CN1870475B - Channel measurement method based on laterally spaced adaptive filter - Google Patents

Abstract

The present invention relates to a channel measurement method based on a horizontal spacing adaptive filter, comprising the steps of: A, using the first horizontal spacing adaptive filter at the first predetermined time t to measure the channel quality value at time t+2; B, Utilize the second horizontal spacing adaptive filter to measure the channel quality value of time t+3 at the first moment t+1; C, the first horizontal spacing adaptive filter and the second horizontal spacing adaptive filter take turns to measure the moment Channel quality values for t+2n and t+2n+1. The present invention measures the performance of the channel in turn by measuring different delay frame numbers and adopting different numbers of horizontally spaced adaptive filters, so that the horizontally spaced adaptive filter has a better measurement function in keeping the adaptive filter in a low-speed environment Under the premise of the premise, it has a better smoothing function in the medium and high speed mobile environment. The method reduces the impact of time delay on the system, thereby improving system performance.

Description

Channel measurement method based on laterally spaced adaptive filter

technical field

The present invention relates to a channel measurement method of mobile communication technology, and more specifically, relates to a channel measurement method based on a laterally spaced adaptive filter.

Background technique

The current protocol version (Rlease 5) stipulates that in the HSDPA of the TD-SCDMA system, there must be at least 9 time slots between the last moment when the UE receives data and the moment when the channel quality information is sent to the network. In addition to the reason of TD-SCDMA frame structure, this kind of regulation mainly takes into account the delay of terminal processing data.

In the high-speed mobile channel, due to the influence of Doppler and multipath, the channel changes rapidly, and the channel quality at the time of measurement will be greatly different from the channel quality at the time when the UE sends channel quality information to the network due to delay, such as in For ITU-VA30 channels, a delay of 10ms will cause a 3-4dB difference in channel quality SINR, which will cause a loss of more than 30% of the throughput of the HSDPA link.

As shown in Figure 1, it is a schematic diagram of the simulation performance comparison of 10ms interval delay and no delay throughput on the VA30 channel. On the ITU-VA30 channel of the TD-SCDMA system, the throughput simulation comparison of the HSDPA link with the theoretical maximum peak value of 1400k bits/s at no delay and delay of 10ms. As can be seen from Figure 1, due to the existence of delay, the system performance will drop sharply.

How to reduce the impact of delay on system performance is a difficult problem that the field of mobile communication has been trying to solve. At present, in relevant literature, it is proposed to use a horizontal measurement adaptive filter to measure channel conditions. There are many kinds of horizontal interval adaptive filters, such as Kalman adaptive filter, recursive least squares adaptive filter and Minimum mean square error filter (LMS, Least mean square), etc. Considering that the LMS filter is easy to implement and has low computational complexity, the following uses the LMS filter as an example to illustrate. Refer to FIG. 2 and FIG. 3 for details. FIG. 2 is a schematic diagram of the time sequence of data frames received by the end user; FIG. 3 is a schematic diagram of the principle of the minimum mean square error filter. Gaps due to latency are reduced by measuring the channel. Its implementation process is as follows:

First, assuming that the measured value is 2 frames (10 ms) different from the sending time, when the terminal measures the channel condition at time t, it can measure the channel quality at time t+2 according to the channel condition information before this time.

进行线性运算，并将运算所得的t+2时刻信道测量值

输出。同样，对于t+3时刻，可以通过输入为u(t+1)，u(t)，.......，u(t-M+2)，以及调整更新后的抽头权值

与滤波器的输入值u(t+1)，u(t)，.......，u(t-M+2)进行线性运算来测量。In the transversal filter, if the channel quality SINR value measured at time t is represented by u(t), then the transversal filter input with M taps is u(t), u(t-1),... ..., u(t-M+1), these inputs form a multi-dimensional space (expressed by φ(t)), then the channel quality value at time t+2 is measured by express. The tap weights of the adaptive filter are used express. The minimum mean square error filter is shown in Figure 3. The tap weight is adjusted by the difference e(t+2) between the measured value and the expected value. After adjusting the tap weight, the filter input values u(t), u(t-1), u(t-2) and u(t-3) and tap weights

Carry out linear operation, and calculate the channel measurement value at time t+2

output. Similarly, for time t+3, you can input u(t+1), u(t), ..., u(t-M+2), and adjust the updated tap weight

It is measured by performing linear operations with the input values u(t+1), u(t), ..., u(t-M+2) of the filter.

