WO2003069852A1

WO2003069852A1 - Data transfer method

Info

Publication number: WO2003069852A1
Application number: PCT/EP2003/001584
Authority: WO
Inventors: Bernhard Raaf
Original assignee: Siemens Aktiengesellschaft
Priority date: 2002-02-15
Filing date: 2003-02-17
Publication date: 2003-08-21
Also published as: US7428406B2; US20050094561A1; AU2003218988A1; JP4280642B2; EP1474898A1; DE60311426D1; KR20040081442A; CN100484086C; KR100945613B1; JP2005518148A; DE60311426T2; EP1474898B1; CN1633788A

Abstract

A data transfer method via a packet oriented channel and a continuous channel in parallel between a mobile station (MS) and a base station (BS), where the transfer via the continuous channel is interrupted whereby at least one transfer gap is formed.

Description

Data Transfer Method

Field of the Invention

The present invention relates to a data transfer or data transmission method, that takes into account the interdependencies between compressed mode and data transfer, especially between compressed mode and packet oriented data transfer e.g. in UMTS (Universal Mobile Telecommunications System) via HSDPA (High Speed Downlink Packet Access), cf. [3] . Compressed mode is applied if inter-frequency measurements, e.g. for handover procedures or OTD measurements, are performed.

Background of the Invention

A terminal or user equipment in a communications system performs measurements also on frequencies different from its actual sending/receiving frequency in order to observe e.g. other base stations or to perform OTD (Observed Time Difference) measurements. During this time (transfer gap) no data transfer takes place. To maintain an average data transfer rate, the data rate outside of the transfer gap is increased in certain time frames. The operational mode in these time frames is related to as compressed mode. This compressed mode obviously influences the data transfer, i.e. interdependencies are occurring.

A base station is a central unit in a cellular communications network, that serves at least terminals or user equipments within a cell of the communications network. It comprises at least a sending/receiving unit. In UMTS it is often referred to as node B.

This situation will be explained further using an example regarding the UMTS System:

While a connection is established between a communication device or user equipment (UE) and a base station (BS or Node B) , the user equipment also observes other base stations in order to find out the base station the optimum connection can be installed to.

For observing another base station, the user equipment has to tune in on frequencies distinct from its actual sending/ receiving frequencies. Thus during the time the user equipment is observing other frequencies, no data are being transmitted or received, at least if the user equipment has only one synthesiser and / or only one RF- part (RF: Radio Frequency) .

However, the user of the user equipment should not notice, that his data transfer has been disrupted in order to create transfer gaps for the so called "inter-frequency measurements" by which frequencies distinct from the actual sending/receiving frequency or frequencies are observed. In the framework of the UMTS standardisation this item is dealt with in [1] .

To maintain a constant average data rate also in the presence of transfer gaps, the net data transfer rate is increased before and after the transfer gaps. By "net data" data actually carrying information are denoted. A certain overhead is added to the net data to ensure that the data can be decoded correctly at the receiver, even if the transmission has not been ideal, i.e. experiences some degradation. The overall data are referred to as gross data, the overhead of data may e.g. consist of parity bits originating from channel coding. Data transfer may be either the transmission or the reception of data or both.

The situation of a transfer disrupted by a transfer gap TG, e.g. a transmission gap, is shown in Fig.l, which is taken from [1] : The transmit power is depicted versus time; the latter is segmented in frames F as time intervals, each frame itself contains several time slots. The frames during which the user equipment listens to another base station and thus cannot be transferring data continuously are referred to as compressed frames, as the transfer rate has to be increased in some timeslots in this frame to achieve an average rate similar to normal frames, that is when the compressed mode is off.

The frames, in which the data are transferred compressed, are referred to as compressed frames, the respective operating mode as compressed mode.

In compressed frames, TGL (transmission gap length) slots from a first slot Nf_irst to a last slot Nι_ast are not used for transmission of data. As illustrated in Fig. 1, the instantaneous transmit power P, which is depicted versus time t,is increased in the compressed frame F_c before and after the transmission gap TG with the length TGL in order to keep the quality e.g. the BER (Bit Error Rate) or the FER (Frame Error Rate) unaffected by the reduced processing gain. F denotes the length of a normal frame. By reduced processing gain is meant that the data are e.g. encoded less safe than during "normal transmission" . The amount of power increase depends on the actually used transmission time reduction method (see [1], subclause 4.4.3).

In Fig.2 an ordinary transmission sequence can be seen, which is used to explain the terms demodulation, coding etc.

