WO2002005446A1

WO2002005446A1 - Rake receiver and method related to a rake receiver

Info

Publication number: WO2002005446A1
Application number: PCT/SE2001/001588
Authority: WO
Inventors: Bo Nyberg; Mats Gustafsson
Original assignee: Telefonaktiebolaget Lm Ericsson (Publ)
Priority date: 2000-07-07
Filing date: 2001-07-06
Publication date: 2002-01-17
Also published as: SE0002587D0; AU2001271183A1

Abstract

The present invention is related to a Rake receiver (R2) for detecting a multipath signal stream the delay spread of which fits within a window (WW2). The window must be long enough to handle a large delay spread. A problem addressed is that a Rake receiver with capacity to handle a large delay spread also imposes a large delay in the rake receiver itself and thereby the time of the power control loop is adversely affected. The problem is solved by a Rake receiver having an adjustable window length. The window length can thereby be adjusted to need of the delay spread of the Rake receiver radio environment. A shorter window length is then used in environments that causes the delay spread to be on an average level, than is needed for the worst case delay spread, and thereby the power control loop is made faster.

Description

RAKE RECEIVER AND METHOD RELATED TO A RAKE RECEIVER

TECHNICAL FIELD OF THE INVENTION

The present invention relates to a radio receiver for a spread spectrum signal sequence and to a method related to such a receiver.

DESCRIPTION OF RELATED ART

In figure 1 is shown a mobile station MS acting as a transmitter and a radio base station BTS acting as a receiver of a radio transmitted data stream. The signal sequence is transmitted over various propagation paths P1-P5 from the transmitter to the receiver, one of the paths Pi in this example is a direct wave while on the remaining propagation paths P2-P4 the signal wave is reflected by various obstacles OB.

Due to different propagation delays on the various propagation paths P1-P5, the signal stream sent from the transmitter will be received by the receiver as a number of versions of the original signal stream that are mutually delayed and thus interfering with each other. The signal energy from the various signal paths can be collected in the receiver. A prerequisite is though, that the various propagation path delays are known.

In figure 2, the channel impulse response CHIR is shown in a diagram of power vs. time. In the further description instead of Channel Impulse Response CHIR the wording delay profile will be used. The various power peaks PPK1-PPK5 in the diagram corresponds to the different propagation paths P1-P5. Their propagation delays T_X, τ₂, τ₃ are also indicated.

A Rake receiver is a radio receiver for a DS-CDMA (Direct Sequence Code Division Multiple Access) signal and comprises a searcher. The searcher is arranged to detect the power peaks PPK1-PPK5 that occurs within a certain search window WWl. The searcher also detects the difference in propagation delays Δτι,Δτ₂,Δτ₃,Δτ₄,Δτ₅ of the power peaks PPK1-PPK5, with respect to the boarder of the window WWl.

In radio communications systems based on DS-CDMA technology a radio spectrum band is shared by multiple users on code division basis. By this is meant that each user is given a unique spreading code that identifies a radio communication channel dedicated to the specific user. The spreading code also functions as to broaden the frequency band of the original user data. This implies that the spreading code rate is considerable higher than the user data rate, since frequency bandwidth is proportional to the data rate. Figure 3a shows the principle of bandspreading the power spectral density of the user data, where the user data power originally is gathered in a narrow user data spectrum band NSB but with the spreading code added to it, the power is spread over a wider band SSB.

Figure 3b shows the principle of multiple access in DS-CDMA technology, where the bandspread signal power of multiple users is transmitted in the same radio frequency band SSB. Radio communication channels RCH of the various users can be resolved in a receiver by the unique spreading codes.

A DS-CDMA transmitter adds to a user data stream the unique spreading code sequence. Thereby a bandspread user data stream is produced.

Figure 5 shows a searcher Si, for estimating the channel impulse response CHIR of figure 2. Estimating the channel impulse response CHIR also includes keeping track of changing propagation conditions. However, the searcher has only capacity for detecting the delay profile within the window WW. The searcher receives on its input the signal from all possible propagation paths P1-P5. The searcher is arranged to find correlation between the various propagation path delays of pilot spreading code stream PN_D sent from the transmitter and its complex conjugate code stream PN_D ^* generated in the receiver.