According to the set delay condition, the expected value at time t+2 should be obtained at time t+4. In this way, the adjustment of tap weights will be delayed and cannot be used to measure the channel quality at time t+3. But for the convenience of research, we still assume that the expected value of this value at time t+2 can be obtained in time to measure the channel quality at time t+3.

From the above analysis, it can be known that using the two-step LMS measurement filter performance in the continuous measurement described in Figure 2 can effectively improve the system performance in low-speed mobile channels (such as in ITU-PA3 channels), but in high-speed mobile conditions, it will decrease. Throughput is lower than that obtained without the measurement filter. Because, in the low-speed mobile state, the channel change of each frame is slow, and the correlation of data between frames is large, so the measured value of the next two steps of continuous measurement is close to the real value, which can improve the system performance. However, in the high-speed mobile state, due to the rapid channel change, the inter-frame data correlation is small, and the value at time t and time t+1 has a large change. While in Figure 3, the tap weights are used Adjusted for the difference from the measured expected value. Therefore, when measuring the value at time t+3, the correlation between the adjusted tap weight coefficient and the measured value at time t+1 is small, resulting in a large error in the measured value at time t+3. By analogy, the error of the measured value has a great influence on the throughput of HSDPA.

Contents of the invention

The technical problem solved by the present invention is to provide a channel measurement method based on an adaptive filter, which measures the quality of the channel in turn through a laterally spaced adaptive filter, and solves the problem of inaccurate measurement in the state of high-speed movement in the prior art The problem of channel quality, so as to reduce the impact of delay on the system in the high-speed mobile environment.

In order to solve the above problems, the present invention provides a channel measurement method based on a horizontally spaced adaptive filter, wherein the horizontally spaced adaptive filter is composed of two spaced valued horizontally adaptive filters, comprising the steps of:

A. Measuring the channel quality value at time t+2 by using the first horizontal interval adaptive filter at the first predetermined time t;

B. Measuring the channel quality value at time t+3 by using the second laterally spaced adaptive filter at time t+1;

C. The first horizontal spacing adaptive filter and the second horizontal spacing adaptive filter measure channel quality values at time t+2n and t+2n+1 in turn.

Described step A specifically comprises:

Adjusting the tap weights of the first horizontal interval adaptive filter according to the preset tap signal difference e'(t) at the first predetermined moment t;

The channel quality measurement value at time t+2 is obtained after linear operation is performed on the adjustment result of the tap weight and the input value of the first horizontally spaced adaptive filter.

The tap signal difference e'(t) is calculated according to the difference between the channel measurement value at time t and the channel expectation value at time t measured at time t-2.

The channel measurement value at time t measured at time t-2 is measured according to the channel quality value at time t and the even-numbered time before it as the input value of the filter.

The formula for adjusting the tap weights of the first horizontal interval adaptive filter is:

ω'(t)=ω(t-2)+μ(t)e'(t)U(t)

Among them, ω(t-2) is the tap weight of the first horizontally spaced adaptive filter at time t-2, μ(t) is the adjustment step size of the minimum mean square error LMS filter, U(t) is the time t Input value for the first horizontally spaced adaptive filter tap.

The value of μ(t) can be a constant or a function value related to the input vector U(t).