The signal may be generated at the source or transmitter TX. In a subsequent analogue to digital converter A/D the signal is digitised, thus the smallest information carrying unit is one bit. Digitising includes e.g. the steps of sampling and quantising the signal. Then various coding steps^' in the encoder C are performed: source coding is performed to get rid of redundancies in the signal or digitised data are used directly (which means no A/D converting or source coding etc needs to be done) ; channel coding is applied to protect the bits. After coding the signal is spread. Now the smallest ' information carrying unit is a chip. Due to spreading the chip rate for a transmission is typically considerably higher than the bit rate.

At the digital modulator DM the data are 'translated' into symbols that differ for the various modulation and coding schemes. The higher a modulation the higher the number of bits that are translated into a symbol.

Now the data are being transferred and subject to influences from noise and interference that can have an impact on the data. For example a previous symbol (1,1) at the digital demodulator might be changed to (0.7,0.9). Hence one speaks about a transfer via an analogous channel AC. At the receiving side the corresponding processes of demodulation at the digital demodulator DD and the decoding at the Decoder D and the digital to analogue conversion at the D/A converter are taking place.

Generally spoken, in compressed mode the transmit power is increased to ensure a safe transmission of the less safe encoded data: By coding the data less, with the same gross data transfer rate a higher net data transfer rate can be achieved. The data bits are e.g. punctuated more than in the frames before or the coding of the data bits has been performed with a lower spreading factor. The compressed mode therefore entails rather complex calculations how the gross data are modified depending on the gap length and on the current data transfer rate - and on the duration of the compressed mode (cf. Fig.l, the time required for the time slots with the higher transmitting power) and in how this modification is realised, e.g. by using a different modulation scheme - a different spreading factor

- puncturing of data, i.e. cutting out individual or group of bits.

It is decided by 'the network which frames are compressed. When in compressed mode, compressed frames can occur periodically or requested on demand. The rate and type of compressed frames is variable and depends on the environment and the measurement requirements. In OSI-layers above the physical layer the knowledge of the scheduling of the compressed frames is existent, thus the above mentioned calculations for the compressed mode can be done. As a further variant for the realisation of compressed frames, it is known that higher layers can also restrict the data rate during frames which will undergo compression on the physical layer, thus making the operation in compressed mode more reliable because less excessive rate matching will be necessary for the compressed frames due to the lower data- rate.

Furthermore, transmission gaps can be placed at different positions, dependent on the purpose such as interfrequency power measurement, acquisition of control channel of another system or carrier and handover operation, cf. [1], section 4.4.4.

For the so called single frame method, the transmission gap is located within the compressed frame. The exact position depends on the length of the transmission gap TGL (transmission gap length) . For the double frame method the transmission gap is overlaps two neighboured frames. In Fig.3a the single frame method is shown, in Fig. 3b an example for the double frame method.

For example, this type of compressed mode is applied in UMTS (Universal Mobile Telecommunications System) to the DPDCH (Dedicated Physical Data Channel) , across which data are transferred by circuit switching. As already explained, the above described methods for compressing the frames require rather complex calculations.

Also packet switched transfer modes, which may be operated in parallel to the circuit switched modes or continuous channels via which, e.g. speech can be transferred, are affected by the transmission gaps TG. This will be detailed below: In a packet switched transfer mode the data are segmented into packets. Each packet is transferred individually. The quality of the reception is decided on basis of various data operations such as demodulation or decoding (cf . Fig.2) . The receiver sends back a receipt of the reception, e.g. an 'ACK'

(Acknowledge) or a 'NACK' (Not acknowledge) depending whether it has recognised a packet as been received correctly or not. A channel with packet switched transfer mode is e.g. in UTMS the HS-DSCH (High Speed Downlink Shared Channel) which is mapped to the physical channel HS-PDSCH. An overview of this technique is provided in [3] .

The HSDPA data channel is basically an enhancement of the existing UMTS downlink shared- channel (DSCH) . HSDPA allows to code multiplex different users or mobile stations on up to 15 codes with spreading factor of 16. The primary multiple access, however, is in the time domain, where different users can be scheduled every transmission time interval (TTI), which corresponds to 3 UMTS slots, i.e., 2 s . Also the number of codes allocated to one user can change from TTI to TTI. Depending on the system load and channel conditions, the base station or Node B adapts modulation and code rate for each user. A certain combination of code rate and modulation is referred to as MCS (Modulation and Coding Scheme) level. The MCS level may change every TTI. It is determined by the base station based on feedback information or channel quality information (CQI) from the user terminal or mobile station, which stems from channel condition measurements. The channel quality information is sent with a periodicity ranging from one to 80 TTIs.

To achieve the high data rates modulation and coding schemes are used which allow a high information bit rate per code. Therefor so called higher modulation techniques are used by which a symbol contains more than 2 bits. One example is 16- QAM (Quadrature Amplitude Modulation) . For these modulation techniques the individual positions for a bit within a symbol are not equally protected. Therefore, there is the ambition to map important bits to well protected positions and less important bits to less protected positions. This is referred to as bit priority mapping and will be detailed below using an example from HSDPA. Furthermore, for channel coding so called turbo codes with rate R = 1/3 are used. The rate indicates the ratio of the total number of bits to the number of load or systematic bits.

