TWI624168B - An intelligent deployment cascade control device based on an fdd-ofdma indoor small cell in multi-user and interference environments - Google Patents

Abstract

The present invention discloses a frequency division duplexing (FDD)-orthogonal frequency division multiplexing access (OFDMA) indoor small cell base station intelligent deployment serial control (intelligent deployment) Cascade control, IDCC) method and device, which enables an indoor small cell to meet multi-user (MU) service reliability, optimized access rate, and cell radius demand, and self-optimized emission minimum in response to environmental changes. Power, easy to install, save energy, and avoid interference with other users outside. The intelligent deployment control device is constructed in an adaptive network-based fuzzy inference system (ANFIS), including a resource configurator, a minimum access rate/edge CQI converter, and an initial transmit power configuration (initial transmit power setting). Control, ITPSC) (ANFIS controller #1), optimized channel quality index (CQI) decision maker (CQIDC) (ANFIS controller #2) and self-optimized power control (SOPC) five units. Wherein, the self-optimizing power control unit includes a transmit power adjustment estimator (transmit power adjustment) Estimator, TPAE) (ANFIS controller #3), power adjustment configurator and self-optimized power control protection mechanism. The service reliability, the indoor coverage radius and the minimum access rate setting are input by the user equipment (UE), and the indoor small cell base deployment is self-optimized in the multi-user interference environment. Orthogonal frequency division multiplexing access-divided full-duplex indoor small cell base station self-optimized deployment control device by user equipment (UE) backhaul reference signal received power (Reference Signal Received Power, RSRP) information, estimate the path loss (PL) and the signal-to-interference-plus-noise ratio (SINR) of the inter-connector, and adaptively control the transmit power of the transceiver The user of the indoor small cell base station can self-optimize to produce the performance of the transmitter that satisfies the multi-user service reliability setting, the optimized access rate, the minimum transmission power and the interference.

The embodiment assumes that after three users are configured by resources, the number of resource blocks is 33, and the minimum access rate requirements are 2.76 Mbps, 7.44 Mbps, and 14.13 Mbps, respectively. The simulation results of the embodiment show that when the invention is deployed in the FDD-OFDMA indoor small cell base station, the indoor small cell base station can have the advantages of high service reliability, low power consumption and low co-channel interference power to neighboring cell users.

Description

Built in a multi-user interference environment built in a frequency division full-duplex-orthogonal frequency division multiplexing Cell intelligent deployment cascade control device

The invention relates to a frequency division full-duplex-orthogonal frequency division multiplexing accessing indoor small cell base station intelligent deployment string constructed in an adaptive network-based fuzzy inference system (ANFIS) Incentive deployment cascade control (IDCC) software and devices, especially a self-optimizing deployment controller that can control the multi-user (MU) service reliability, transmission power and interference of small cell transfer machines .

With the rapid development of wireless communication requirements and applications, the flow of social multimedia messages has increased rapidly, and the requirements for links with high transmission rates and high reliability have increased day by day, especially for multimedia communication and global seamless communication coverage. The mainstream status of wireless broadband communication systems has been established, and the demand for spectrum resources in broadband communication systems has also increased. As a result, spectrum resources have become increasingly crowded and useless, and it is likely to become a bottleneck restricting the development of wireless broadband communication. The short-range femtocell has a small service area and has the advantages of low cost, low power consumption, high spectral efficiency and high capacity compared with the microcell system. It is a very marketable wireless network technology. [Note 1]. [Note 2] The research report predicts that the global small cell market will grow from 690 million US dollars in 2014 to 4.8 billion US dollars in 2019, with an average annual growth rate of 41.7. %. Therefore, short-range nanocell technology will play a very important role in the development and application of next-generation wireless communication networks.

[Note 1] Jie Zhang, Guillaume de la Roche, et al. Femtocells - Technologies and Deployment, Wiley, Jan. 2010. [Note 2] Small Cells Market and Femtocell Industry 2019 Forecasts in New Research Reports, DALLAS, October 7, 2014 /PRNewswire/.

The number of nanocell base stations is much larger than that of large cells. In order to save costs, the general cell deployment method will not be applicable to the deployment of nanocell base stations. Therefore, the microcell base station must be deployed by the user, and an intelligent deployment cascade control (IDCC) software and device must be designed so that the nano cell base station can be assisted with minimal human assistance. Only need plug-and-play, you can automatically configure system parameters, self-optimize the base station transmit power in the interference environment, achieve the goal of saving energy, reducing co-channel interference, and can meet multi-user service reliability. Corresponding to the user's input setting of the minimum access rate of the cell edge CQI of the indoor small cell transfer machine and the cell radius coverage of the matching room size. The present invention mainly considers a frequency division duplexing-orthogonal frequency division multiple access (FDD-OFDMA) nanocell base station in a single indoor interference environment, and uniformly distributes Indoor multi-user, designed self-optimized base station deployment device built on adaptive network fuzzy inference system.

[Note 3] A method for deploying adaptive coverage of nanocells is proposed to reduce the increase in mobility signals of the core network. The nanocell base station uses self-optimized deployment of nanocell coverage using data from large cell users and indoor users moving through the cell-cell coverage area.

[Note 4] A self-optimized nanocell network base station group coverage method based on genetic algorithm is proposed, based on traffic statistics and interference level between adjacent base stations of a nanocell network. The centralized dynamic update of the downlink pilot power of each base station in the nanocell network effectively optimizes the coverage of the deployment of the nanocell network base station group in the enterprise environment.

[Note 3] Holger Claussen et al., Self-optimization of Coverage for Femtocell Deployments, Bell Labs Technical Journal - Core and Wireless Networks, Volume 14 Issue 2, August 2009, Pages 155-183.

[Note 4] Lina S. Mohjazi et al., Self-Optimization of Pilot Power in Enterprise Femtocells Using Multi Objective Heuristic, Journal of Computer Networks and Communications, Volume2012.

General wireless multimedia communication network spectrum and resource management, mainly including channel and transmit power assignment. [Note 5] The application of the ANFIS method to control the power of the direct sequence code division multiplex multimedia communication system can estimate and compensate the influence of the attenuation channel, thereby achieving the goal of high successful transmission rate and high data transmission rate. The two input variables of the ANFIS power control mechanism are signal-to-interference-plus-noise ratio (SINR) error( e ) and SINR error change(△ e ), respectively, and each uses 7 Gaussian functions. To blur the attribution function, there are 49 fuzzy inference rules. The ANFIS power control mechanism adjusts the signal-to-interference ratio setpoint and uses two input variables, SINR error e(n) and SINR error change Δe(n), to track the SINR setpoint. The fixed value of the target SINR is set to 1.5 dB, which makes the power control process less flexible, and the ANFIS input parameter only depends on the SINR control, and cannot effectively coordinate with the current channel environment for power control. The technology also does not consider network multi-user (MU) service and service reliability (SR) performance.

[Note 5] C. H. Jiang, J. K. Lian, R. M. Weng, C. H. Hsu, "Multi-rate DS-CDMA with ANFIS-assisted power control for wireless multi-media communications," International Journal of Innovative Computing, Information and Control, vol. 6, no. 8, pp. 3641-3655, Aug. 2010.

The present invention is a multi-user (MU) frequency division full-duplex-orthogonal frequency division multiplexing multiplexed indoor small cell constructed in an adaptive network-based fuzzy inference system (ANFIS) Base station intelligent deployment cascade control (IDCC) software and devices, especially a self-optimizing deployment controller that can control the multi-user service reliability, transmission power and interference of small cell transmission machines. The main purpose of the present invention is that the user can input the service reliability, the indoor coverage radius and the minimum access rate requirement (corresponding to the cell edge CQI 1~15) and the user equipment (UE) backhaul orthogonal frequency division. Path frequency (PS) and signal-to-interference-plus-noise ratio (SINR) information, adaptive control transfer The transmit power of the machine enables users in the small cell mobile communication network to meet service reliability, minimum transmit power and interference requirements.