The channel measurement value at time t+2 after the linear operation satisfies the equation:

$\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; p</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; \&phi;</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>$

Wherein, the ω'(t) is the tap vector of the first horizontally spaced adaptive filter, and the U(t) is the input value of the tap of the first horizontally spaced adaptive filter at time t.

Described step B specifically comprises the steps:

Adjust the tap weights of the second horizontal interval adaptive filter according to the preset tap signal difference e"(t+1) at time t+1;

The channel quality measurement value at time t+3 is obtained after linear operation is performed on the adjustment result of the tap weights and the input value of the second horizontally spaced adaptive filter.

The tap signal difference e″(t+1) is calculated according to the difference between the measured channel value at time t+1 and the expected channel value at time t+1 measured at time t-1.

The channel measurement value at time t+1 measured at time t-1 is measured according to the channel quality value at time t+1 and the odd-numbered time before it as the input value of the filter.

The formula for adjusting the tap weights of the second horizontal interval adaptive filter is: ω″(t+1)=ω(t-1)+μ(t+1)e″(t+1)U(t +1)

Among them, ω(t-1) is the tap weight of the second horizontally spaced adaptive filter at time t-1, μ(t+1) is the adjustment step size of the LMS filter, and U(t+1) is the time t+1 Input value for the second horizontally spaced adaptive filter tap.

The value of μ(t+1) can be a constant or a function value related to the input vector U(t+1).

The measured channel value at time t+3 after the linear operation satisfies the equation:

$\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; p</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; \&phi;</span>}}^{<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}^{<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

Wherein, the ω″(t+1) is the tap vector of the second horizontally spaced adaptive filter, and the U(t+1) is the input value of the tap of the second horizontally spaced adaptive filter at time t+1.

The first predetermined time t is the time when the channel quality value is actually received.

Compared with the prior art, the present invention has the following beneficial effects: the present invention measures the performance of the channel in turn through different measurement delay frame numbers and adopts different numbers of horizontal interval adaptive filters, so that the horizontal interval adaptive filter is It has a better smoothing function in medium and high-speed mobile environments, and it reduces the impact of time delay on the system in high-speed mobile environments, thereby improving system throughput and frequency band utilization; the present invention can also be used in low-speed environments It has the function of measuring the channel. Therefore, the channel quality measurement method of the present invention can not only reduce the loss caused to the system due to time delay to a certain extent, improve the performance of the system, but also has the characteristics of low computational complexity and easy implementation. It is a practical Optimize channel measurement method.

Description of drawings

Fig. 1 is a schematic diagram of the simulation performance comparison between 10 ms interval delay and no delay throughput on the VA30 channel in the prior art;

FIG. 2 is a schematic diagram of a time sequence of data frames received by a terminal user in the prior art;

Fig. 3 is the schematic diagram of the principle of the minimum mean square error filter in the prior art;

Fig. 4 is the flow chart of the channel measurement method that adopts horizontally spaced adaptive filter in the present invention;

FIG. 4A is a flow chart of measuring the channel quality value at time t+2 by using the first horizontally spaced adaptive filter at time t in the present invention;

FIG. 4B is a flowchart of measuring the channel quality value at time t+3 by using the second laterally spaced adaptive filter at time t+1 according to the present invention;

FIG. 5 is a schematic diagram of performance comparison of various schemes under a low-speed mobile channel (such as an ITU-PA3 channel);

Fig. 6 A is the schematic diagram of performance comparison after HSDPA link adopts 4 kinds of processing methods described in Fig. 6 under medium-speed mobile channel (ITU-VA30);

FIG. 6B is a schematic diagram of performance comparison of the HSDPA link under the high-speed mobile channel (ITU-VA120) after adopting the four processing methods described in FIG. 6 .

Detailed ways

In a cellular mobile communication system, the wireless channel is a multipath time-varying channel (including propagation loss, fast fading, slow fading, and interference changes, etc.), so the quality of the received signal is also a time-variant affected by channel conditions. In order to improve system performance and overcome the impact of channel time variation on system performance, link adaptive technology can be used.

