CN106304381A - A kind of LTE system resource regulating method - Google Patents

Abstract

The present invention provides a kind of LTE system resource regulating method, relate to the dispatching method being specially adapted for real time business, it it is dispatching method based on marginal utility principle, step is: the parameter of QoS guarantee time delay be incorporated in marginal utility principle, obtains reflecting business S type function of delay sensitive change in queue；The signal received according to user i, calculates user i at the n-th TTI at the channel gain of all carrier waves；Calculate the user i Signal to Interference plus Noise Ratio at each subcarrier；The Signal to Interference plus Noise Ratio of all carrier waves in occupied each RB is converted into the effective signal-to-noise ratio of this RB；Calculate the transfer rate of user i；Calculate user's Mean Speed at (n-1)th TTI；Calculate user i in the n-th TTI and, in the scheduling priority factor of each RB, select the business of the user of TP top priority；Update user's Mean Speed at the n-th TTI, and repeat said process.Instant invention overcomes the deficiency that may not apply to real time business in existing method, the scope of application is wider, and systematic entirety can promote.

Description

A kind of LTE system resource regulating method

Technical field

Technical scheme relates to being specially adapted for LTE wireless communication technology field, specifically a kind of LTE system resource regulating method.

Background technology

Within 2004,3GPP starts Long Term Evolution project LTE (Long Term Evolution), it is intended to pay close attention to The technological evolvement problem of wave point and wireless networking framework.2013, China Mobile, CHINAUNICOM, in State's telecommunications all obtains TD-LTE licence plate, and then in 2015, CHINAUNICOM and China Telecom obtain FDD-LTE Licence plate, so far commercial for 4G gate is formally opened, and Consumer's Experience also becomes operator 4G commercialization success or not Essential condition.

Compared to 3G technology, LTE system is obviously improved in terms of power system capacity, transmission speed, at net Simplified in network system architecture, can be supported that multiple different QoS (Quality of Service) requires class The business of type, including VoIP business, Streaming Media, game, HTTP/FTP, e-mail etc., to user with more Good experience, simultaneously these business or pair time Ductility Requirement higher or higher to bit error rate requirement, this is right The scheduling of Radio Resource it is also proposed higher requirement.

In LTE system, being attached by Uu interface between base station and terminal, wireless resource scheduling occurs At medium education (MAC) layer of base station, being realized by wireless resource scheduling device, it can be descending common Enjoy channel (DL-SCH) and Uplink Shared Channel (UL-SCH) distribution physical layer resources, DL-SCH and UL-SCH Different schedulers is used to be scheduling respectively.The performance indications that general scheduling of resource considers have throughput of system, Fairness between user, time ductility etc..

The current user accessing wireless network and business get more and more, and Radio Resource is nervous and channel quality is unstable Fixed, the performances such as throughput of system, user fairness, service delay, or energy can not be taken into account when scheduling resource Take into account but overall performance is the highest, cause a lot of in the case of be not well positioned to meet user's QoS requirement, reach Less than good Consumer's Experience.

In the last few years, some time domains and frequency domain resource scheduling method are suggested in succession, and more classical have three classes: The first kind is polling dispatching, and the method is exactly that base station provides service, user for community user in a certain order Between fairness be best, but owing to not accounting for channel quality, cause system entire throughput very poor； Equations of The Second Kind is max carrier to interference scheduling, and channel condition quality as overriding concern condition, is caused channel by the method The user of inclement condition may can not get scheduling of resource always, and between user, fairness is worst；3rd class is that ratio is public Heibei provincial opera degree, the method is the compromise of polling dispatching and max carrier to interference dispatching method, is ensureing that all users' is flat All ensure that while volume of transmitted data maximum that not havinging certain user can not get the phenomenon of scheduling, the party for a long time Method is disadvantageous in that the performance of system entirety is the highest, and may not apply to real time business.

Unlimited resources scheduling problem is proved to be a NP-hard problem, but improves LTE system overall performance, Utilize limited Radio Resource to realize the inherent value of Radio Resource the most to greatest extent, be currently to need badly The problem solved.

Summary of the invention

The technical problem to be solved is: provide a kind of LTE system time domain and frequency domain resource scheduling side Method, considers the factors such as QoS time delay, throughput of system, fairness, transfer rate, channel condition, In LTE system, the dispatching priority grade of each user, selects the business of the highest user of priority, thus enters Row scheduling of resource.This method is on the premise of ensureing certain handling capacity, by the marginal utility principle in economics Make systematic entirety energy optimum, between user, between fairness and throughput of system, find a kind of preferably balance. Compared with the conventional method, Time Delay of Systems reduces, and the applicable scope of business is wider, and systematic entirety can carry Rise, overcome existing method and may not apply to the deficiency of real time business.