The invention adopts an adaptive network-based fuzzy inference system (ANFIS) method to design a serial control structure of a small cell intelligent deployment control device, including a resource configurator and a minimum access rate. /Edge CQI converter, initial transmit power setting control (ITPSC) (ANFIS controller #1), optimized channel quality index (CQI) decision maker (CQIDC) (ANFIS controller# 2) Self-optimizing power control (SOPC) five units. The self-optimizing power control unit includes a transmit power adjustment estimator (TPAE) (ANFIS controller #3), a power adjustment configurator and a self-optimizing power control protection mechanism. The user inputs the service reliability, the indoor coverage radius and the minimum access rate setting, and the self-optimized deployment control device estimates the path loss between the transmission and the transmission machine by using the information returned by the user terminal, and adaptive control transmission The transmitting power of the pick-up allows the indoor small cell base station self-optimized deployment control device to meet the service reliability, optimized access rate, minimum transmit power and interference transfer performance.

The resource configuration unit of the present invention configures the resource block number of the small base station on average according to the number of indoor users and the set system bandwidth, and the minimum access rate/cell edge CQI conversion unit according to the minimum (cell edge) access rate requirement set by each user in the room. The corresponding indoor small cell transfer cell edge CQI is generated, and the interface cell edge CQI corresponds to a receiver block error rate (BLER) equal to a signal-to-interference ratio (SINR _th ) of 0.1.

For the initial transmit power controller (ANFIS #1), each indoor user has three input parameters: indoor coverage radius (R _u ), resource block number (nRB _u ), and cell edge CQI (CQI _{min, u} ), each input parameter uses a three-shaped (Generalized bell-shaped) membership function (MF), defined as a radius of 0m~15m, divided into three levels; the number of defined resource blocks is 1~100, divided into 3 levels; define the cell edge CQI as 1~15, which is divided into 3 levels, and there are 27 fuzzy inference rules. We define the coverage radius to be less than 5 meters for L, 5 meters to 10 meters for M, 10 meters to 15 meters for H, and the number of resource blocks from 0 to 25 for L and 26 to 74 for M. 75~100 are H, cell edge CQI demand is 1~5 is L, 6~10 is M, 11~15 is H, and its output is the initial minimum transmit power corresponding to the uth user ( _{Pini, u} )value. The objectives and conditions for optimizing the initial transmit power controller of the u-th user ANFIS can be expressed as:

For the optimized channel quality indicator decision maker (ANFIS #2), each indoor user has three input parameters: the average path loss between the transceivers. The initial transmit power setting ( P _{ini , u} ) and the number of resource blocks ( nRB _u ), each of which uses three Gaussian membership functions. In the optimized CQI decision controller unit, the average path loss is defined as 30dB~70dB, which is divided into three levels; the initial transmit power is -75dBm~20dBm, which is divided into three levels; the number of resource blocks is 1. ~100, divided into 3 levels, a total of 27 fuzzy inference rules. The optimized CQI decision controller unit generates the best CQI target in the interference-free environment according to the formula (2), and provides a self-optimized control power controller (SOPC) unit to perform power adjustment in the interference environment.

We define the average path loss to be 30dB~40dB for L, 41dB~60dB for M, 61dB~70dB for H; the initial transmit power is -75dBm~-51dBm for L, -50dBm ~-4dBm for M, -3dBm~20dBm for H The number of resource blocks is 0 to 25 for L, 26 to 74 for M, and 75 to 100 for H, and the output is the best CQI (CQI _{best, u} ) corresponding to the uth user in a non-interference environment. The optimization goals and conditions corresponding to the u-th user-optimized CQI decision controller unit can be expressed as:

1. For the Transmit Power Adjustment Estimator (TPAE) (ANFIS #3), each indoor user has three input parameters: the optimal CQI and SINR measurement average for cell edge CQI requirements and interference free environment. Value, each input parameter uses 3 bell-shaped attribution function, defines the cell edge CQI demand as 1~15, which is divided into three levels; the best CQI is 1~15, which is divided into three levels; SINR measurement The average value is -25dB~45dB, which is divided into three levels, a total of 27 fuzzy inference rules. We define cell edge CQI requirements as 1 to 5 for L, 6 to 10 for M, and 11 to 15 for H. The best CQI for undisturbed environments is L at 1 to 5, M at 10 to 10, and 11 to 15 at The average value of H and SINR is L at -25dB~-5dB, M is -5dB~25dB, and H is 25dB~45dB. The output is the minimum transmit power adjustment value corresponding to the uth user in the interference environment. The optimization target and condition corresponding to the uth user corresponding to the uth user transmit power adjustment amount estimator can be expressed as:

Designing a self-optimizing power control protection mechanism controller enables the intelligent deployment cascade control (IDCC) device to self-optimize the transmit power adjustment for indoor small cell users, avoiding interference with outdoor neighboring cell users, thus reducing indoor small cells System performance within.

In order to further explain the present invention, it will be helpful to review the review by the following illustrations, illustrations, and detailed description of the invention.

[Note 6] Jyh-Shing Roger Jang, "ANFIS: Adaptive-Network-Based Fuzzy Inference System", IEEE Transaction on System, Man, and Cybernetics, Vol. 23, NO. 3, June 1993.

IDCC‧‧‧Smart deployment cascade control

CQI‧‧‧ channel quality indicators

SR‧‧‧Service reliability

MU‧‧‧Multiple users

ITPSC‧‧‧Initial Transmit Power Configurator

CQIDC‧‧‧ channel quality indicator decision maker

SOPC‧‧‧ self-optimizing power controller

R _u ‧‧‧The radius requirement of the u user

nRB _u ‧‧‧Number of resource blocks for the uth user

nUE‧‧‧The total user equipment side in the same room

T _{min, u} ‧‧‧The minimum access rate requirement for the u-user

CQI _{min, u} ‧‧‧The minimum CQI requirement for the u-user

CQI _best,u ‧‧‧The best CQI estimate for the u user

P _ini,u ‧‧‧U-user initial transmit power configuration

PL‧‧‧path loss

‧‧‧The average path loss of the uth user

△P‧‧‧The transmission power adjustment amount of the uth user

SINR‧‧‧ News

SINR(t)‧‧‧ transmits the machine-to-machine ratio at time t

‧‧‧The average user's average interference ratio

TTI‧‧‧ transmission time interval

ANFIS‧‧‧Adaptive Network Fuzzy Inference System

Figure 1 System architecture of indoor small cell intelligent deployment serial control device

Figure 2 Initial transmission power controller unit architecture diagram

Figure 3 Optimized CQI decision controller unit architecture diagram

Figure 4 Self-optimized power controller unit architecture diagram

Figure 5. BLER performance curve of small cell down-chain transmission in indoor office A environment

Figure 6. Access rate and system capacity of small cell downlink transmission in indoor office A environment the amount

Figure 7. Plan of the indoor small cell base station power measurement environment

Figure VIII ITU-R indoor office path loss model and measured path loss curve

Figure 9. Initial training power controller (ITPSC) unit training data: in the case of fixed coverage radius of 5 meters and minimum CQI (3, 7, 10), different resource blocks (1~100) Minimum transmit power

Figure 10 (a) Initial Transmit Power Controller (ITPSC) unit initial attribution function

Figure 10 (b) The initial transmit power controller (ITPSC) unit is trained by the hybrid learning algorithm

Figure 10 (c) Initial transmit power root mean square error of the initial transmit power controller (ITPSC) unit

Figure 11 Optimized CQI Decision Controller (CQIDC) unit training data: When the resource blocks are 50, the initial transmit power is -40dBm, -20dBm and 0dBm, respectively, different average path loss (30dB) ~60dB) corresponding optimized CQI

Figure 12 (a) Optimized CQI decision controller (CQIDC) unit initial attribution function

Figure 12 (b) The attribution function of the optimized CQI decision controller (CQIDC) unit after training through the hybrid learning algorithm

Figure 12 (c) Optimum CQI Root Mean Square Error for Optimized CQI Decision Controller (CQIDC) Unit

Figure 13: Training data of one of the transmit power adjustment estimator units: average value of different SINR measurements under the condition of service reliability requirement of 90%, minimum CQI _min of 6, and optimized CQI _best of 7, 10 Corresponding optimal power adjustment

Figure 14 (a) Initial power transfer function of the transmitter power adjustment estimator

Figure 14 (b) The attribution function of the transmit power adjustment estimator unit after training through the hybrid learning algorithm