At present, in the wireless communication system, the link adaptive technology mainly adopts two working modes. The first mode is the power adaptive mode, and the transmitting end changes the transmission power of the transmitted data to adapt to the change of the channel condition; In this way, the sender adapts to channel changes by changing the data transmission code rate and modulation mode.

The AMC technology is a key technology that determines the performance of HSDPA, and how the Node B accurately obtains the information of the transmission channel condition becomes an important factor that determines the system performance. Normally, the Node B bases on the measurement channel information provided by the UE (such as sending the recommended transport block size to the Node B), as well as its own channel resources, sending packet scheduling algorithms, power control information provided by the DPCH channel, and the response ACK fed back by the UE /NACK information to determine the target UE for sending data, the transmission format MCS and the transmission block size, among which the channel condition information provided by the UE is the main determining factor.

Especially in the HSDPA of the TD-SCDMA system, there needs to be at least 9 time slots between the last moment when the UE receives data and the moment when the channel quality information is sent to the network. In order to reduce the impact of delay on the system, this The invention provides a channel measurement method using a lateral interval adaptive filter. By taking values at intervals, the channel measurement method of the present invention has a better smoothing function in medium and high-speed mobile environments, and it reduces the The impact of time delay on the system in the mobile environment, thereby improving the throughput and frequency band utilization of the system; in addition, the present invention can also have the function of measuring channels in a low-speed environment.

The present invention will be further described below in conjunction with the accompanying drawings.

Please refer to Fig. 4, Fig. 4A and Fig. 4B, which are respectively the flowchart of the channel measurement method using the horizontal spacing adaptive filter in the present invention, and the present invention uses the first horizontal spacing adaptive filter to measure the time t+2 at time t The flow chart of the channel quality value of and the flow chart of the present invention measuring the channel quality value of time t+3 by using the second laterally spaced adaptive filter at time t+1.

The channel measurement method using the horizontally spaced adaptive filter includes:

Step 10: Measuring the channel quality value at time t+2 by using the first horizontally spaced adaptive filter at the first predetermined time t;

Step 11: Measure the channel quality value at time t+3 by using the second laterally spaced adaptive filter at time t+1;

Step 12: The first horizontal spacing adaptive filter and the second horizontal spacing adaptive filter measure channel quality values at time t+2n and t+2n+1 in turn.

The step S10 specifically includes (as shown in FIG. 4A ):

Step 100: Adjust the tap weights of the first horizontally spaced adaptive filter according to the preset tap signal difference e'(t) at the first predetermined time t;

Step 101: Perform a linear operation on the adjustment result of the tap weight and the input value of the first horizontally spaced adaptive filter to obtain a channel quality measurement value at time t+2.

The step 11 specifically includes (as shown in Figure 4B):

Step 110: Adjust the tap weights of the second horizontally spaced adaptive filter according to the preset tap signal difference e"(t+1) at time t+1;

Step 111: Perform a linear operation on the adjustment result of the tap weight and the input value of the second horizontally spaced adaptive filter to obtain a channel quality measurement value at time t+3.

Wherein, the flow charts of the specific steps of step S10 and step S12 are shown in Fig. 4A and Fig. 4B.

表示。First, for the convenience of description, the present invention still takes the minimum mean square error filter in FIG. 3 as an example, that is, receives data at time t and measures data with a difference of 2 frames (time t+2). In order to measure the data at time t+2, there is a first horizontally spaced adaptive filter with M taps, and the inputs are u(t), u(t-2),......, u(t- 2M+2), that is, to take values at intervals, these inputs form a multi-dimensional space (expressed by φ′(2t)), then measure the channel quality value at time t+2 with

express.