The HS-DSCH is shared among several users. The respective transfer rate for each of the users is decided on basis of the individual channel quality. One of the multiple access possibilities is in the time domain, where different users can be scheduled every transmission time interval (TTI) , which corresponds to three UMTS slots (UMTS: Universal Mobile Telecommunication System), that is 2ms.

The transport channel HS-DSCH is mapped - as above mentioned to the physical channel HS-PDSCH (High Speed Physical Downlink Shared Channel) , to which a compressed mode can be applied. For higher data rates, where a single HS-PDSCH cannot carry the entire data rate, a set of HS-PDSCH channels can be used, in this case all the HS-PDSCHs of the set are transmitted simultaneously and they can be distinguished because they use different spreading codes. The invention is however not affected whether one HS-PDSCH or a set is used.

In principle the above described compressed mode can be applied to packet switched data, too. Therefor calculations need to be done cf . [2] . However, a simpler process is desirable to make the calculations less complex. Short Description of the Invention

It is an object of the invention to provide an easy solution to handle the interdependencies between compressed mode and a packet switched data transfer channel,

It is a further object of the invention, to provide a transmission method, where information is transmitted in packets, and where transmission or reception gaps are provided, whereby no transmission or reception is performed during transmission or reception gaps.

The object of the invention will be achieved with a method and an arrangement which are characterised by what is disclosed in the appended claims. Advantageous embodiments and modifications of the invention will be presented in the dependent claims .

It is relevant, that data are transferred or transmitted in packets outside of transmission or reception gaps that can be used by the mobile station to tune to other frequencies.

By a data transfer method according to an embodiment of the invention data are transferred between a mobile station and a base station via a via a packet oriented channel and a continuous channel in parallel. Thereby the transfer via the continuous channel is interrupted such that at least one transfer gap is formed.

In a first embodiment, after the reception of a data packet no receipt, e.g. an 'ACK' or a 'NACK', is sent back by the receiver after a first processing time. The first processing time is often called UE-processing time and denotes the time between the end of reception of a signal and the start of transmission of a successive or subsequent signal, which may be an ACK or NACK signal as in the case of the UMTS system.

In UMTS 5ms are allocated for the UE-processing time. The timing structure of HSDPA is shown in Fig.3, where also the length of 1 TTI can be seen, which corresponds to 2ms in

UMTS.

During the first processing time the signal is e.g. demodulated, that is a set of symbols is assigned to a set of incoming data, and to allot a probability to each symbol or bit, that the decision for a symbol or bit has been correct. By not sending a receipt during the transmission gap the gap is maintained despite of the HSDPA transmission and the gap can be used for the initially described observation of other frequencies .

In a second embodiment of the invention a data packet is stored by the receiver, e.g. in the receiver, thus representing a first data set. A decoding is done, when the data packet has been received repeatedly, thus at least a second data set is provided. By a joint decoding of the data sets the performance of the decoding process is improved in respect to decoding e.g. the first data set alone.

In a third embodiment of the invention the receipt is sent back after the processing time plus a certain delay. This also ensures, that no receipt is transferred in the transmission gap.

Thus, it is another object of the invention to propose a method how to restrict the scheduling of HSDPA transmissions in order not to compromise compressed mode operation and still sacrifice as little data throughput, especially HSDPA throughput, as possible.

Description of Figures and Embodiments

Further aspects, features, objects and advantages of the invention will become more apparent from the detailed description set forth below, especially when taken in conjunction with the drawings wherein show:

Fig.l a scheme of a compressed mode transmission;

Fig.2 block diagram of transmission process;

Fig.3 possible positions of the time transmission interval, for a) the single frame method and for b) the double frame method; Fig. 4 the throughput for HSDPA depending on the

Modulation and coding scheme; Fig.5 a timing structure of HSDPA uplink timing, from figure 14 of [3] . Fig. 6 Schematic diagram of a communications network showing data transfer between a mobile station and a base station via a continuos channel and a packet oriented channel, respectively, according to one preferred embodiment of the invention.

Detailed description of the Invention

For a thorough understanding of the invention's scope it is differentiated now between various applications of the compressed mode and its implications on the HSDPA transmission. Generally spoken, methods according to the invention propose not to send any feedback during a transmission gap. This is done either by not sending any acknowledgement, e.g. an ACK or an NACK, that would be transmitted during a transmission gap, or by sending it with a delay sufficient to postpone its transmission after the transmission gap. As explained above, the feedback is normally used in packet oriented transmission to acknowledge the receipt of a transmission (ACK) or the failure to receive a packet correctly.

