The chain circuit self-adaptive method of predicting based on Signal to Interference plus Noise Ratio during high speed downlink packet inserts
Technical field
The present invention relates to a kind of high speed downlink packet insert in based on the chain circuit self-adaptive method of Signal to Interference plus Noise Ratio prediction, be particularly related to a kind of 1.28Mcps of being applicable to TDD (Time Division Duplex, time division duplex) link adaptation scheme of predicting based on Signal to Interference plus Noise Ratio during the high speed downlink packet of HSDPA (High Speed Downlink Packet Access, high speed downlink packet inserts) system inserts.
Background technology
Begin from the R5 version of 3GPP (3rd Generation Partnership Project, three generations's partnership relation) technical standard, HSDPA as a kind of enhancement techniques of 3G mobile communication system formally by standardization.1.28Mcps TDD HSDPA has mainly adopted AMC (Adaptive Modulation andCoding, Adaptive Modulation and Coding), FCS (Fast Cell Selection, the fast cell selection), defeated, short TTI (the Transmission Time Interval of multi-antenna transmitting, Transmission Time Interval) (5ms) and HARQ (Hybrid Automatic Repeat Request mixes automatic repeat requests) wait some novel technology to improve the throughput (Throughput) of system.
AMC (Adaptive Modulation and Coding, Adaptive Modulation and Coding) belongs to a kind of link adaptation techniques, its basic principle is exactly the transmission rate according to the channel situation self-adapted adjustment system, when signal to noise ratio when higher, adopt higher order of modulation and code rate, otherwise promptly adopt lower order of modulation and code rate.
The realization of AMC will face several challenges.The combination of at first, too much modulation, encoding scheme (MCS) can increase the complexity of system.At 3GPP technical standard 25.321 (Technical SpecificationGroup Radio Access Network; Medium Access Control (MAC) protocolspecification V5.10.0.) in, stipulated clearly that 1.28Mcps HSDPA only adopts QPSK (Quadrature Phase Shift Keying, four phase shift keyings) and 16QAM (QuadratureAmplitude Modulation, quadrature amplitude modulation) two kinds of modulation systems, and the access capability of UE is divided into three classes, the transmission block size of each class in a TTI is defined as 64 kinds of selections, and this has reduced the complexity of system design effectively.Secondly, AMC is very responsive to the sum of errors time-delay of channel measurement.In order to select suitable modulation and code rate, the scheduling on upper strata must have accurately quality of channel to be understood.Otherwise might be because of having adopted too high modulation, code rate and cause frame error rate too high, or have adopted lower modulation, code rate and caused the loss of power system capacity.In the 1.28McpsHSDPA system, standard has been stipulated the size of recommending transmission block among modulation system that institute should adopt and each TTI by HS-SICH (Shared InformationChannel for HS-DSCH, the special-purpose information channel of sharing of HS-DSCH) channel to the base station by UE.Because the physical channel of mobile communication became when having, even UE is accurately to the measurement of channel, but because UE to the hysteresis quality that the base station report is produced, can produce the unmatched phenomenon of modulation, encoding scheme and channel that is adopted equally.At this phenomenon, the notion of channel estimating has been carried, and the theoretical foundation of channel estimating is that channel is a continually varying, and the phenomenon that can occur suddenling change not that is to say between the continually varying channel sample point it is the (see figure 1) of being correlated with.And relevant power is relevant with the Doppler frequency shift size of system, that is to say when the speed of a motor vehicle (corresponding Doppler frequency shift is bigger) when accelerating slowly, correlation between the channel sample point in same sampling interval is by dying down by force, obviously when the speed of a motor vehicle is very fast, the precision of channel estimating becomes and is difficult to guarantee, so all Forecasting Methodologies all can only satisfy the performance requirement of system in certain vehicle speed range.Present channel estimating scheme all has complexity than the ivory-towered shortcoming of physical channel model higher or that adopted.Document " Prediction of Mobile Rado Channels " for example, T.Ekman, Licentiate thesis, Department of Signals and Systems, Uppsala University ofTechnology, Dec.2000. and " Prediction of Fast Fading Parameters by Resolvingthe Interference