与t+2时刻真实期望数值的差值e′(t+2)来调整第一横向间隔自适应滤波器的抽头权值 所述e′(t+2)是测量值与测量值之差，在本申请文件中把测量值作为期望值，通过两者的差来调制横向间隔自适应滤波器的抽头权值。When the user equipment UE is at time t+2, the measured It is sent to the network as the channel quality value at time t+2. and pass the measured value

The difference e'(t+2) from the real expected value at time t+2 is used to adjust the tap weight of the first horizontally spaced adaptive filter The e′(t+2) is the difference between the measured value and the measured value. In this application document, the measured value is taken as the expected value, and the tap weight of the lateral spacing adaptive filter is modulated by the difference between the two.

表示。In the same way, the present invention is provided with another M-tap second transverse adaptive filter. At time t+1, the user inputs u(t+1), u(t-1),  … , the data of u(t-2M+3) measures the channel quality at this moment, and these inputs form a multi-dimensional space (expressed by φ″(2t)), then the channel quality value at the moment t+3 is measured by

express.

When the UE is at time t+3, it will measure the resulting value It is sent to the network as the channel quality at time t+3. and use it later The difference e″(t+3) from the expected value at time t+3 to adjust the tap weight corresponding to the input data

The channel measurement method of the four-tap LMS filter is taken as an example to describe below. Assuming that the system measures the channel quality value u(t) at time t, the channel quality at time t+2 is measured using the first horizontally spaced adaptive filter. Its implementation steps include:

M0: filter according to the channel values u(t), u(t-2), u(t-4) and u(t-6) at time t and before t-2, t-4 and t-6 input value of the device.

M1: The channel measurement at time t will be measured at time t-2 Subtract it from u(t) to get the difference e'(t).

进行调整；M2: The tap weight of the LMS filter according to e′(t)

make adjustments;

M3: After adjusting the tap weights, the filter input values u(t), u(t-2), u(t-4) and u(t-6) and the tap weights perform linear operations;

输出。M4: The calculated channel measurement value at time t+2

output.

进行调整，其调整过程为：In the above step M2, the tap weights of the first horizontal interval adaptive filter according to e'(t)

To adjust, the adjustment process is:

Let the number of tap slots of the first horizontally spaced adaptive filter be That is, the tap vector is

The input values of the M taps at time t of the first horizontal interval adaptive filter are u(t),...,u(t-M+1), that is, the input vector is

U(t)=[u(t), u(t-1), . . . , u(t-M+1)] ^T . (2)

与时刻t的信道期望值u(t)之差，即The tap signal difference e'(t) is the channel measurement value at time t-2 measurement time t

The difference from the channel expectation value u(t) at time t, namely

The tap value can be adjusted according to e′(t) obtained from Equation 3

ω'(t)=ω(t-2)+μ(t)e'(t)U(t).  (4)

Among them, μ(t) is the adjustment step size of the LMS filter. According to different LMS algorithms, the value of μ(t) can be a constant or a function value related to the input vector U(t).

Therefore, after adjusting the tap weights, the filter input values u(t), u(t-2), u(t-4) and u(t-6) are compared with the tap weights Carry out linear operation (step M3), the process of its linear operation is:

则该值满足等式Let the measured value of the first transverse spacing adaptive filter at time t be

Then the value satisfies the equation

Here, the vectors ω'(t) and U(t) satisfy the above equations (1) and (2), respectively.

From the above calculation, the channel measurement value at time t+2 can be obtained

Similarly, according to the channel quality value measured at time t+1, we can use the second transverse interval filter to measure the channel quality value at time t+3. The specific steps are as above, except that the input parameter values of the filters are different, and will not be repeated here. When measuring the channel quality at time t+4, the first horizontally spaced adaptive filter is used again, and overlapped repeatedly in this way, so as to achieve the purpose of measuring channel quality.