Figure 14 (c) Radiated root power error of the transmit power adjustment amount of the transmit power adjustment estimator unit

Figure 15 Flow chart of power configuration unit power adjustment control algorithm

Figure 16 IDCC training simulation flow chart

Figure 17 (a) CCRF distribution of SINR measurements for fixed transmit power and smart deployment controllers with interference power of -100dBm

Figure 17 (b) The CCRF distribution of the SINR measurement value of the intelligent deployment of the cascade controller in the interference power of -90dBm

Figure 17 (c) CCRF distribution of SINR measurements for fixed transmit power and smart deployment controllers with interference power of -80dBm

Figure 18 (a) CCDF distribution map of fixed transmission power and smart deployment controller in the interference power of -100dBm

Figure 18 (b) CCDF distribution map of fixed transmission power and intelligent deployment controller in the interference power of -90dBm

Figure 18 (c) CCDF distribution map of fixed transmission power and intelligent deployment controller in the interference power of -80dBm

Figure 19 Average access rate of fixed transmit power and smart deployment cascade control under different interference power environments

Figure 20 Average transmission power of smart deployment cascade control under different interference power environments

The detailed structure of the present invention and its connection relationship are illustrated in conjunction with the following diagrams to facilitate an understanding of the audit committee.

The main structure of the small cell intelligent deployment control device of the present invention is shown in Figure 1, which is a series-connected adaptive controller. Its architecture is divided into resource configurator, minimum access rate/edge CQI converter, initial transmit power setting control (ITPSC) (ANFIS controller #1), optimized channel quality index (CQI). Decision maker (CQIDC) (ANFIS controller #2) and self-optimized power control (SOPC) five units.

The resource allocation unit of the small cell intelligent deployment controller device of the present invention considers that the indoor multi-user resource configuration allocates resource blocks according to the indoor user number and the base station system bandwidth average; the minimum access rate/edge CQI conversion unit, according to Configure the edge CQI corresponding to the user's minimum access rate requirement conversion. The invention uses the adaptive network established by Jyh-Shing Roger Jang in 1993 using the fuzzy inference system, called the adaptive neuro-fuzzy inference system (ANFIS) [Note 6], to realize intelligent self-optimizing power control. ANFIS uses a hybrid learning approach in which the input and output modes mimic the human nervous system, which is characterized by learning to adjust the weights to the appropriate values to perform the functions that people want to achieve. In the initial transmit power configuration (ITPSC) unit, the user sets the radius of coverage, the number of configured resource blocks, and the cell edge CQI requirement according to the size of the indoor room, and as the input parameter of the ANFIS initial transmit power controller, the maximum transmit power at the small cell base station. Under the limitation, the transmission power that can meet the above requirements is output in a non-interference environment. The optimized channel quality indicator decision maker (CQIDC) unit is used as the optimal CQI decision according to the initial setting of the initial transmit power output by the transmit power controller, the number of configured resource blocks, and the path loss average between the transmitters of the user end measurement. The input parameters of the controller, the output satisfies the best CQI with a block error rate (BLER) less than 0.1 in a non-interference environment. My optimized power control unit includes three parts: transmit power adjustment estimator (TPAE) (ANFIS controller #3), power adjustment configurator and self-optimized power control protection mechanism. The transmit power adjustment estimator (TPAE) unit is set by the user's service reliability requirements, cell edge CQI requirements, optimized CQI decision controller output in the interference-free environment, the best CQI and the user-side measurement SINR The average value is used as the input parameter of the self-optimizing minimum transmit power adjustment estimator, and the output satisfies the user service reliability setting and BLER. The minimum transmit power adjustment value of 0.1.

Figure 2 is a block diagram of the initial set transmit power controller unit corresponding to the uth user. The figure contains a five-layer architecture with three input parameters and one output parameter. For the initial transmit power controller unit, the input parameters are respectively u user room The radius, the number of allocated resource blocks and the cell edge CQI requirement are included, and the output parameter is the initial transmission power of the uth user; FIG. 3 is the architecture diagram of the optimized CQI decision controller unit corresponding to the uth user, which includes The five-layer architecture has three input parameters and one output parameter. For the optimal CQI decision maker unit, the input parameters are the average path loss between the uth user relays, the number of configured resource blocks, and the initial set transmit power. The output parameter is the best CQI corresponding to the uth user; FIG. 4 is the self-optimized minimum transmit power controller unit architecture diagram corresponding to the uth user, the figure includes a five-layer architecture, and has three input parameters and one output parameter. For the self-optimizing minimum transmit power controller unit, the input parameters are the cell edge CQI requirement of the uth user, the optimal CQI and SINR average measurement value of the non-interference environment, and the output parameter is corresponding to the uth user. Transmit power adjustment value. The SINR of the present invention is mainly estimated by the RSRP parameter of the UE measurement, and the SINR is reported to the base station [Note 7], and the path loss of the transmission is generated by subtracting the Reference Signal transmitted by the base station from the RSRP [ Note 8].

(A) ANFIS controller architecture

The present invention will be described by taking the self-optimizing minimum transmit power controller unit's transmit power adjustment estimator (TPAE) as an example:

The first layer: in this layer contains three m- _th input parameters x _j,m and the output of the attribution function A _j,n , the attribution function is a bell-shaped function, as shown in (1),

a _j,n , b _j,n , c _j,n are the parameters of the front part, and the parameters will be adjusted using the gradient descent formula.

Second layer: This layer W _i i-th _{output, m} is one received from the pen input is generated by multiplying the first m, and represents the information rate controller ANFIS total of 27 rules.

w _{i , m} = O _{2, i} = A _{1, p} ( x _{1, m} ) × A _{2, q} ( x _{2, m} ) × A _{3, r} ( x _{3, m} ) for i =1, 2..., 27; p =1,2,3; q =1,2,3; r =1,2,3 (5)

The third layer: the action of normalizing the second layer of income.

The fourth layer: O _{4, i} is the output of the fourth layer.

α _i , β _i , γ _i , ω _i are the post-part parameters, and the parameters are adjusted using the least squares estimate LSE. The fuzzy rules R _i , for i =1~27 of the 27 fuzzy inference rule bases are expressed as follows: R ₁ : If( x _1,m is A ₁₁ )and( x _2,m is A ₂₁ )and( x _{3, m} is A ₃₁ )then(output is f _1,m )R ₂ :If( x _1,m is A ₁₁ )and( x _2,m is A ₂₁ )and( x _3,m is A ₃₂ )then(output Is f _2,m )R ₃ :If( x _1,m is A ₁₁ )and( x _2,m is A ₂₁ )and( x _3,m is A ₃₃ )hen(output is f _3,m )⋮R ₂₆ : If( x _1,m is A ₁₃ )and( x _2,m is A ₂₃ )and( x _3,m is A ₃₂ )then(output is f _26,m ) R ₂₇ :If( x _1,m Is A ₁₃ )and( x _2,m is A ₂₃ )and( x _3,m is A ₃₃ )then(output is f _27,m ) (8) Using the designed ANFIS rule base, the system operator will be replaced Produce optimal decisions.

The fifth layer: the output of the fourth layer, using the formula (9) to obtain the final data rate output parameter G _m .

[Note 7] S. Hämäläinen and H. Sanneck, LTE: Self Organising Networks (SON): Network Management Automation for Operational Efficiency , Wiley, Jan. 30, 2012

[Note 8] 3GPP Technical Specification 136.213, "Evolved Universal Terrestrial Radio Access (E-UTRA); Physical layer procedures" (Release 8), www.3gpp.org

(B) Minimum access rate/edge CQI conversion unit

In order to meet the user access rate setting and the receiver block error rate (BLER) is less than 0.1, we must obtain the correspondence between the access rate and the SINR threshold under different CQI, so the indoor small cell transfer machine is performed. Chain transmission block error rate (BLER) and pass rate performance simulation. In the simulation of the embodiment, the system simulation parameters are shown in Table 1. The basic parameters of the small cell downlink transceiver are shown in Table 2.