Pattern; " T.Eycoez, A.Duel-Hallen, H.Hallen, Signals, Systems﹠amp; Computers, Conference Record of the Thirty-First Asilomar Conference on, Volume:1,2-5 Nov 1997 Page (s): each channel coefficients at the tapped delay channel model among the 167-171 is taked AR (Autoregressive, autoregression) or ARMA (Autoregressive-movingAverage, autoregressive moving average) modeling, and approach by Wiener filtering or Kalman filtering, the last low pass filter that is limited to Doppler frequency by band again carries out smoothly, this method needs the single order and the second-order statistic of estimated channel tap coefficient, and to all to estimate each tap coefficient, computation complexity is very high, especially to the TDD system, because what adopt is the mode of inserting pilot tone, this final convergence to Wiener filtering or Kalman filtering has caused very large difficulty.Document " Channelprediction for adaptive coded modulation in Rayleigh fading; " G.E. ien.H.Holm, and K.J.Hole, in Proc.European Signal Processing Conference (EUSIPCO), (Toulouse, France), Sept.2002 and " An analysis of pilot symbolassisted modulation for Rayleigh fading channels; " J.K.Cavers, IEEETransactions on Vehicular Technology, vol.40, pp.686-693, the prerequisite that November 1991 carries out channel estimating is to have supposed that the auto-correlation function of channel tap is known, and this obviously is infeasible in the system of reality.
Summary of the invention
The present invention changes from the SINR (Signal to Interference plus Noise Ratio) of prognoses system, by calculating the slope of Signal to Interference plus Noise Ratio localized variation, predict following Signal to Interference plus Noise Ratio with this, and according to the result that forecasting institute gets determine the base station coding, the modulation scheme that should adopt, in certain vehicle speed range, obtain the raising of throughput of system thus by the not high prediction scheme of complexity.
For achieving the above object, based on the chain circuit self-adaptive method of Signal to Interference plus Noise Ratio prediction, it mainly comprised following steps during high speed downlink packet provided by the invention inserted:
Step 1, at 3GPP technical standard 25.321, each fast size of transmission (TB size) of stipulating among the V5.10.0. is determined corresponding with it Signal to Interference plus Noise Ratio under the Gaussian channel by emulation, the criterion of its foundation is that Signal to Interference plus Noise Ratio is minimum on guaranteeing that frame error rate is less than 0.01 basis; This will be as the foundation of selecting the transmission block size in the later step according to Signal to Interference plus Noise Ratio;
Step 2, utilize current TTI (Transmission Time Interval, Transmission Time Interval) TS0, HS-DSCH (High Speed Downlink Shared Channel, high speed descending sharing channel) midamble of occupied time slot (pilot tone in the time slot) part and DwPTS (Downlink Pilot Time Slot, descending pilot frequency time slot) estimate the Signal to Interference plus Noise Ratio at these time slots;
Step 3 calculate the slope that the Signal to Interference plus Noise Ratio on these time slots changes, and weighted average is to obtain the average Signal to Interference plus Noise Ratio of current TTI;
Step 4, according to the Signal to Interference plus Noise Ratio of the slope prediction next frame that is obtained, utilize then corresponding relation between Signal to Interference plus Noise Ratio and the transmission block size obtain next TTI modulation, the encoding scheme that should adopt;
Computation complexity of the present invention is low, it is by calculating the slope of Signal to Interference plus Noise Ratio localized variation, predict following Signal to Interference plus Noise Ratio with this, and according to the result that forecasting institute gets determine the base station coding, the modulation scheme that should adopt, in certain vehicle speed range, obtain the raising of throughput of system thus by the not high prediction scheme of complexity.
Description of drawings
Fig. 1 PA3 channel, DPCH (DPCH) code channel number is 2, DPCH0 (special-purpose disturb physical channel) code channel number be 8 be Signal to Interference plus Noise Ratio over time, wherein transverse axis n represents n times of slot time.
Fig. 2 .a is a TTI structural representation that adopts four descending time slot HS-DSCH; Fig. 2 .b is the structural representation of a time slot;
Fig. 3 then is realization flow figure of the present invention.
Embodiment
Following according to Fig. 2 .a, 2.b, Fig. 3 illustrates a better embodiment of the present invention.Can further understand content of the present invention accurately.