The present invention solves this technical problem and be the technical scheme is that time domain and frequency in a kind of LTE system of offer Territory resource regulating method, is LTE resource regulating method based on marginal utility function, and its step is as follows:

Step one, joins the parameter of QoS guarantee time delay in the middle of marginal utility principle, obtains reflecting industry Business is the S type function of delay sensitive change, the dispatching sequence of the business for waiting in line in queue in queue Carry out regular, make the business having QoS delay requirement in queue and the business near time delay thresholding obtain relatively authority Weight；

Step 2, the signal received according to user i, calculate user i at the n-th TTI (Transmission Time Interval, Transmission Time Interval) at the channel gain of all carrier waves；

Step 3, calculates the user i Signal to Interference plus Noise Ratio at each subcarrier；

Step 4, is converted into this RB by the Signal to Interference plus Noise Ratio of all carrier waves in occupied each RB (Resource Block) Effective signal-to-noise ratio；

Step 5, according to user i at the Signal to Interference plus Noise Ratio of each subcarrier, calculates the transfer rate of user i DRC_i(n)；

Step 6, calculates user's Mean Speed at (n-1)th TTI according to the transfer rate of user；

Step 7, calculates user i in the n-th TTI and, in the scheduling priority factor of each RB, selects the highest The business of the user of priority level；

Step 8, updates user's Mean Speed at the n-th TTI, and repeats said process.

Accompanying drawing explanation

The present invention is further described with embodiment below in conjunction with the accompanying drawings.

Fig. 1 is the schematic flow sheet of the inventive method.

Fig. 2 is to reflect business S type function figure of delay sensitive change in queue in the inventive method.

Detailed description of the invention

The fixed-bandwidth arranging system is B, and Resource Scheduler connects i terminal, and in the method, business is n-th The parameters such as handling capacity before individual TTI, queue length can be obtained by rlc layer.

Step one, such as Fig. 2, joins the parameter of QoS guarantee time delay in the middle of marginal utility principle, obtains energy Enough reflection business S type function U of delay sensitive change in queue_i(n):

$U_i\left(n\right)=\frac{1}{1+e^{\mathrm{\δ}(L-l_i\left(n\right))}}$

Wherein, L is the patient maximum delay of user i；l_iN () is that the business of user i is when the queue of relief area Prolong；δ is constant, span [0,1], can adjust according to the size of delay sensitive in actual schedule Joint, and δ is the biggest, the gradient of S type function is the steepest.

Step 2, the signal received according to user i, calculate user i at the n-th TTI at all carrier waves Channel gain

${\mathrm{Gain}}_{i,k}\left(n\right)={10}^{\left[\frac{P_{\mathrm{ti},k}\left(n\right)}{10}\right]+\left[\frac{{\mathrm{\ξ}}_{i,k}\left(n\right)}{10}\right]+\left[\frac{{\mathrm{mapth}}_{i,k}\left(n\right)}{10}\right]}$

Wherein, p_{Li, k}N () is the path loss of r subcarrier；ξ_{I, k}N () is produced by the r subcarrier shadow fading Raw gain；mpath_{I, k}N () is gain produced by the r subcarrier multipath fading.

Step 3, utilizes with the carrier channel gain obtained, and calculates user i and makes an uproar the letter of each subcarrier is dry Compare SINR_{I, k}(n):

${\mathrm{SINR}}_{i,k}\left(n\right)=\frac{P_k*{\mathrm{Gain}}_{i,k}\left(n\right)}{I+N_o}$

Wherein, N_oIt is noise power, P_kReferring to the base station transmitting power at r subcarrier, I is between community Jamming power；

Step 4, is converted into effective noise of this RB by the Signal to Interference plus Noise Ratio of all carrier waves in occupied each RB Compare SNR_{I, j}(n):

${\mathrm{SNR}}_{i,j}\left(n\right)=\mathrm{EESM}({\mathrm{SINR}}_{i,k}\left(n\right),\mathrm{\σ})=-\mathrm{\σ}\·\left[\frac{1}{M}\mathrm{\Σ}{e}^{-\frac{\mathrm{SINRi},k\left(n\right)}{\mathrm{\σ}}}\right]$

Wherein, M is the subcarrier number included in a Resource Block, and for different frame structures, M value is the most not With, according to common prefix, M=7；If frame structure uses expanded prefix, M=6；σ is the alterable factor, Carry out value according to modulation coding mode, specifically can inquire about table 1:

Table 1

Step 5, according to user i at the Signal to Interference plus Noise Ratio SINR of each subcarrier_{I, k}N (), calculates user i's Transfer rate DRC_i(n):

${\mathrm{DRC}}_i\left(n\right)=\underset{j}{\mathrm{\Σ}}{\mathrm{DRC}}_{i,j}\left(n\right)$

DRC_{I, j}(n)=log₂(1+SINR_{I, j}(n))

Wherein, DRC_{I, j}N () is user i transfer rate in jth RB；

Step 6, calculates Mean Speed R of (n-1)th TTI according to the transfer rate of user i_i(n):

$R_i\left(n\right)=(1-\frac{1}{n_c})\·R_i(n-1)+\frac{1}{n_c}{\mathrm{DRC}}_i\left(n\right)$

Wherein, n_cFor updating time span, typically it is required to meet change and the service delay requirement covering rapid fading；