The searcher in figure 4 comprises a buffer B51 for buffering the sequence of the received signal (r) to operate on and a code generator CG arranged to produce the complex conjugate code stream PN_D . It further comprises a delaying means DM5 with an input from the code generator CG and a multiplier M51 with an input from the delaying means DM5 and an input from the buffer B51 for the received signal stream. In order to find the power peaks PPK1-PPK5 in the first window WWl, the complex conjugate code stream PN_D ^* is combined with the received signal sequence (r) in the multiplier M51 and the combined sequence is fed to a correlation detector module CDM for the delay profile to be detected in a well know process.

Figure 5 shows a Rake receiver Rl arranged for separately demodulating the various propagation paths P1-P5 of the spectrum spread user data stream UDS that fit into the window WW indicated in figure 2. The Rake receiver Rl comprises a set of fingers fl-f5, each of the fingers fl-f5 for handling one of the propagation paths P1-P5. The Rake receiver Rl also comprises a code generator CG for generating the complex conjugate code stream PN_D ^* of the spreading code sequence used in the transmitter for bandspreading the user data stream UDS. A delaying means DM1 at the output of the code generator CG is arranged to delay the conjugate code stream PN_D ^* a length corresponding to the delay at the end of the window WWl, i.e. the maximum delay possible to detect within the window.

In each of the fingers fl-f5 the received radio signal streams are delayed for a time that corresponds to the difference in delay Δτι,Δτ_2/ Δτ₃,Δτ₄, Δτ₅ between the actual propagation path P1-P5 handled in the respective finger f1-f5 and the delay TMAX at the end of the window. Thereby, the bandspread user data stream of the various propagation paths is time aligned.

The delayed conjugate code stream PN_D ^* is combined in each finger fl-f5 with the time aligned data streams. Since the conjugate code stream is correlated with the time aligned data streams, the user data stream is thereby despread and its energy is regathered in the narrow spectrum band NSB.

In the narrow spectrum band NSB an integrator ITR collects energy over short periods and feeds the collected energy to a demodulator DMD. At its output the demodulator produces a complex value. The real part of complex value of all fingers fl-f5 are added and based on this addition a decoding decision is made on every bit in the user data stream UDS. Such combining before the decision is often referred to as soft combining.

As indicated in fig 3b mobile terminals MS in DS-CDMA mobile communication system transmits signal streams in the same frequency band SSB. The radio base station BTS receives signal streams from all the mobile terminals that it serves as well as uplink transmission from mobiles served by other radio base stations, all of which interfere with a wanted signal stream from one mobile station. The wanted signal stream can be detected in the radio base station BTS provided that after despreading the signal is higher than the background interference. Therefore the base station BTS controls the uplink power from the mobile stations MS that it serves, this is called uplink power control. If the power of a wanted signal is to low, the signal cannot be detected. If, instead, the power is higher than needed for detection, it unnecessary increases the overall interference in the communication system and thereby decreases the number of mobile terminal possible to serve by the system. In addition to the problem of interference, fast fading on the radio link causes the signal strength to change rapidly in the receiver. In DS-CDMA communication system a fast power control loop functions to mitigate the fast fading influence on the received uplink signal strength. The fast power control loop evaluates the signal strength received in the base station and sends commands to the mobile station that in turn adjusts its power according to the commands.

One of the big challenges in DS-CDMA technology is to make the fast uplink power control real fast and thereby increase the overall traffic capacity of the communication system. One of the difficulties is though that the radio transmission time effects the time of the fast power control loop especially as propagation delays of the wanted signal can vary greatly.

In order to make a correct evaluation of the received power all essential propagation paths have to be detected in the radio base station BTS.

SUMMARY OF THE INVENTION

A problem related to a prior art Rake receiver is how to fulfil the conflicting requirement on signal detection window long enough for detecting various signal paths in all possible radio environments on one hand and the requirement on uplink power control to be fast on the other hand. If the window is made long enough for detecting all essential propagation paths in an environment with big differencies in propagation delays this will make the process for detecting the received power longer and adversely effect the power control loop.

The essence of the present invention is to enable the signal detection window to be adjustable in length to enable the length to be adjusted to correspond to the need in the radio environment of the Rake receiver. The present invention solves this problem with a Rake receiver that has a signal detecting window arranged to be possible to redefine in length.

The present invention also solves this problem with method related to a Rake receiver having a first window for estimating a delay profile and a second window for detecting the signal stream of various propagation paths. The method includes steps of estimating the delay profile, evaluating the delay spread of power peaks in the delay profile and redefining the second window length based on the result of the evaluating step.