Based on the chain circuit self-adaptive method of Signal to Interference plus Noise Ratio prediction, it comprised the steps (as shown in Figure 3) during a kind of high speed downlink packet provided by the invention inserted:
Step 1, by the mapping table between emulation formulation Signal to Interference plus Noise Ratio and the transmission block size, the criterion of its foundation is that Signal to Interference plus Noise Ratio is minimum on guaranteeing that certain frame error rate frame error rate is less than 0.01 basis;
Step 2, estimate m (m=1,2,3 ...) individual TTI goes up Signal to Interference plus Noise Ratio SINR3, SINR4, SINR5, the SINR6 (as Fig. 2 .a, shown in Fig. 2 .b) of different time-gap;
Wherein SINR3, SINR4, SINR5 and SINR6: represent that m TTI goes up midamble according to TS3, TS4, TS5, TS6 and estimate the Signal to Interference plus Noise Ratio that obtains;
Step 3,
At first, comprise step 31, judge whether current TTI (Transmission Time Interval) is first; Be whether m equals 1; In this way, then carry out step 32, as not being then to leap to step 33;
Step 32: calculate m (m=1,2,3 ...) Signal to Interference plus Noise Ratio slope (step 321) and average Signal to Interference plus Noise Ratio (step 322) on the individual TTI, can be expressed as respectively with formula (1) and (2) respectively:
SINR
1=(SINR3+SINR4+SINR5+SINR6)/4 (2)
Wherein:
S
1: the 1st TTI goes up the Signal to Interference plus Noise Ratio slope that obtains;
t
34, t
45, t
56: represent the time interval between TS3 and the TS4 respectively, the time interval between TS4 and the TS5, the time interval between TS5 and the TS6; Please note that the starting point in these time intervals and terminal point all are that center with time slot is as the reference point;
SINR
1: the 1st TTI goes up the average Signal to Interference plus Noise Ratio that obtains.
Step 33: calculate m (m=2,3 ...) Signal to Interference plus Noise Ratio slope (step 331) and average Signal to Interference plus Noise Ratio (step 332) on the individual TTI, can be expressed as respectively with formula (3) and (4) respectively:
SINR
m=(SINR3+SINR4+SINR5+SINR6)/4 (4)
Wherein:
S
M-1: m-1 TTI goes up the Signal to Interference plus Noise Ratio slope that obtains;
S
m: m TTI goes up the Signal to Interference plus Noise Ratio slope that obtains;
t
34, t
45, t
56: represent the time interval between TS3 and the TS4 respectively, the time interval between TS4 and the TS5, the time interval between TS5 and the TS6; Please note that the starting point in these time intervals and terminal point all are that center with time slot is as the reference point;
F: value is the forgetting factor between 0 to 1.
SINR
m: m TTI goes up the average Signal to Interference plus Noise Ratio that obtains.
Step 4 is according to S
mAnd SINR
mCan predict the SINR of m+1 TTI
M+1
SINR
m+1=S
m*t
TTI+SINR
m (3)
T wherein
TTIRepresent the time that a TTI continues, SINR
M+1Represent that m+1 TTI goes up the average Signal to Interference plus Noise Ratio that obtains;
Step 5 according to the form that the first step obtains, is determined the transmission block size of m+1 TTI;
Step 6 judges that whether transmission finishes, and as not, then jumps to step 2; In this way, then finish computational process (step 7).
The present invention mainly is applicable to 1.28Mcps TDD (Time Division Duplex, time division duplex) HSDPA (High Speed Downlink Packet Access, high speed downlink packet inserts) system, but it also can be used for 3GPP (3rd Generation Partnership Project, three generations's partnership relation) in FDD (Frequency Division Duplex, Frequency Division Duplexing (FDD)) and the 3.84Mcps TDD HSDPA system.At the VA30 channel, and 1.4Mbps UE transmittability,
(
The expression down receiving signal is at the power spectral density of UE end, I
OcRepresent the descending band limited white noise power spectral density of measuring gained at UE antenna place) time when not adopting channel estimating the throughput of emulation gained be 200kbps, be 234kbps and adopted after the present invention, obtained obvious effects.