For measuring the channel quality of the next two frames, the horizontal interval adaptive filter of the present invention can be used, that is, the first and second two horizontal interval adaptive filters are used to measure in turn, and the input is the data of the interval value . For other cases, if the channel quality of the next three frames is measured, three horizontally spaced adaptive filters are used to measure in turn. and so on. As the number of intervals increases, the inter-frame data correlation also gradually decreases, and the corresponding number of taps of the horizontal interval adaptive filter also decreases, which will affect the performance of the horizontal interval adaptive filter. However, according to the current usage of the terminal equipment, the data processing time with a delay of two frames (10 ms) is sufficient.

The research shows that the transverse spacing filter can improve the throughput of the system no matter in the low-speed or medium-high-speed channel conditions, especially in the medium-high-speed environment, the improved system is particularly obvious.

Please also refer to FIG. 5 , which is a schematic diagram of performance comparison of various schemes in a low-speed mobile channel (such as an ITU-PA3 channel). In the low-speed mobile channel, although the measurement performance of the lateral spacing adaptive filter is not as good as that of the existing continuous measurement filter, it is worse than that without the continuous measurement filter. As shown in Figure 5, it can be known that the number of taps M of the filter is 4, and under the ITU-PA3 channel, the continuous measurement filter is better than the performance of the horizontal interval adaptive filter of the present invention, and the horizontal interval adaptive filter The performance is better than direct feedback without measurement filter. This is because in a low-speed environment, the filter mainly plays the role of measurement. Since the correlation between the input data of the continuous measurement filter is worse than that of the horizontal interval adaptive filter, its measurement performance is better, while the horizontal interval adaptive filter Although the correlation of the input data is not as good as that of the continuous measurement filter, the filter still plays a certain measurement role, so the performance is better than the link without the measurement filter.

The Fig. 5 also compares the performance of the system under the data smoothing method at the same time. The data smoothing processing refers to performing averaging processing on all received data within a sliding time window. And use them respectively as the channel information reported by the user equipment UE at the next moment. From the simulation results in the figure, in low-speed mobile channels (such as ITU-PA3 channels), this smoothing process has poor performance because it has no measurement function.

与测量期望值的差值来调整的。因此，在测量t+3时刻的值时，调整后的抽头权值系数与t+1时刻测量值相关性小，造成t+3时刻的测量值误差大。以此类推，测量值的误差对高速下行分组接入HSDPA的吞吐量造成较大的影响。由此可见，现有的连续测量滤波器在中、高速信道环境中会损害系统性能，而本发明采用横向间隔测量滤波器在中高移动速度环境下，可以获得更好的系统吞吐量，从而提高系统的性能。如图6A、6B所示。所述图6A、6B分别是HSDPA链路在中速移动信道(ITU-VA30)和高速移动信道(ITU-VA120)下采用图5中所述的4种处理方法后性能比较的示意图。As the mobile speed of the UE increases, the channel change also accelerates, and the measurement performance of the existing continuous measurement filter cannot keep up with the channel change, because in the low-speed mobile state, the channel change of each frame is slow, and the inter-frame data The correlation is large, so the measured value of the next two steps of continuous measurement is close to the real value, which can improve the system performance. However, in the high-speed mobile state, due to the rapid channel change, the inter-frame data correlation is small, and the value at time t and time t+1 has a large change. In the prior art, the tap weight is used

Adjusted for the difference from the measured expected value. Therefore, when measuring the value at time t+3, the correlation between the adjusted tap weight coefficient and the measured value at time t+1 is small, resulting in a large error in the measured value at time t+3. By analogy, the error of the measured value has a greater impact on the throughput of the high-speed downlink packet access HSDPA. It can be seen that the existing continuous measurement filter will damage the system performance in the medium and high-speed channel environment, but the present invention can obtain better system throughput by using the horizontally spaced measurement filter in the medium and high moving speed environment, thereby improving system performance. As shown in Figure 6A, 6B. 6A and 6B are schematic diagrams of HSDPA link performance comparison after adopting the four processing methods described in FIG. 5 under medium-speed mobile channel (ITU-VA30) and high-speed mobile channel (ITU-VA120).