The invention will perform BLER performance simulation of 1x1 SISO-OFDM transmission machine under different CQI conditions, the system bandwidth is set to 20MHz, and the channel mode is Indoor Office A (IOA) [Note 9], channel estimation is used. The Least square (LS) is estimated, and the equalizer is equalized by minimum mean square error (MMSE). The user's moving speed is assumed to be 10 km/hr, and 1000 sub-frames are used for simulation under different SINR conditions. Five is shown. The invention can generate BLER under different CQI conditions by using FIG. The SINR threshold of 0.1 is shown in Table 3. The physical resource block (PRB) in Table 3 is generated based on the 20 MHz system bandwidth and three indoor user configurations.

According to the BLER performance of the above-mentioned transfer machine, the simulation results of the access rate are simulated, as shown in Fig. 6. The present invention considers a frequency division full-duplex mode, and separates uplink and downlink signals in the frequency domain, and the resources in one direction are continuous in time, and as the mobile communication technology evolves, the system access rate is gradually approaching. Shannon upper limit, the capacity of the downlink system can be expressed as [Note 10] Where BW and BW_eff are the system bandwidth and the effective bandwidth of the system respectively; η is the correction parameter; SINR and SINR_eff are the signal-to-interference ratio and the effective signal-to-interference ratio; nRB _total is the total resource block number of the system and nRB _u is the u The number of resource blocks that the user has assigned. In the present invention, the simulation parameter setting of the SISO downlink transmission system is as shown in Table 1. The BW is 20 MHz, the BW_eff is 0.83, the η is 0.43, and the SINR_eff is 2.5119 (4 dB). The system capacity is shown by the black dotted line in Figure 6, and the CQI is different. The actual transmission rate of the actual transmission machine is shown in the solid line of Figure 6.

When the system bandwidth is 20MHz, the total number of resource blocks is 100. Assuming that three indoor users are configured by resources, the number of resource blocks is 33, and the system access corresponding to different SINR thresholds is calculated by (10). The rate is shown in Table 3. Therefore, when the user input access rate requirement range is 1.99~2.76Mbps, 6.76~7.44Mbps, 12.99~14.13Mbps, according to Table 3, the cell edge CQI 3, 7, 10 respectively. Therefore, the present invention can obtain a minimum access rate/edge CQI conversion by using (10) for different resource blocks of different users and obtaining corresponding correspondence between different access rates and SINR thresholds.

(C) Initial setting transmit power controller (ITPSC) unit

In order to control the initial transmit power of the small cell base station, so that the intracellular users can meet the user's set requirements, we will use the indoor small cell transfer machine to transmit the block error rate (BLER) performance simulation results. Training information for the controller.

The invention considers the multi-user service reliability requirement of the indoor small cell attenuation environment, and the signal intensity P _r of the intracellular base station is log-normally distributed.

In the middle And σ _W are the average signal strength and standard deviation, respectively. The average value of the signal strength is

Where K is the (minimum) average signal strength (dBm) of the receiver at d = R , and N is the path loss index. For mobile users who are at a distance d from the base station _, the probability that P _{r is} greater than the threshold value P _r,min

Where R is the cell radius and KP _r,min (dB) is Fade Margin(FM) for d = R to ensure the reliability of the communication link.

The service reliability is determined in cells with a cell radius R , and the received signal strength P _{r of the} user equipment (UE) receiver is greater than the percentage of P _r,min . Service reliability can be expressed as [Note 11]

(13) Bringing in (14) the service reliability is

make

The controller will use the link estimation to generate the minimum transmit power under different coverage radius ( R ), service reliability ( SR ), and cell edge CQI ( CQI _min ) requirements, which is the initial transmit power, which corresponds to the uth user. The link estimate can be expressed as follows: P _{ini , u} = P _{r min, u} ( CQ I _{min, u} ) + L _t - G _t + PL ( R _u ) + FM ( SR _u )- G _r + L _r (17 Where P _{ini, u} is the initial transmit power (dBm) of the uth UE, P _{rmin, u} ( CQI _{min, u} ) is the receiving sensitivity of the uth UE at the cell edge CQI equal to CQI _{min, u} , L _t is the base station cable transmission loss (dB), G _t is the base station antenna gain (dBi), PL (R _u ) is the uth user path loss at the transmission machine distance equal to R , FM(SR _u ) is The uth user has a service reliability equal to the fade margin (FM) (dB) of SR _u , G _r is the client antenna gain (dBi), and L _r is the user body loss (dB). The attenuation margin value can be calculated by the equation (15)(16).

The present invention assumes that the basic parameters of the indoor small cell downlink transceiver are as shown in Table 2. In this embodiment, the single antenna mode (SISO), the present invention is also applicable to the multi-antenna mode (MIMO), with the parameters of Table 2 and Transmitter BLER performance simulation, which can calculate the u-th user receiving sensitivity of CQI requirements at different cell edges in a non-interfering environment

P _{r min, u} ( CQ I _{min, u} )= P _{N , u} + SNR _th ( CQ I _{min, u} ) (18) where P _N,u is the noise energy (dBm) of the uth user, SNR _th (CQI _{min, u} ) is the SNR threshold of BLER=0.1 for the uth user under different CQI _{min, u} (as shown in Table 3). The noise power P _N,u can be expressed as follows: P _{N , u} = NF × N ₀ (W) = NF × kT ₀ × BW _{r , u} (W) = NF (dB) + (-174) + 10log ₁₀ ( BW _r,u )(dBm) (19) NF is the noise figure (NF) of the UE, N ₀ is the thermal noise power spectral density (dBm/Hz) at the receiving end, and k is the Boltzmann constant of 1.3804 x 10 ^-23 Joule/°K, T ₀ is the absolute temperature 290°K, BW _{r, u} is the uth user receiving bandwidth (Hz), which is expressed as follows: BW _{r , u} =15kHz×12× nRB _u (20) Where nRB _u is the number of resource blocks configured by the uth user.

Reference ITU-R indoor path loss estimation model of the present invention is an analog link, may be represented by the following formula [Note 12] PL (d) = 20log 10 (f) +10 N log 10 (d) + L f (n) -28 ( dB ) (21) where d is the distance between the transmissions (m); f is the carrier frequency (MHz), which is used to match the actual small cell base station experimental measurement, the carrier frequency is set to 2350MHz; N is the path loss index L _f (n) is the floor penetration loss (dB), n is the number of floors, and the floor penetration loss is not considered in this experiment.

Experimental measurement

In order to make the two parameters of the indoor power path loss and the standard deviation of the received signal strength closer to the actual environment, we carry out small cell base station power measurement in the laboratory. The basic parameters of the transfer machine are listed in Table 2, and the experimental environment considers There are two cases: line of sight (LOS) and non-line of sight (NLOS). The environment is shown in Figure 7. The laboratory is about 20.98 meters long. The width is about 7.30 meters, the small cell base station is placed on the right side of the laboratory, and the mobile phone measurement end is 1~19 meters away from the small cell base station. The star pattern is the measurement point. Figure 7 (a) is 19 point direct view. The LOS measurement point, Figure 7(b) is the 36-point non-direct-view (NLOS) measurement point, and each of the different measurement points will perform 2000 Reference Signal Received Power (RSRP) measurements.

The invention uses the mobile phone to measure the signal transmitted by the base station, and performs the difference statistics on the RSRP measured by the different transmission machine spacing, and then corrects the path loss model in the ITU-R office to approximate the path loss of the signal transmission in the current environment. The experimental results of the path loss are shown in Figure 8, where the solid black line is the path loss model in the ITU-R office; the dotted line is the path loss measured by the actual direct visibility (LOS), and the square dotted line is the actual non-direct view. (NLOS) The measured path loss is estimated by using the RSRP measured at the base station and the average of 2000 RSRP values for each distance measurement, that is, the actual measured path loss at each distance. The solid line is a circle approximation of the actual measured distribution of-sight (LOS) the amount of the ITU-R office path loss model is modified _{as: PL (d) = 20log 10} (f) + 19.5log 10 (d) -36 (dB) ( 22) wherein the carrier frequency f is 2350 MHz and the path loss index N is 1.95. Solid squares approximation of the actual line was measured non-distribution-of-sight (the NLOS) the amount of the ITU-R office path loss model is modified _{as: PL (d) = 20log 10} (f) + 28log 10 (d) -36 (dB) (23) wherein the carrier frequency f is 2350 MHz and the path loss index N is 2.8.