In consequence the data packets whose acknowldege ents are affected are either being re-transmitted or not retransmitted at all.

The above mentioned continuous data channel may be the DCH (Dedicated Channel) and the corresponding Dedicated Physical Data Channel and Dedicated Physical Control Channel by which e.g. speech can be transferred.

These considerations apply to uplink and downlink compressed mode scenarios as well. Now it will be differentiated between uplink and downlink compressed mode:

Restrictions on HSDPA transmission due to downlink compressed mode (cf. embodiment 1)

The restrictions which have to be accepted for (downlink) HSDPA transmission can be described as follows: During the transfer gaps provided by compressed mode, the user equipment is free to tune to other frequencies and therefore cannot receive any signals on its currently assigned frequency.

In general during the transfer gaps the user equipment needs not receive any signal. In other words, the base station may send signals during the transfer gap, but it cannot be assumed, that the signals are received by the user equipment. Therefore for one embodiment this simple fact is specified, disregarding specific details of e.g. the offset of the DPCH against the HS-DSCH and HS-SCCH (High Speed Synchronisation Control Channel) or the specific details of the compressed mode patterns or the different frame structure types, as this is not necessary.

For any given offset and for given parameters of the compressed mode, it will be quite easy to calculate whether HS-SCCH or HS-SDCH overlap with the transfer gap. Note that ' an overlap is already given if any single chip of the HSDPA transmission would overlap with any single chip (see- explanations to Fig.2) of the transfer gap. Further more, because the HI transmission (HI:HS-DSCH Indicator) always fully overlaps the corresponding HS-SCCH transmission (if the HI is transmitted at all, the system can also be designed without the HI), it is sufficient to consider only the latter.

In another embodiment the base station does not transmit any data- in this case. While this will be the optimum solution if the base station knows that the user equipment does not listen during the time in question, this does not necessarily have to be the case. For example, if the transfer gap of the compressed mode only partially overlaps with a HSDPA TTI, then it may be possible for the user equipment to still receive parts of the TTI and, due to the error correcting coding which is applied, it might be possible to decode the data packet also from this partial reception. This is in particular possible, if the packet was already a retransmission, in this case it might be possible to decode the packet based on the received information of the earlier transmission (s) plus that partial one. It may also be the case that successful decoding is not possible due to the part of the TTI that is missed during the gap, however, the received part of the TTI, together with a subsequent' transmission may make it possible to decode the data.

Therefore, in an other embodiment, the user equipment receives at least part of the data even if the TTI (partly) overlaps with a transfer gap by compressed mode. The user equipment can then either perform a complete decoding or only store these data in the user equipment soft buffer (hpw this is done will be described in more detail in an other embodiment). In an other variant the user equipment ' may, at least for some transfer gaps, decide that it does not need to interrupt communication as long as the entire transfer gap, but only for a shorter time. This may be possible if the actions that are scheduled for a particular transfer gap can also be performed in this shorter time. This may be the case e.g., if the user equipment has a synthesiser that can be tuned to the other frequencies more quickly than was assumed when the requirements were decided. An other possibility is, that the action scheduled for a particular transfer gap can inherently be done in a shorter time, e.g. a so called BSIC - verification on an other GSM frequency (BSIC= Base Station Identification Code) . This can be done in a rather short time, but the exact position of this time within the transfer gap may not be known at the base station when the transfer gap is being scheduled, so that a larger transfer gap is scheduled than will actually be necessary in the end.

Restrictions on HSDPA transmission due to uplink compressed mode (cf. e.g. second embodiment) The restrictions due to uplink compressed mode are somewhat different, because it is not the downlink HSDPA transmission that is causing the trouble but the uplink sent receipt, e.g. the ACK/NACK (Acknowledge/Not acknowledge) transmission. Of course, during the transmission gaps for uplink compressed mode, the user equipment is not able (and therefore not requested) to transmit anything, in particular not a receipt.

However, this does not rule out the possibility for the base station, which inUMTS is also often called Node B, to transmit an associated downlink HSDPA transmission before the transfer gap.

The only restriction a base station has to take into account is the fact, that it cannot expect the user equipment to transmit any receipt, e.g. either ACK or NACK, in response to this HSDPA transmission. As a consequence, the base station will not get an information about whether the user equipment has correctly received the packet and will therefore have to retransmit the packet in any case. At first sight sending an HSDPA transmission in this case may look like a waste of resources, but actually the throughput can be increased: The base station will not use a MCS (Modulation an Coding Scheme) where there is a fair chance that the user equipment will decode the initial transmission, but it will use a MCS where the second transmission has such a fair chance.