Step 7, calculates user i in the n-th TTI in the scheduling priority factor of each RB, and selects The business of the user i of TP top priority, scheduling priority factor is:

$P_i\left(n\right)=\frac{{\mathrm{DRC}}_{i,j}\left(n\right)}{R_i(n-1)}\·U_i\left(n\right)\·Q_i{\left(n\right)}^{-1}\·S_i{\left(n\right)}^{-1}$

Wherein, Q_i(n), S_iN () is the rate factor for controlling different business speed, and

$Q_i\left(n\right)=\min [{\left(\frac{R_i(n-1)}{R_L}\right)}^{\mathrm{\α}},1]$

$S_i\left(n\right)=\max [{\left(\frac{R_i(n-1)}{R_H}\right)}^{\mathrm{\β}},1]$

Wherein, R_i(n-1) it is user i transfer rate meansigma methods before the n-th TTI；R_HIt is user QoS The upper limit of rate requirement；R_LIt it is the lower limit of user's QoS rate requirement；α, β are all higher than 1, are function parameters, It is used for controlling the convergence rate of owner priority.If the transfer rate of user is less than the lower limit of qos requirement, i.e. R_i(n-1) ＜ R_L, then Q_i(n) ＜ 1, S_iN ()=1, the dispatching priority of this user can be elevated；If the biography of user Defeated speed is higher than the upper limit of qos requirement, i.e. R_i(n-1) ＞ R_H, then Q_i(n)=1, S_i(n) ＞ 1, the tune of this user Degree priority can be lowered, thus ensures the fairness of scheduling between user.

Step 8, updates user's Mean Speed at the n-th TTI, and repeats said process.

Claims (6)

1. a LTE system resource regulating method, it is characterised in that be LTE resource based on marginal utility principle Dispatching method, its step is as follows:

Step one, joins the parameter of QoS guarantee time delay in the middle of marginal utility principle, obtains reflecting industry Business is the S type function of delay sensitive change, the dispatching sequence of the business for waiting in line in queue in queue Carry out regular, make the business having QoS delay requirement in queue and the business near time delay thresholding obtain relatively authority Weight；

Step 2, the signal received according to user i, calculate user i at the n-th TTI at all carrier waves Channel gain；

Step 3, calculates the user i Signal to Interference plus Noise Ratio at each subcarrier；

Step 4, is converted into effective noise of this RB by the Signal to Interference plus Noise Ratio of all carrier waves in occupied each RB Ratio；

Step 5, according to user i at the Signal to Interference plus Noise Ratio of each subcarrier, calculates the transfer rate of user i；

Step 6, calculates the Mean Speed of (n-1)th TTI according to the transfer rate of user；

Step 7, calculates user i in the n-th TTI and, in the scheduling priority factor of each RB, selects the highest The business of the user of priority level；

Step 8, updates user's Mean Speed at the n-th TTI, and repeats said process.

2. according to a kind of LTE system resource regulating method described in claim 1, it is characterised in that: reflection is used The S type function of family business delay sensitive change in queue

$U_i\left(n\right)=\frac{1}{1+e^{\mathrm{\δ}(L-l_i\left(n\right))}}$

Wherein, L is the patient maximum delay of user i；l_iN () is the business team in relief area of user i Row time delay；δ is constant, span [0,1], can enter according to the size of delay sensitive in actual schedule Row regulation, and δ is the biggest, the gradient of S type function is the steepest.

3. according to a kind of LTE system resource regulating method described in claim 1, it is characterised in that: user i Transfer rate at the n-th TTI

${\mathrm{DRC}}_i\left(n\right)=\underset{j}{\mathrm{\Σ}}{\mathrm{DRC}}_{i,j}\left(n\right)$

Wherein, DRC_{I, j}N () is user i transfer rate in jth RB.

4. according to a kind of LTE system resource regulating method described in claim 3, it is characterised in that: user i Transfer rate in jth RB

DRC_{I, j}(n)=log₂(1+SINR_{I, j}(n))

Wherein, SINR_{I, k}N () is the user i Signal to Interference plus Noise Ratio at each subcarrier.

5. according to a kind of LTE system resource regulating method described in claim 4, it is characterised in that: user i Signal to Interference plus Noise Ratio at each subcarrier

${\mathrm{SINR}}_{i,k}\left(n\right)=\frac{P_k*{\mathrm{Gain}}_{i,k}\left(n\right)}{I+N_o}$

Wherein, N_oIt is noise power, P_kReferring to the base station transmitting power at r subcarrier, I is community Between jamming power, Gain_{I, k}N () is user i at the n-th TTI at the channel gain of all carrier waves.

6. according to a kind of LTE system resource regulating method described in claim 5, it is characterised in that: user i At the n-th TTI at the channel gain of all carrier waves

Wherein, p_{Li, k}N () is the path loss of r subcarrier；ξ_{I, k}N () is the r subcarrier shadow fading Produced gain；mpath_{I, k}N () is gain produced by the r subcarrier multipath fading.