It can be seen from the simulation results in Figs. 6A and 6B that under high-speed channels, the performance of the laterally spaced adaptive filter of the present invention is better than that of the continuous measurement filter of the prior art and the case of no filter. At this time, the horizontal interval adaptive filter mainly plays the role of weighted average processing and smoothing of the previous data, while the measurement function takes a back seat. However, comparing the performance of the horizontal interval adaptive filter and the data smoothing method, The two are very close.

In addition, in terms of algorithm complexity, the computational complexity of the horizontally spaced adaptive filter is the same as that of the continuous measurement filter with the same number of taps, but considering that the horizontally spaced adaptive filter requires more storage space for the tap coefficients , for example, when two horizontally spaced adaptive filters are used, the storage capacity is twice that of the continuous measurement filter. Since the horizontally spaced adaptive filter has better smoothing performance in medium and high-speed mobile environments, as shown in Figures 6A and 6B, the performance of the 4-tap horizontally spaced adaptive filter is better than that of the 8-tap continuous measurement filter Performance, therefore, the laterally spaced adaptive filter can achieve better performance with lower computational complexity. Because in the linear adaptive filter, the complexity of calculation is proportional to the number of taps, the more the number of taps, the greater the amount of calculation.

From the above analysis results, it can be seen that the laterally spaced adaptive filter has two functions: channel measurement in low-speed mobile environment and data smoothing in medium and high-speed mobile environment, so it can improve system performance in various environments. Especially in HSDPA, the adaptive filter is used to measure the channel performance by measuring the horizontal interval. Because the lateral interval measurement adaptive filter has a better smoothing function in the channel environment of medium and high moving speeds, it has the effect of measuring the channel in the low-speed environment. In addition, the present invention can also use different numbers of laterally spaced adaptive filters to measure in turn according to different measurement delay frames, and this new signal processing method can reduce the impact of delay on the system in medium and high-speed mobile environments. performance impact. The continuous measurement filter can only improve the performance of the system in the low-speed mobile channel, but it will reduce the throughput of the system in the medium and high-speed channel environment. The horizontal spacing adaptive filter of the present invention directly adopts the data smoothing method, which is just opposite to the situation of the continuous measurement filter. It can be seen that the horizontal interval adaptive filter is a compromise between the two, and can improve the performance of the system to a certain extent.

The invention proposes an effective channel measurement method, which can improve system throughput and frequency band utilization ratio when applied to HSDPA link. Since there is no simple and effective method for the performance loss caused by the delay between measurement and transmission, this method can not only reduce this loss to a certain extent, but also has low computational complexity and is easy to implement. It is a practical optimization method.

The above is only a preferred embodiment of the present invention, it should be pointed out that for those of ordinary skill in the art, without departing from the principle of the present invention, some improvements and modifications can also be made, and these improvements and modifications should also be It is regarded as the protection scope of the present invention.

Claims (14)