A modulation and coding scheme describes in particular, which transmission method is being used for the packet, in particular the amount of redundancy that is used and that is available for error correction purposes (coding) , and the modulation scheme which describes how many bits are transmitted with a single symbol. E.g. for QPSK (Quadrature Phase Shift Key) two bits are transmitted while for 16 QAM (Quadrature Amplitude Modulation) transmission 4 bits are transmitted. The code rate R describes how many redundancy is being used, R is defined as

R = Number of bits before of coding / Number of bits after coding

Table 1: Modulation and Coding Scheme (MCS)

MCS modulation code info bit level rate rate

R per code

5 3/4 720 kbps

16-QAM 4 1/2 480 kbps

3 3/4 360 kbps

2 QPSK 1/2 240 kbps

1 1/4 120 kbps

The previous table gives an example of a possible set of Modulation and Coding Schemes that can be used and Fig. 4 shows the throughput that can be achieved depending on the channel quality for these Modulation and Coding Schemes.

The abscissa I_or/Io_C is the ratio of the total power spectral density of the downlink to one user terminal to the power spectral density of the band limited noise and interference. Io / o is defined as

/ E_b Λ 1 _o N₀ R_c g_d

where Eb / No is the bit energy to spectral noise density, R_b is the bit rate, and R_c is the chip rate. The factor g_d is fraction of radiated power devoted to the HSDPA data channel.

An total overhead of 20% is assumed (e.g., for signalling channels and pilot channels which are used to determine the influence of the transmission medium on the received signals by the user equipment) leading to g_ - 0.8. Thus the abscissa denotes basically the signal to noise ratio.

The steps in the throughput curves are caused if the initial transmission fails and a retransmission has to be done, which will make the throughput drop to the half for the second transmission, to a third for the third transmission and so on. As can be seen from the graph, after one retransmission a level 2 Modulation and Coding Scheme achieves a throughput of 129 kbps (kilo bits per second) after the second transmission, just as MCS level 1 for the first transmission, in a similar range of the channel quality. In that range the same throughput can be achieved with two transmissions of MCS level 2 as with a single transmission of MCS level 1.

Then of course the initial transmission will have a rather low success probability, which means the ACK/NACK signal that is 'lost', because it is not sent due to the uplink compressed mode transfer gap is most likely a NACK signal, so the information would be redundant .

Note that a scheduler, the control unit in the base station which decides which user equipment to serve during the next TTI and which MSC to use, is anyhow free to always use such an operating point or transfer mode, which also has the advantage of increased time diversity due to combination of two transmissions at different times. The throughput of a higher MCS with retransmission is very similar to the throughput of a MCS carrying half as much payload for an initial transmission. If this scheme is employed, uplink compressed mode does not degrade the throughput of a HSDPA session.

To recapitulate, the base station shall not schedule a HSDPA transmission for a user equipment which is using compressed mode, if part of the HS-SCCH information or the corresponding HS-DSCH TTI (TTI: Transmission Time Interval) overlaps with a transfer gap of downlink compressed mode.

In respect with Fig. 5 it will be detailed, how a part of the HS-SCCH or the HS-DSCH can overlap with a transfer or transmission gap:

A detailed timing diagram of the HSDPA channels in relation to other UMTS channels is given in Fig. 5, which is taken from figure 14 of [3] . Fig. 5 shows the timing offset between the downlink HS-DSCH (High Speed Downlink Shared Channel) and the uplink DPCCH (Dedicated Physical Control CHannel) . The code-multiplexed uplink HS-DPCCH starts m*256 chips after the start of the uplink DPCCH with m selected by the UE such that the ACK/NACK transmission (of duration 1 timeslot) begins within the first 0-255 chips after 7.5 slots following the end of the received HS-DSCH. The UE processing time (indicated as τ_UEP ) is therefore maintained at 7.5 slots (5.0 ms) as the offset between DPCCH and HS-DPCCH varies. The ACK bit is sent on the first slot of the code multiplexed uplink HS-DPCCH. Every first slot on the HS-DPCCH, chosen according to the parameters above, are reserved for ACK/NACK signalling (marked with A/NA in Fig. 5) . The other two slots on the HS- DPCCH can be used for CQI transmission (marked with QI in Fig. 5). T_siot indicates the duration of one slot (i.e. 0.67ms) . From Fig. 5 it can be seen, that the receipt, that is the ACK or the NACK is sent after the data. Therefore, the receipt can overlap with a transmission gap even if the actual data do not .

If a part of an ACK/NACK signal overlaps with a transfer gap of uplink compressed mode, the user equipment is not requested to transmit it. Instead, the user equipment may discontinuously transmit (DTX: discontinuous transmission) the affected timeslot on the uplink HS-DPCCH. Furthermore the user equipment does not need to attempt to decode the transmitted packet, but is only requested to store the data of the corresponding HS-DSCH TTI in the virtual user equipment buffer to be able to combine them with data sent in subsequent TTIs.