Thereby the signal detection window need not correspond to the window used by the searcher for detection of power peaks and for finding the propagation delays.

An advantage with the present invention is that the power control loop can be faster in radio environments allowing for the signal detection window to be shortened. A long signal detection window results in the received signal being delayed a corresponding length in the receiver to be time aligned. When the signal detection window is shortened the delay is also shortened and thus the received power is more quickly obtained in the receiver.

The radio receiver is manufactured with a window long enough for handling bad radio environments, if it is then used in a fairly good radio environment the signal detection window length is decreased and the power control loop is thereby made faster.

DESCRIPTION OF THE DRAWINGS

Figure 1 shows an area view of various wave propagation paths from a transmitter to a receiver. Figure 2 shows a diagram of the radio channel impulse response showing received power as a function of time.

Figure 3a is a diagram illustrating the principle of frequency bandspreading signal energy that is adopted in DS- CDMA technology.

Figure 3b is a diagram illustrating the multiple access principle of DS-CDMA technology.

Figure 4 is a block diagram of a DS-CDMA prior art searcher.

Figure 5 is a block diagram of a prior art Rake receiver.

Figure 6 is a modified version of figure 2.

Figure 7 is a block diagram of an inventive Rake receiver.

DESCRIPTION OF PREFERRED EMBODIMENTS

Figure 6 shows a delay profile CHIR with peaks PPK1-PPK3 that are more narrow than the peaks of the delay profile shown in figure 2. The propagation delay spread as shown in the delay profile CHIR of figure 2 occurs in bad radio environments. For most environments where a radio base station is situated the differences in propagation delays are smaller and the delay profile CHIR of figure 2 is more likely.

In figure 6 two windows WWl, WW2 are indicated. The first window WWl has a first length Ll, or time depth Ll, for the delay profile CHIR to be detected. The second window WW2 has a second length L2, or time depth L2, for detection of signals. Signal streams received over propagation paths PI^¬ PS, the peaks of which PPK1-PPK3 fits into the second window WW2 are to be detected.

The delay profile CHIR is detected by a searcher Si in a Rake receiver R2 shown in figure 7. The signal streams are detected by the fingers fl-f3 of the same Rake receiver R2. The searcher S2 uses the fixed length Ll of the first window WWl for detecting the delay profile CHIR.

The second window WW2 length L2 is, however, made adjustable in order to enable the second window WW2 length to be selected to be shorter than the first window length Ll when the radio environments are fairly good.

The inventive Rake receiver R2 of figure 7 mainly comprises the same parts as the prior art Rake receiver Rl of figure 5, i.e. in addition to the searcher S2 and the number of fingers fl-f3 the Rake receiver R2 also includes a buffer B2, a code generator CG and delaying means DM2 for delaying a code generated PN_D by code generator CG.

The buffer B2 receives and buffers the received signal stream according to the well known FIFO principle (First-In, First-Out) . The purpose of the buffer is to time align the signal streams of different propagation paths P1-P3, and to this end different delays of the received signal stream are tapped from the buffer B2 and fed to corresponding fingers fl-f3. The searcher S2 controls the delay Δτι-Δτ at which the signal stream is tapped to each of the fingers fl-f3. The signals are delayed in the buffer a time that corresponding to the difference between the actual propagation delay and the maximum propagation delay x^x that is possible to detect within the second window WW2. At the output from the buffer the signal of various propagation paths P1-P3 is thus given at total delay, i.e. propagation delay and buffering delay, that is equal to the to the delay at the second window end.

The description so far just describes the function common with the buffer Bl of the prior art Rake receiver R2. The inventive Rake receiver R2 differs in that the maximum delay depth X_MΆX used in the buffer B2 is adjustable. The searcher S2 controls that no more of the buffer depth will be used than the selected maximum depth that corresponds to the selected length L2 of the second window WW2.

The delaying means DM2 of the inventive Rake receiver S2 applies an adjustable delay to the code PN_D ^* generated by the code generator. The delay is controlled by the searcher S2 to correspond to the maximum delay X_MAX of the buffer B2. If the maximum delay of the buffer B2 is changed to a new value, the delay of the code PN_D ^* must be given the corresponding value, in order for the code PN_D ^* to be correlated with the time aligned signal streams in the fingers fl-f3.