The present invention considers that the indoor environment is mostly blocked by a plurality of partitions, so the non-direct visibility (NLOS) correction model (23) is used as the following simulation and design.

The received power shading weak standard deviation σ _W can be generated by using the RSRP calculated by the different measuring points to calculate the average standard deviation 4.27dB, which is used as the masking weak standard deviation of the test environment. Therefore, the shadow attenuation standard deviation σ _W and the path loss constant N are brought into equations (15) and (16), and the attenuation margin value (FM) corresponding to 90% of the service reliability can be obtained by 2.14 dB.

In order to meet the user's set requirements and achieve the minimum transmit power, the present invention simulates the BLER results of the transfer machine by the basic parameters of the indoor small cell transfer machine, as listed in Table 2, as listed in Table 3. Combined with Tables 2 and 3 and (18)(19)(20)(21)(23), the ratios of the different coverage radii (2.5, 5, 7.5, 10, 12.5, 15 meters) and the number of resource blocks (1) are generated. ~100) and the minimum transmit power under different minimum CQI (1~15) conditions, as the training data of the initial transmit power controller. Figure 9 shows one of the training data of the initial set transmit power controller: the minimum corresponding to different resource blocks (1~100) under the fixed coverage radius of 5 meters and the minimum CQI (3, 7, 10). Transmit power. Taking 50 resource blocks as an example, the minimum transmit powers corresponding to different minimum CQIs (3, 7, 10) are -35.33 dBm, -27.33 dBm, and -18.33 dBm, respectively.

In the initial transmit power controller unit, the definition of the coverage radius is less than 5 meters, L, 6 meters to 10 meters is M, 11 meters to 15 meters is H, and the number of resource blocks is 0 to 25 L, 26~74 are M, 75~100 are H, cell edge CQI demand is 1~5 is L, 6~10 is M, 11~15 is H. The fuzzy inference rules are shown in Table 4.

The purpose of initially setting the transmit power controller unit is to set the initial minimum transmit power of the base station. In the initial non-interference environment, the coverage radius, service reliability and cell edge CQI requirements of the u-th user are met. For ITPSC, the inputs are the radius of coverage ( R _u ), the number of resource blocks ( nRB _u ), and the cell edge CQI ( CQI _{min, u} ). Each input parameter uses three bell-shaped assignment functions (Generalized bell-shaped). )defined as There are 27 fuzzy inference rules. The fuzzy inference rules are shown in Table 4. The target output corresponding to the u-th user is the initial transmit power (P _{ini, u} ) .

Figure 10 (a) and Figure 10 (b) are the ITPSC initial attribution function and the attribution function after training through the hybrid learning algorithm. After 500 trainings, the rms error of the initial transmit power is reduced to 0.9 dBm, which converges after about 300 trainings, as shown in Figure 10 (c).

[Note 9] ITU-R, Recommendation ITU-R M.1225, GUIDELINES FOR EVALUATION OF RADIO TRANSMISSION TECHNOLOGIES FOR IMT-2000, 1997

[Note 10] Mogensen, P., et al, “LTE Capacity compared to the Shannon Bound”, Vehicular Technology Conference, 2007. VTC2007-Spring. IEEE 65th, pp.1234-1238, April 2007.

[Note 11] Wiliam C. Jakes, Jr., Microwave Mobile Communications, 1974

[Note 12] ITU-R, Recommendation ITU-R P.1238-7 (02/2012) Propagation data and prediction methods for the planning of indoor radiocommunication systems and radio local area networks in the frequency range 900 MHz to 100 GHz, 02 /2012

(D) Optimized CQI Decision Controller (CQIDC) unit

In the actual wireless channel environment, the indoor small cell base station smart deployment will face the same frequency interference of the external large base station or adjacent small cells. Therefore, in order to prevent the communication quality degradation caused by the interference, the self-optimized deployment tandem control device uses the optimized CQI decision controller (CQIDC) unit, and the decision can satisfy the BLER in the interference-free environment. The best CQI of 0.1. In addition, in the interference environment, the self-optimized control power controller (SOPC) unit continuously tracks the measured SINR, and self-optimizes the transmission power, so that the user meets the target requirements of service reliability and minimum transmission power.

In the interference-free environment, the CQIDC unit estimates that the location of the user terminal can satisfy the BLER in a non-interference environment based on the link estimation. The best CQI of 0.1 is to give a CQI target in an ideal environment, so that the controller can adjust the power in the interference environment by this target to compensate for the degradation of communication quality caused by the interference. In order to determine the optimal CQI of the user's location in a non-interfering environment, we must first estimate the SNR _{u of} the uth user in a non-interfering environment, which can be expressed as follows: SNR _u = P _{r , u} (W) / P _{N , u} (W) (25) where P _r,u is the average received power of the uth user in a non-interference environment, which can be expressed as follows

When we know the initial transmission power of the u-th user, P _{ini, u} , the average path loss of the u-th user and the base station As shown in Table 2, other parameters can be used to calculate the received power P _{r,u of} the u-th user in the interference-free environment using (26); the noise power P _{N of} the u-th user _{, u is} (19) )formula.

After estimating the SNR in a non-interfering environment, we can find the u-th user to meet the BLER by Table 3. The best CQI of 0.1, the decision mode is as follows

The CQIDC unit is used to determine the optimal CQI that should be met in a non-interfering environment. The present invention simulates the BLER results of the transfer machine by the basic parameters of the indoor small cell transfer machine, as listed in Table 2, as listed in Table 3. Combined with Table 2, Table 3, and (25)(26)(27), different average path loss (30dB~70dB), initial transmit power (-75dBm~20dBm) and resource block number (1~100) are generated. The best CQI output is used as training material. Figure 11 shows the training data of one of the optimized CQI decision controllers: when the resource blocks are 50, and the initial transmit power is -40dBm, -20dBm and 0dBm, respectively, the average path loss (30dB~70dB) Corresponding to the optimized CQI, it can be observed that as the path loss increases, the optimal CQI of the decision will become smaller, and if the initial transmit power is larger, the decision of the same average path loss is optimized. The larger the CQI.

In the optimized CQI decision controller unit, the average path loss is defined as 30dB~40dB for L, 41dB~60dB for M, and 61dB~70dB for H; the initial transmit power is -75dBm~-51dBm for L, -50dBm~- 4dBm is M, -3dBm~20dBm is H; the number of resource blocks is 0~25 for L, 26~74 for M, and 75~100 for H. The fuzzy inference rules are shown in Table 5.

Optimized CQI decision controller unit The minimum transmit power of the uth user in the interference-free environment output by ITPSC, the path loss estimated based on the position of the uth user from the previous location, and the resource block currently used by the uth user The number determines the best CQI that the uth user should meet in a non-interfering environment. For the optimized CQI decision controller unit, the input parameters are the average value of the path loss estimate between the uth user terminals. The initial transmit power ( P _{ini, u} ) and the uth user resource block number ( nRB _u ) output by ITPSC , each input parameter is defined by using three Gaussian attribution functions as There are 27 fuzzy inference rules, and the fuzzy inference rules are shown in Table 5. The target output is the best CQI ( CQI _{best, u} ) for the uth user in a non-interfering environment.

Figure 12 (a) and Figure 12 (b) are the initial attribution letters of the ANFIS CQI optimization decision controller. The number and the attribution function after training through the hybrid learning algorithm. After 300 trainings, the rms error of the best CQI is reduced to 0.46, which converges at about 250 training sessions, as shown in Figure 12(c).

(E) Self-optimizing power controller (SOPC) unit

The self-optimizing minimum transmit power controller consists of a transmit power adjustment estimator (TPAE) and a power adjustment configurator. TPAE unit through user demand setting, CQIDC optimization decision controller output ( CQI _{best, u} ) and user measurement information The transmission power adjustment is made so that the indoor small cell users can still meet the user setting requirements and the communication quality requirements of the transmission machine in the interference environment, and the self-optimized control indoor small cell base station transmits the minimum power, thereby reducing the interference to neighboring cells.

