defined by [Equation 3]; A method for optimizing the resolution of a cell-free massive multiple input/output transmission device, characterized in that it further comprises.
an uplink channel setting unit for defining a channel for an i-th uplink with respect to an m-th access point (AP) by [Relational Expression 1];
an uplink spectral efficiency setting unit defining spectral efficiency (SE) of the i-th uplink as a closed solution obtained by substituting [Equation 1] into [Equation 2]; and
An uplink resolution derivation unit for extracting a resolution having maximum energy efficiency using the uplink spectral efficiency defined by the uplink spectral efficiency setting unit and energy efficiency (EE) defined by [Equation 3]; A system for optimizing the resolution of a cell-free massive multi-output transmission device comprising:

Images