The searcher S2 corresponds to the to the prior art searcher S2, except for the processor CPU that comprises means for controlling the buffer B2 and the delaying means DM2 as described above.

The length L2 of second window WW2 is selected by the operator when putting a new radio base station into operation. The selection is made at the man machine interface of a radio base station BTS incorporating the inventive Rake receiver R2.

Alternatively the searcher S2 itself selects the second window WW2 length L2 based upon statistics of the delay spread of the power peaks PPK1-PPK3 found in a number of delay profiles CHIR. Another alternative would be that the searcher S2 selected the second window length L2 based upon the power delay profile CHIR for each communication set up. However, a selected second window length L2 should not be changed often during a communication session since data is lost in the buffer B2 when the second window length L2 is changed.

In an alternatively embodiment the Rake receiver R2 includes several parallel buffers Bl with different time depths, each buffer corresponding to a fixed second window length L2. Just one of the buffers at a time is tapped for feeding the time aligned signal of various propagation paths to the Rake fingers fl-f3. When a new fixed second window length L2 is selected, the corresponding buffer B2 is used instead of the previous one, for tapping signals to the fingers fl-f3. When the second window length is changed also the delay of the code from code generator CG must be changed accordingly. For this purpose each of the buffers B2 is associated with a corresponding delaying means DM2, and as the buffer B2 is exchanged also the delaying means DM2 is changed.

Some prior art Rake receivers are designed to time align the signal of various propagation paths at the output of the fingers instead of at the input of the fingers, as in the Rake receiver Rl and R2 so far described in this description. However, this invention is applicable also when the buffering means are connected to the output of the Rake fingers fl-f3 for time aligning the signal stream of the various propagation paths. After time alignment the signal of the various propagation paths is combined.

Claims

1. A Rake receiver arranged to detect a multipath signal stream, the delay spread of which fits within a window (WW2) with a defined window length (L2)

c h a r a c t e r i s e d in

an arrangement for redefinition of the window length.

2. A Rake receiver as defined in claim 1 including buffering means (B2) that is arranged to delay the received multipath signal stream to compensate for the difference in propagation delay of the various propagation paths (P1-P3) relative to the maximum propagation delay (x_maχ) that fits within the window (WW2) .

3. A Rake receiver as defined in claim 2 wherein the buffering means have a buffer depth that is adjustable to correspond to a selected window length (L2) .

4. A Rake receiver as defined in claim 3 wherein a delaying means is arranged to generate a delay of a despreading code generated in the Rake receiver and that is arranged to adjust the delay to correspond to the maximum propagation delay (x_max) that fits within the window (WW2) .

5. A Rake receiver as defined in claim 2, 3 or 4 wherein the buffering means is arranged at the input of fingers (fl-f3) in the Rake receiver.

6. A Rake receiver as defined in claim 2 or 3 wherein the buffering means (B2) is arranged at the output of fingers (fl-f3) in the Rake receiver.

7. A Rake receiver as defined in claim 2 wherein the buffering means (B2) comprises several buffers of different depth each of which corresponds to a defined window length (12) and one of the buffers are alternatively feeding fingers in the Rake receiver with the propagation delay compensated signal.

8. A Rake receiver as defined in claim 1 wherein the redefinition of the window length (L2) is arranged to be made in response to an estimated delay spread.

9. A Rake receiver as defined in claim 8 including a searcher for estimating the delay spread.

10. A method related to a Rake receiver having a first window (WWl) with a first length (Ll) for estimating a delay profile (CIR) and a second window (WW2) with a second length (L2) for detecting a multipath signal, the delay spread of whitch fits into the second window (WW2), wherein the method comprises the steps of:

estimating the delay profile (CHIR) ;

evaluating the delay spread of the multipath signal in the delay profile (CHIR) ;

redefining the second window length (L2) based upon the result of the evaluating step;

detecting the multipath signal;

estimating the delay profile (CHIR) within the first window (WWl) ;

repeating the detecting and estimating steps.

11. A rake receiving method as claimed in claim 10 wherein the detecting step includes delaying the signal stream to time align the ^■ various propagation paths with the maximum delay ( _A ) possible to detect within the second window (WW2) .

12. A rake receiving method as claimed in claim 10 wherein the detecting step further includes adjusting the delay of a generated despreading code (PN_D ) to correlate it with the time aligned signal.