The self-optimizing power controller consists of two parts, the power adjustment amount estimation and the power adjustment amount configuration. The power adjustment amount estimation is mainly for each user to calculate the minimum transmission power adjustment amount in accordance with the user setting requirement and the communication quality of the transmission machine in the interference environment; the power adjustment amount configuration is for the sum of the powers after the adjustment of each user exceeds the maximum transmission. At power, the controller will increase or decrease the power of each user in a step-by-step manner so that the transmit power of each user is fairly configured within the maximum transmit power limit.

(i) Transmit power adjustment estimator

In the estimation part of the transmission power adjustment amount, in order to satisfy the BLER in the interference environment 0.1 and service reliability requirements, so the u-th target SINR threshold should not be set below the minimum SNR threshold SNR _th ( CQI ) that satisfies the service reliability ( SR _u ) and the edge CQI ( CQI _{min, u} ) _{Min, u} )+ FM ( SR _u ), so the uth user target SINR threshold is defined by the following equation SINR _{th , u} =max{ SNR _th ( CQI _{min, u} )+ FM ( SR _u ), SNR _th ( CQI _{Best , u} )}(dB) (29) Then, by this target SINR threshold, the minimum transmit power self-optimization adjustment is performed, and the transmit power adjustment amount is expressed as follows among them The average value of the SINR measured for the uth client.

In order to meet the user-set SR requirement and output the minimum transmit power in the interference environment, the TPAE unit generates the signal-to-interference ratio threshold of the corresponding cell edge CQI of Table 3 by the indoor small cell transfer machine BLER simulation result. Combined with Tables 2, 3 and (29) (30), different CQIs (1~15) and different CQIs (1~15) are obtained in different cell edge CQIs (1~15). The minimum transmit power adjustment amount under the SINR measurement average (-25dB~45dB) is used as training data. Figure 13 shows the training data for a group of self-optimizing power controllers: in the case of service reliability requirement of 90%, minimum CQI _{min, u} is 6, optimized CQI _{best, u} is 7, 10, different SINR measurements The optimal power adjustment corresponding to the average value shows that the power adjustment amount decreases as the measured SINR average value increases to achieve the minimum transmission power design goal. The solid curve of the circle is adjusted according to the formula (28). The target SINR threshold should be the user edge CQI demand CQI _min, the SINR threshold corresponding to _u = 6 (12dB) plus the FM corresponding to the 90% service reliability requirement. =2.14dB, the summed SINR threshold is 14.14dB greater than CQI _{best, u} =7 corresponds to the SNR threshold (13dB), so the adjustment target SINR threshold is 14.14dB: the solid line curve is due to CQI _{Best, u} = 10 corresponds to the SNR threshold (22dB) greater than 14.14dB, so the adjustment target SINR threshold is 22dB. Figure 13 shows that when the power adjustment amount is 0 dB, the average SINR of the corresponding CQI _{best, u} = 6, 10 is 14.14 dB and 22 dB, respectively.

In the transmit power adjustment estimator, the optimal CQI is defined as 1 to 5 for L and 6 to 10 for M. 11~15 is H, and the average value of SINR is L at -25dB~-5dB, M is -5dB~25dB, and H is 25dB~45dB. The fuzzy inference rules are shown in Table 6.

The transmit power adjustment estimator sets the average CQI and the SINR of the user equipment in the interference-free environment of the CQIC output by the user. In the multi-user interference environment, the assignment meets the service reliability requirement. The minimum transmit power is given to the transceiver. For the transmit power adjustment estimator, the uth user input is the cell edge CQI demand (CQI _{min, u} ), the best CQI (CQI _{best, u} ) and the SINR measurement average. Each input parameter uses three bell-shaped attribution functions as shown in equation (4). There are 27 fuzzy inference rules. The fuzzy inference rules are shown in Table 6. The u-th user target output is the minimum transmit power adjustment. Quantity ( △P _u ).

Figure 14 (a) and Figure 14 (b) show the initial assignment function of the transmit power adjustment estimator and the attribution function after training by the hybrid learning algorithm. After 500 trainings, the root mean square error of the minimum transmit power adjustment is reduced to 0.86 dB, which converges at about 300 training sessions, as shown in Figure 14 (c).

(ii) Power adjustment configuration algorithm

In the power adjustment configuration part, in order to control the sum of the adjusted powers of each user after the power modulation estimation does not exceed the maximum transmission power of the indoor small cell base station, the flow chart of the power configuration algorithm is shown in FIG. First, each user compensates its power adjustment amount ΔP _u to the previous time transmission power P _u (m-1), then calculates the sum of powers that all users should transmit, and determines whether the total transmission power P _total is greater than the maximum transmission power P _max . If it is less than P _max , the power allocation is performed for the compensated transmission power P _u (m) of each user; if it is greater than P _max , the power reversal step adjustment is performed. Where P _tmp,u (i) is the i-th ring temporary storage value of each user's previous time transmission power, and each cycle P _tmp,u (i) will increase or decrease the power of △ , and ± △ in this implementation The example assumes ±0.1dB, when ΔP _u >0, ie P _tmp,u (i) increases by 0.1dB in each loop; when ΔP _u <0, then P _tmp,u (i) in each The circle is reduced by 0.1 dB. After summing the powers P _tmp,u (i+1) of each user through the step adjustment, it is determined whether the total transmission power P _total is greater than P _max , and if not greater than P _max , then for P _{tmp, u} (i+1 ) Continue to adjust with △ power adjustment; if it is greater than P _max , set the user transmit power temporary value P _{tmp, u} (i) of the previous circle to the user transmit power P _u (m), and the power can be completed. Configuration.

(F) Self-optimizing power control protection mechanism

In order to make the IDCC device only self-optimize the transmission power adjustment for indoor small cell users, avoid interference with outdoor neighboring cell users, and reduce the performance of the IDCC device. The invention sets a self-optimizing power control protection mechanism, and calculates an average road according to the estimated value of the return path loss of the user equipment end. The path loss is then estimated according to the indoor path loss model (23) to estimate the distance between the client and the base station (d). If the user equipment end is within the radius (R), the self-optimized minimum transmit power control is performed.

(G) Simulation experiment results

The invention simulates and verifies the reliability of multi-user service. In the experiment, three users individually measure multiple times in the indoor small cell coverage area covering the radius R, and each measurement point is measured 1000 times. The total number of measurement points is defined as follows: On the circumference of the radius r=1m, the 7-point measurement is performed uniformly; on the circumference of the radius r=2, 3, 4, 5 meters, the points are uniformly distributed, 14, 21, 28, 35 points respectively. Measurement; when the user sets a radius of 5 meters and the total number of measurement points is 105 points, the total number of measurement points will increase with the radius of the user-set coverage.

The invention verifies that the performance of the SINR measurement, the access rate and the service reliability measured by the user uniformly distributed in different locations in the coverage area will be observed by the complementary cumulative distribution function (CCDF). Taking the SINR measurement value CCDF as an example, it is defined as follows: F(SINR _th ) = P (SINR measurement value > SINR _th ) (32) CCDF has the same meaning as service reliability, and corresponds to SINR measurement value = SINR _th The probability value is the ratio of all indoor multi-user SINR measurements greater than SINR _th , that is, service reliability. According to formulas (15) and (16), the meaning can be shown to be consistent with service reliability.

The embodiments of the present invention all assume that there are a total of three users in the indoor small cells, and the radius of the coverage is set. Set to 5 meters, service reliability requirements are set to 90%, each user sets different minimum access rate requirements: UE1 is 2.76Mbps, UE2 is 7.44Mbps, UE3 is 14.13Mbps, and the following simulation is performed.

The present invention will compare the performance of a fixed transmit power in a different interference environment, i.e., using only the output of the initial transmit power controller (ITPSC) and the performance of the intelligent deployment cascade controller (IDCC). The resource allocation part is assumed to be an average configuration, that is, three users are configured with 33 resource blocks; and the minimum access rate is converted by the minimum access rate/minimum CQI conversion unit, and the minimum CQI corresponding to each user is 3, 7, respectively. 10; The output power of the initial transmit power controller is the minimum transmit power that satisfies the user's coverage radius, service reliability, and minimum access rate requirement in a non-interference environment, respectively, UE1 is -36.83dBm, UE2 is - 27.72dBm, UE3 It is -19.03dBm.