The detailed operation of the scheduler in the base station will be vendor specific, and it also needs not to be specified totally in detail. Specific implementations of the scheduler can be done by those versed in the art in accordance with the rules set forward herein. All that needs to be specified, is the fact that the user equipment is not requested to transmit any receipt, e.g. a ACK/NACK signal, that overlaps an uplink compressed mode transfer gap. Note that again not even a single chip of overlap is allowed. As a consequence, it will be invisible from the outside, whether the user equipment could correctly decode the HSDPA transmission and therefore the user equipment does not even have to attempt to decode this packet, all it has to do is store the soft decision values in the virtual user equipment buffer where they will be combined with the next transmission which will then be processed as usual, including transmission of an ACK/NACK Signal. This is the intention behind the second embodiment, as described above. Restrictions on QI transmission due to uplink compressed mode

Similar restrictions due to uplink compressed mode as for the receipt (e.g. ACK/NACK signal) affect also any other uplink HSDPA transmission, i.e. the Quality Indicator (QI) transmission.

The QI conveys information regarding the quality of the downlink channel as seen from the user equipment and is used by the base station to detect, which of the user equipments attached to a base station have a good reception and are therefore suited for HSDPA transmission.

Again, during the transfer gaps for uplink compressed mode, the user equipment is not able (and therefore not requested) to transmit anything, in particular not the QI transmission.

The QI transmission is intermittent or not continuous due to the measurement feedback cycle, even if it is possible to request the user equipment to transmit a QI in every TTI as a special case.

Further more the base station anyhow has to cope with corrupted or undecodable QI transmissions, so no special actions or provisions are expected to be necessary for the QI transmission, it can simply be omitted if there is a collision with an uplink compressed mode transfer gap. Note again that not even a single chip of overlap is allowed.

To sum up, the user equipment is not requested to transmit a Quality Indicator signalling, if a part of it overlaps with a transfer gap of uplink compressed mode. Instead the user equipment may simply not transmit anything during the affected timeslots on the uplink HS-DPCCH. This behaviour is also called DTX (Discontinuous Transmission) .

Restrictions due to simultaneous uplink and downlink compressed mode

If both uplink and downlink compressed mode are activated at the same time, or if multiple compressed mode patterns are activated, then the restrictions due to the individual transfer gaps exist in parallel. This means a downlink or uplink transmission is only feasible, if it is compatible with all the transfer gaps. Thus any receipt is not to be sent in any transfer gap.

Further aspects and embodiments

In the above text, it is proposed to simply not transmit an uplink transmission (any receipt, e.g. ACK or QI) if it would overlap with a transmission gap.

As a further embodiment, the affected transmission can also be delayed until a time, when it is possible to do the transmission. The delay can in principle take any value, e.g. the delayed transmission could start immediately after the, end of the transfer gap. However, in order to ease the implementation, it may be more desirable to delay the transmission by an integer number of timeslots. In case that more than one data packet has been transmitted during a compressed mode operation, that is more than one receipt is being delayed, the individual receipts are sent after the delay with an interval of one timeslot instead of one TTI to catch up for the delay faster than it would be the case with an interval of one or more TTIs.

In another embodiment of the invention, the delayed transmission is not delayed by an integer number of timeslots, but by an integer number of TTIs. This will ease the implementation at both the user equipment and base station, because the delayed transmission will then be received at a time when this type of transmission is anyhow due. Therefore the entire transmission respectively reception chain can be easier implemented. At first sight delaying for the next TTI instead of slot looks like introducing yet another delay. However, the entire system is designed and implemented that it works optimum for the given delays, i.e. when the transmission is sent at the nominal time. Any further delay will mean, that the action which is done in response to the ACK/NACK or the QI will not be possible any more at the time without any delay, but only at a later TTI. This action will in particular be the scheduling decision on the next TTI, i.e. which user equipment to transmit user data, which modulation and coding scheme to use and whether to transmit a new packet or to repeat an old one. If the ACK/NACK command is delayed by an integer number of TTIs, then obviously the mentioned response can also only be sent an integer number of TTIs later. If the delay is however less, i.e. an integer number of TTIs minus a fraction of a TTI, then the response can not be sent one entire TTI earlier, because the ACK/NACK came after the ^deadline' for that TTI. In other words, trying to optimise the delay of a transmission of an ACK/NACK (likewise for the QI which also influences subsequent transmissions) by a fraction of a TTI is pointless. As the implementation of the delay of an integer number of TTIs is easier, that should therefore be preferred.