1. A channel measurement method based on a horizontally spaced adaptive filter, said horizontally spaced adaptive filter is made up of two horizontally spaced adaptive filters, characterized in that it comprises steps: A. Measuring the channel quality value at time t+2 by using the first horizontal interval adaptive filter at the first predetermined time t; B. Measuring the channel quality value at time t+3 by using the second laterally spaced adaptive filter at time t+1; C. The first horizontal spacing adaptive filter and the second horizontal spacing adaptive filter measure channel quality values at time t+2n and t+2n+1 in turn. 2. according to the described channel measurement method based on horizontal spacing adaptive filter of claim 1, it is characterized in that, described step A specifically comprises: 21) Adjusting the tap weights of the first horizontal interval adaptive filter according to the preset tap signal difference e'(t) at the first predetermined time t; 22) Perform a linear operation on the adjustment result of the tap weight and the input value of the first horizontally spaced adaptive filter to obtain the channel quality measurement value at time t+2. 3. according to the described channel measuring method based on horizontal spacing adaptive filter of claim 2, it is characterized in that, described in step 21) tap signal difference e ' (t) is according to the channel of time t-2 measurement time t It is calculated by the difference between the measured value and the expected value of the channel at time t. 4. according to the described channel measurement method based on horizontal spaced adaptive filter of claim 3, it is characterized in that, the channel measurement value of measuring time t at time t-2 is according to the channel of time t and the even-numbered time before it The quality value is measured as the input value of the filter. 5. according to the described channel measurement method based on horizontal interval adaptive filter of claim 2, it is characterized in that, the formula of the tap weight of described adjustment first horizontal interval adaptive filter is: ω'(t)=ω(t-2)+μ(t)e'(t)U(t) Among them, ω(t-2) is the tap weight of the first horizontally spaced adaptive filter at time t-2, μ(t) is the adjustment step size of the minimum mean square error LMS filter, U(t) is the time t Input value for the first horizontally spaced adaptive filter tap. 6. The channel measurement method based on the horizontal spacing adaptive filter according to claim 5, wherein the value of μ(t) can be a constant or a function value related to the input vector U(t). 7. according to the described channel measurement method based on horizontal interval adaptive filter of claim 2, it is characterized in that, the channel measurement value of moment t+2 after linear operation described in step 22) satisfies equation: $\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; p</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; \&phi;</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>$ Wherein, the ω'(t) is the tap vector of the first horizontally spaced adaptive filter, and the U(t) is the input value of the tap of the first horizontally spaced adaptive filter at time t. 8. according to the described channel measurement method based on horizontal spacing adaptive filter of claim 1, it is characterized in that, described step B specifically comprises the step: 81) Adjust the tap weights of the second horizontal interval adaptive filter according to the preset tap signal difference e"(t+1) at time t+1; 82) Perform a linear operation on the adjustment result of the tap weight and the input value of the second horizontally spaced adaptive filter to obtain the channel quality measurement value at time t+3. 9. according to the described channel measurement method based on horizontally spaced adaptive filter of claim 8, it is characterized in that, described tap signal difference e " (t+1) in step 81) is according to the moment t-1 measurement moment t The difference between the measured channel value at +1 and the expected channel value at time t+1 is calculated. 10. The channel measurement method based on the horizontal spacing adaptive filter according to claim 9, characterized in that, the channel measurement value at time t+1 measured at time t-1 is based on time t+1 and before The channel quality values at odd time instants are measured as the input values of the filter. 11. according to the described channel measurement method based on horizontal interval adaptive filter of claim 10, it is characterized in that, the formula of the tap weight of described adjustment second horizontal interval adaptive filter is: ω″(t+1)=ω(t-1)+μ(t+1)e″(t+1)U(t+1) Among them, ω(t-1) is the tap weight of the second horizontally spaced adaptive filter at time t-1, μ(t+1) is the adjustment step size of the LMS filter, and U(t+1) is the time t+1 Input value for the second horizontally spaced adaptive filter tap. 12. The channel measurement method based on the horizontally spaced adaptive filter according to claim 11, wherein the value of μ(t+1) can be a constant or a function value related to the input vector U(t+1). 13. according to the described channel measurement method based on horizontal spacing adaptive filter of claim 11, it is characterized in that, the measurement channel value of moment t+3 after described linear operation satisfies equation: $\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; p</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; \&phi;</span>}}^{<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}^{<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$ Wherein, the ω″(t+1) is the tap vector of the second horizontally spaced adaptive filter, and the U(t+1) is the input value of the tap of the second horizontally spaced adaptive filter at time t+1. 14. The channel measurement method based on a laterally spaced adaptive filter according to any one of claims 1-13, wherein the first predetermined time t is the time when the channel quality value is actually received.

Images