Figure 17 shows the use of smart deployment cascade control and fixed transmit power control. As the interference power increases, its SINR measurement performance is compared. By comparing Fig. 17(a) with Fig. 17(c), the solid line part is the SINR performance using the smart deployment cascade control, and the broken line part is the SINR performance of the fixed transmission power control. Figure 17 (a) shows that in the approximate interference-free (-100dBm) environment, the ratio of the SINR measurement value (5dB) corresponding to the UE1 SINR measurement value greater than the minimum CQI (CQI _min = 3) is about 90%, fixed. The transmit power is about 87%; the UE2 SINR measurement is greater than the minimum CQI (CQI _min = 7) corresponding to the SINR threshold (13dB) is about 90%, and the fixed transmit power is about 89%; UE3 SINR The measured value is greater than the minimum CQI (CQI _min = 10) corresponding to the SINR threshold (22dB) ratio of about 90%, the fixed transmission power is about 89%; as the interference becomes larger, Figure 17 (c) It is shown that in the environment where the interference power is -80 dBm, the ratio of the SINR measurement value corresponding to the UE1 SINR measurement value to the minimum CQI (CQI _min = 3) is about 89%, and the fixed transmission power is about 14%; The ratio of the UE2 SINR measured value to the SINR threshold (13dB) corresponding to the minimum CQI (CQI _min = 7) is approximately 89%, and the fixed transmit power is approximately 15%; the UE3 SINR measured value is greater than the minimum CQI (CQI) _Min = 10) The corresponding SINR threshold (22dB) is approximately 89%, and the fixed transmit power is approximately 15%.

Figure 18 shows the use of smart deployment cascade control and the above fixed transmit power control. As the interference power increases, its access rate performance is compared. The solid line part is the access rate performance using self-optimized deployment tandem control, and the dotted line part is the access rate performance of fixed transmission power control. By comparing Fig. 18(a) with Fig. 18(c), Fig. 18(a) shows that the UE1 access rate is greater than the minimum access rate requirement (2.76Mbps) in an approximately interference-free (-100dBm) environment. For 90%, the fixed transmit power is about 87%; the UE2 access rate is greater than the minimum pass rate requirement (7.44Mbps) is about 90%, the fixed transmit power is about 89%; and the UE3 pass rate is greater than the minimum pass rate. The ratio of demand (14.13Mbps) is about 90%, and the fixed transmission power is about 89%.; As the interference becomes larger, Figure 18(c) shows that the UE1 access rate is greater than the minimum accessibility in the case of interference power of -80dBm. The ratio of demand (2.76 Mbps) is about 89%, and the fixed transmission power is about 14%; the ratio of UE2 access rate is greater than the minimum access rate requirement (7.44 Mbps) is about 89%, and the fixed transmission power is about 15%; The UE3 access rate is greater than the minimum access rate requirement (14.13 Mbps), which is approximately 89%, and the fixed transmit power is approximately 15%.

It can be observed from Figure 17 and Figure 18 that with the increase of interference power, the service reliability of only fixed transmit power can no longer maintain the user's requirement for setting service reliability, and using smart deployment cascade control can make The three users of indoor small cells continue to maintain a demand that is approximately 90% of the user-defined service reliability.

Figure 19 shows the average access rate and average total access rate of users with fixed transmit power and intelligent deployment of cascaded controllers under different interference environments. The solid line portion is the average access rate performance using the smart deployment cascade controller, and the dotted line portion is the average throughput performance of the fixed transmission power control. Figure 19 shows the use of a smart deployment cascade controller when the interference power is -100dBm. The average access rate of UE1 is about 10.48 Mbps, the average access rate of UE2 is about 17.38 Mbps, the average access rate of UE3 is about 24.59 Mbps, and the total average access rate is about 52.45 Mbps; the average access rate of fixed transmit power UE1 is about 10.86 Mbps, UE2 The average access rate is 17.38 Mbps, the average access rate of UE3 is about 24.08 Mbps, and the total average access rate is 52.45 Mbps. As the interference power increases, when the interference power is -80dBm, the intelligent deployment of the serial controller is used. The average access rate of UE1 is 11.48Mbps, the average access rate of UE2 is 17.31Mbps, and the average access rate of UE3 is 22.97Mbps. The access rate is 51.75Mbps; the average access rate of UE1 with fixed transmit power is reduced to 2.97Mbps, the average access rate of UE2 is reduced to 6.93Mbps, the average access rate of UE3 is reduced to 12.41Mbps, and the total average access rate is reduced to 22.32Mbps. From the above results, the fixed transmission power can be found, and the average access rate decreases significantly with the increase of the interference. However, with the intelligent deployment of the serial controller, the average access rate can be maintained even in different interference environments. Performance.

Figure 20 shows the average transmit power in a different interference (-100dBm~-80dBm) environment using a smart deployment cascade controller. Figure 20 shows the use of a smart deployment cascade controller, when the interference power increases, its transmit power will gradually increase. When the interference power is -100dBm, the average output power of UE 1 is about -39.99dBm, the average output power of UE 2 is about -29.85dBm, the average output power of UE 3 is about -22.36dBm, and the average total transmit power is -21.58dBm; As the interference becomes larger, the average output power of UE 1 is about -2.46 dBm when the interference power is -80 dBm, the average output power of UE 2 is about -13.56 dBm, and the average output power of UE 3 is about -5.886 dBm, the average total emission. The power is -5.154 dBm, and the results show that the intelligent deployment of the cascade controller can adaptively adjust the transmit power to maintain the multi-user service reliability requirements as the environmental interference power changes.

Therefore, the simulation results show that the indoor small cell smart deployment cascade control device includes the initial transmit power controller unit, the optimized CQI decision controller unit and the self-optimization function. The rate controller unit can not only meet the service reliability requirements set by the user in the multi-user interference environment, but also transmit the signal with the minimum transmission power as much as possible to meet the user's demand, thereby saving energy and reducing the generation of neighboring users of heterogeneous networks. The goal of co-channel interference.

In the simulation of this embodiment, the basic parameters of the transceiver are as shown in Table 2, which is single antenna mode (SISO), and the present invention is also applicable to multi-antenna mode (MIMO) and other different channel environments.

Claims (9)