The latter is particularly the case if there is simultaneously a transfer gap in both uplink and downlink compressed mode, because the backlog of the outstanding ^* ACK/NACK signals which was built up before the transfer gap can be finished after the transfer gap before new ACK/NACK signals associated with HSDPA transmissions after the transfer gap have to be transmitted. Basically the reason for this is that the HSDAP downlink transmission is longer than the ACK/NACK transmission, therefore the transfer gap necessarily blocks at least as many downlink transmissions as ACK/NACK transmissions. However, in uplink only compressed mode, i.e. if there are Gaps only in the uplink direction, this is not the case because no downlink frames are blocked. In order to avoid a backlog also in this case, it is proposed to simply not transmit the ACK/NACK signal or, alternatively, to transmit the ACK/NACK signal with a delay which is not necessarily a multiple of a TTI. Still from an implementation point of view it is desirable to select a multiple of a timeslot. For the same reasons as detailed above it is not advantageous to select a delay which is not a multiple of a timeslot: None of the transmitted ACK/NACK signals would be received earlier than a fraction of a timeslot earlier compared to the slot aligned case. As has been detailed above, even a delay reduction of a fraction of a TTI is not beneficial for the system, this argument of course in particular applies for a fraction of a timeslot.

In a further embodiment the delay can be set to be a multiple of slots which is not at the same time a multiple of a TTI. In this way it can be avoided that delayed ACK/NACK transmissions block the timeslots which are allocated for current ACK/NACK transmissions which would then themselves have to be delayed.

In a further embodiment, only every second TTI can be used for transmission of delayed ACK/NACK signals, together with a new ACK/NACK signal which needs not to be delayed, leaving two timeslots of every other frame available for QI transmission.

In an even further embodiment, it is proposed to transmit first one or several delayed ACK/NACK transmission after the transfer gap, subsequently followed by QI transmission, in particular if a QI transmission would also have been scheduled during the transfer gap or during a delayed ACK/NACK signal.

In another embodiment, ACK/NACK and/or QI transmissions are sent after the transfer gap, possibly at a time when other ACK/NACK signals or QI transmissions would be scheduled. In this case these latter transmissions are also delayed, until they can be sent.

In a different embodiment, ACK/NACK transmission is prioritised over QI transmission, i.e. if both an ACK/NACK signal is pending and a QI transmission the latter is further delayed and the pending ACK/NACK is transmitted before of the QI transmission.

In a further embodiment, only a single QI transmission is sent if a QI transmission has been delayed until the nominal transmission time of a later QI transmission. Instead of transmitting two or more QI transmissions one after the other, it may be sufficient to send only one QI . Transmitting two QIs in short succession will anyhow be redundant, because the channel quality will not have changed significantly during such a short time.

In a further embodiment, a QI transmission which cannot be sent at its nominal time due to a compressed mode transfer gap is delayed until a timeslot also allocated for QI transmission. Note that normally no QI may be sent during this particular timeslot because QI transmission is only done intermittently) . Basically this means that the QI transmission is delayed by an integer number of TTIs rather than an integer number of timeslots.

Generally spoken the following possibilities exist, that can be combined: a) the delay is such, that the receipt is sent immediately after the transfer gap. b)the delay has the length of an integer number of time slots. c)the delay has the length of an integer number of transmission time intervals (TTI).

The invention has been described in relation to HSDPA and compressed mode, however, as will be apparent to those skilled in the art, the same principles can easily be applied to other systems or scenarios, where a packet communication is established, but some time intervals are not available for uplink or downlink communication due to constraints from other aspects of the system. Such constraints can come from an other communication e.g. a circuit switched connection which runs in parallel. It can also be another system, which runs in parallel and is incompatible due to some reasons with the packet transfer and therefore causes some time intervals not to be available for the packet transfer. Such other system can be another communication system, which competes for some limited resources e.g. a receiver or causes mutual interference. It can also be a completely different action, which competes for e.g. energy sources or computing resources or other resources .

In Fig. 6 the transmission of data in a communications system or network according to the UMTS standard comprising a terminal or mobile station MS and a base station BS is schematically depicted. For example with the introduction of the HSDPA transmission there exists now in parallel the data transmission via a packet oriented channel and a continuous channel. Both channels have to provide or to establish transmission gaps during which the user equipment tunes to other frequencies. For a packet oriented channel the following aspects can be considered a) No transmitting of packets during the transmission gaps b) b) Receipts, e.g. ACKs or NACKs that are sent during the transmission gaps will not be received. Therefor, the further packet transfer has to be carried out without knowledge of that receipt.