An indoor small cell intelligent deployment serial control device constructed in a multi-user interference environment built in frequency division full duplex and orthogonal frequency division multiplexing access, enables an indoor small cell base station to satisfy the transfer block The error rate (BLER) is less than 0.1, the multi-user (MU) service reliability, the best access rate and the cell radius requirement, and the self-optimized deployment tandem control device includes: The transceiver uses a frequency division full duplex (FDD) to separate the uplink and downlink signals; a base station uses orthogonal frequency division multiple access (OFDMA) to allocate resources; a resource configuration unit, Configuring a small cell base station resource block according to the number of indoor users and the system bandwidth; a minimum access rate/cell edge channel quality index (CQI) conversion unit is connected to the resource configuration unit and generates a cell edge Channel quality indicator requirements; an initial set transmit power configuration (ITPSC), which allows each indoor user input parameter to be an indoor coverage radius ( R _u ), resource block number ( nRB _u ), and cell edge channel quality indicator ( CQI _{mi) n, u} ) demand; an optimized channel quality index (CQI) decision maker (CQIDC), which is connected in series with the initial set transmit power configuration; a self-optimized power controller (SOPC), which is connected in series An optimized channel quality indicator decision maker; and a self-optimizing power control unit comprising a transmit power adjustment amount estimator (TPAE), a power adjustment configurator and a self-optimizing power control protection mechanism; The adjustment quantity estimator (TPAE) is connected in series with the power adjustment configurator; a self-optimizing power control protection mechanism controller controls the activation of the power adjustment configurator; wherein, by the user inputting the service reliability, the The indoor coverage radius and the minimum access rate requirement are intelligently deployed in a multi-user interference environment. The indoor small cell base station intelligent deployment control device estimates by the user's returned information. The average path loss (PL) and the Signal-to-Interference-plus-Noise Ratio (SINR) of the inter-connector are adaptively controlled by an orthogonal frequency division multiplexing The transmit power of the pick-up allows the user of the indoor small cell base station to self-optimize the minimum transmit power to meet the multi-user service reliability and optimize the access rate requirements of the cell edge, and to minimize the transmit power and adjacent Co-channel interference between cells; and the adaptive network fuzzy inference system design of the optimized channel quality index (CQI) decision maker (CQIDC) unit is based on an initial setting of the transmit power controller The initial transmit power of the output, the average path loss between the transmitters of the client and the number of resource blocks required by the user are used as input parameters of the optimized channel quality indicator decision maker unit, and the output parameters are in a non-interference environment. The best channel quality indicator with block error rate less than 0.1, there are three input parameters and one output parameter; for the optimized channel quality indicator decision unit, each indoor user has three input parameters, respectively Average path loss between the relays of the uth user ( ), initial transmit power setting (P _{ini, u} ) and resource block number (nRB _u ), each input parameter uses Gaussian attribution function, each Gaussian attribution function is divided into 3 levels, a total of 27 fuzzy inference rules, The output is the optimal channel quality indicator (CQIbest, u) corresponding to the uth user in the interference-free environment; the optimized target and condition of the optimized channel quality indicator decision maker unit corresponding to the u-th user can be expressed as : BLER in a non-interfering environment 0.1, the optimal CQI decision target optimization assignment for the u-th user Restricted conditions: { , P _ini,u , nRB _u }30dB 70dB -75dBm P _ini,u 20dBm 1 nRB _u 100 CQI _best,u {1~15}. The indoor small cell intelligent deployment serial control device constructed in the frequency division full duplex and the orthogonal frequency division multiplexing access device in the multi-user interference environment described in claim 1 of the patent scope, wherein the initial setting transmit power The configuration is based on an adaptive network-based fuzzy inference system (ANFIS) architecture. There are three input parameters and one output parameter. For an initial setting of the transmit power controller unit, each indoor user The input parameters are the indoor coverage radius ( R _u ), the number of resource blocks ( nRB _u ) and the cell edge channel quality indicator ( CQI _{min, u} ), and the block error rate in accordance with the OFDM transmission machine is lower than a preset value ( 0.1) Below, the output parameter corresponds to the initial minimum transmit power value ( P _{ini, u} ) set by the uth user is assigned to the _transceiver, and each input parameter uses a generalized Bell-Shaped function, each The bell-shaped attribution function is divided into three levels, and there are 27 fuzzy inference rules. The target and condition for optimizing the initial transmission power controller of the u-user ANFIS can be expressed as: corresponding to the u-th user base. The minimum initial transmission power is assigned the best optimization goal Restricted conditions: { R _u , nRB _u , CQI _min,u }0m< R _u 15m 1 nRB _u 100 1 CQI _min,u 15 P _ini,u { 20dBm}. The indoor small cell intelligent deployment serial control device constructed in the frequency division full duplex and the orthogonal frequency division multiplexing access device in the multi-user interference environment described in the second application scope, wherein the initial setting transmit power configuration According to the adaptive network fuzzy inference system architecture design, the parameters of the front part and the post parts are trained and adjusted by the following training data; in order to meet the user setting and block error rate in the non-interference environment The minimum transmit power of 0.1 demand, corresponding to the initial transmit power training data of the uth user, can be estimated by the following formula; P _{ini, u} = P _{rmin, u} (CQI _{min, u} ) + L _{t -} G _t + PL (R _u )+FM(SR _u )-G _r +L _r The training data required for the initial setting of the transmit power configuration unit by satisfying the service reliability by 90% and covering the radius (2.5, 5, 7.5, 10, 12.5, 15 gong) The ruler), the number of resource blocks (1~100) and the minimum transmit power (P _{ini, u} ) corresponding to the cell edge channel quality index (1~15) are different. The indoor small cell intelligent deployment serial control device constructed in the multi-user interference environment described in the first application of the scope of the application of the frequency division full duplex and the orthogonal frequency division multiplexing access, wherein the optimized channel quality The indicator decision maker unit is designed according to the adaptive network fuzzy inference system. The parameters of the front part and the post part are trained and adjusted by the following training data; in the non-interference environment, the block error rate is satisfied. The optimal channel quality indicator ( CQI _{best, u} ) training data of 0.1 can be estimated and determined by the SNR _{u of} the u-th user position. The training data required for the optimized channel quality indicator decision maker unit is determined by the average path loss (30dB~70dB), initial transmit power (-75dBm~20dBm) and resource block number (1~100) of different users. Corresponding to the best channel quality indicators. In the multi-user interference environment described in the scope of claims 1 and 4, the indoor small cell intelligent deployment serial control device constructed in the frequency division full duplex and the orthogonal frequency division multiplexing access device is included in the office. modifying the average path loss ITU (path loss exponent of 2.8) the following model _{formula: PL (d) = 20log 10} (f) + 28log 10 (d) -36 (dB). The self-optimized power control is constructed in a multi-user interference environment as described in claim 1 of the invention, which is constructed in a frequency division full-duplex and orthogonal frequency division multiplexing access indoor small-cell intelligent deployment control device. The device is composed of the transmit power adjustment amount estimator, the power adjustment configurator and the self-optimizing power control protection mechanism controller. The self-optimized power control is constructed in a multi-user interference environment as described in claim 1 of the invention, which is constructed in a frequency division full-duplex and orthogonal frequency division multiplexing access indoor small-cell intelligent deployment control device. The transmit power adjustment estimator is designed according to an adaptive network fuzzy inference system. There are three input parameters, which are the cell edge channel quality indicator requirements ( CQI _{min, u} ) set by the uth user. The optimal channel quality index ( CQI _{best, u} ) and the average of the user's measurement signal in the interference-free environment output by the quality channel decision maker unit of Jiahua Channel ( As the input parameter of the transmit power adjustment amount estimator, the output satisfies the uth user service reliability setting and the block error rate in the interference environment. The minimum transmit power adjustment value of 0.1; each input parameter uses a bell-shaped attribution function, each bell-shaped attribution function is divided into three levels, a total of 27 fuzzy inference rules, and the output is corresponding to the u-th user in the interference environment. The minimum transmit power adjustment amount; the optimization target and condition of the transmit power adjustment amount estimator corresponding to the uth user can be expressed as: in the interference environment, the minimum power estimate corresponding to the uth user transfer transmitter Optimization goal is to optimize the assignment Restricted conditions: { CQI _{min, u} , CQI _{best, u} and }1 CQI _{min, u} 15 1 CQI _best,u 15-25dB 45dB +△ P _u { 20dBm}, P _u is the previous transmit power (dBm), and nUE is the number of indoor users. The indoor small cell intelligent deployment tandem control device constructed in the multi-user interference environment described in the seventh application scope of the application is divided into a full-duplex full-duplex and an orthogonal frequency-division multiplexed access device, wherein in a multi-user interference environment The transmit power adjustment amount estimator is designed according to the adaptive network fuzzy inference system, and the front part parameter and the post part parameter are trained and adjusted by the training data; in order to satisfy the block error rate in the interference environment For the demand of 0.1 and service reliability (SR _u ), the uth user target interference ratio threshold (SINR _{th, u} ) is defined as SINR _{th, u} = max{SNR _th (CQI _{min, u} ) + FM ( SR _u ), SNR _th (CQI _{best, u} )} (dB), the uth transmit power adjustment amount (ΔP _u ) is the target signal to interference ratio threshold (SINR _{th, u} ) and the user signal to interference ratio Measuring average The difference between the subtraction; the training data required by the transmit power adjustment estimator, from the service reliability (90%) and the cell edge channel quality index (1~15), the best channel quality index (1~15) The amount of transmission power adjustment corresponding to the average value of the signal-to-interference ratio. The self-optimized power control is constructed in a multi-user interference environment as described in claim 1 of the invention, which is constructed in a frequency division full-duplex and orthogonal frequency division multiplexing access indoor small-cell intelligent deployment control device. Protection mechanism to avoid indoor small cells When the self-optimized transmit power is adjusted, it interferes with the neighboring cell users, thus reducing the efficiency of the self-optimized deployment of the serial control device. The intelligent deployment serial control device inputs the average path loss of the user back, and then uses the indoor The path loss model estimates the distance between the client and the base station (d). If the client is within the radius (R), the self-optimized minimum transmit power control unit can be activated.