Further aspects of the invention in respect of the UMTS standardisatio

In the last RAN WG1 (Radio Access Network Working Group 1) meeting one contribution [2] highlighted the interdependencies of compressed mode and HSDPA (High Speed Downlink Packet Access) . While the details of this contributions were left for further study, in particular the question whether the formulation might be too restrictive and the exact wording to be used in the specification, the conclusion from [2] was generally agreed (see [4]): During compressed frames (or gaps) on the DCH (Dedicated Channel) there is no HS-DSCH (High Speed Downlink Shared Channel) activi ty for the UE (User

Equipment) (incl uding signalling in DL/UL (Downlink /

Uplink) ) ,. exact details to be defined later. (Note :

Compressed mode is applied on DCH only) .

This contribution proposes how to restrict the scheduling of HSDPA transmissions in order not to compromise compressed mode operation and still sacrifice as little HSDPA throughput as possible. At the same time a compact formulation to be used in the Technical Report [5] and the working change requests (CR) for the specifications 25.2xx is proposed.

Restrictions on HSDPA transmission due to downlink compressed mode

The restrictions which have to be accepted for (downlink) HSDPA transmission are actually quite simple to describe: During the gaps provided by compressed mode, the UE (User Equipment) is free to tune to other frequencies and therefore cannot (and consequently also needs not) receive any signals on its currently assigned frequency. We therefore propose to specify exactly this simple fact.

Admittedly the arithmetic in the most general case, which TTIs (Transmission Time Interval) would be allowed or disallowed will be somewhat more involved. In the contribution [2] a proposal has been made, but as neither the offset of the DPCH against the HS-DSCH and HS-SCCH (High Speed Synchronisation Control Channel) nor the specific details of the compressed mode patterns or the different frame structure types have been taken into account, it is unclear whether this approach is generally applicable. Anyhow we don't have to strive for such a general formula if we can also state the above mentioned simple fact.

For any given offset and for given parameters of the compressed mode, it will be quite easy to calculate whether HS-SCCH or HS-SDCH overlap with the gap. Note that an overlap is already given if any single chip of the HSDPA transmission would overlap with a any single chip of the gap. Further more, because the HI transmission always fully overlaps the corresponding HS-SCCH transmission, it is sufficient to consider only the latter.

Restrictions on HSDPA transmission due to uplink compressed mode

The restrictions due to uplink compressed mode are somewhat different, because it is not the downlink HSDPA transmission that is causing the trouble but the uplink ACK/NACK (Acknowledge / Not acknowledge) transmission. Of course, during the gaps for uplink compressed mode, the UE is not able (and therefore not requested) to transmit anything, in particular not the ACK/NACK signal. However, this does not strictly speaking rule out the possibility for the Node B to transmit an associated downlink HSDPA transmission before of the gap .

The only restriction a Node B has to take into account is the fact, that it cannot expect the UE to transmit either ACK or NACK in response to this HSDPA transmission. As a consequence, the Node B will not get an information about whether the UE has correctly received the packet an will therefore have to retransmit the packet in any case. At first sight this may look like a waste of resources, but actually the throughput can be increased: The Node B will not use a

MCS (Modulation an Coding Scheme) where there is a fair chance that the UE will decode the initial transmission, but it will use a MCS where the second transmission has such a fair chance. Then of course the initial transmission will have a rather low success probability, which means the

ACK/NACK signal that is lost (more precisely that is not sent) due to the uplink compressed mode gap is most likely a NACK signal, so the information would anyhow be redundant. Note that the scheduler is anyhow free to always use such an operating point, which also has the advantage of increased time diversity due to combination of two transmissions at different times. The throughput of a higher MCS with retransmission is anyhow very similar to the throughput of a MCS carrying half as much payload for an initial transmission. If this scheme is employed, uplink compressed mode does not degrade the throughput of a HSDPA session.

From a specification point of view it is again easy to specify such a behaviour: The detailed operation of the scheduler in the Node B is of course out of the scope of the specification but will be vendor specific. All that needs to be specified, is the fact that the UE is not requested to transmit an ACK/NACK signal that overlaps an uplink compressed mode gap. Note that again not even single chip of overlap is allowed. As a consequence, it will be invisible from the outside, whether the UE could correctly decode the

HSDPA transmission and therefore the UE does not even have to attempt to decode this packet, all it has to do is store the soft decision values in the virtual UE buffer where they will be combined with the next transmission which will then be processed as usual. Restrictions on QI transmission due to uplink compressed mode

Similar restrictions due to uplink compressed mode as for the ACK/NACK signal of course also affect other uplink HSDPA transmission, i.e. the Quality Indicator (QI) transmission. Again, during the gaps for uplink compressed mode, the UE is not able (and therefore not requested) to transmit anything, in particular also not the QI transmission.

The QI transmission is intermittent anyhow due to the measurement feedback cycle (even if it is possible to request the UE to transmit a QI in every TTI as a special case) . Further more the Node B anyhow has to cope with corrupted or undecodable QI transmissions, so no special actions or provisions are expected to be necessary for the QI transmission, it can simply be omitted if there is a collision with an uplink compressed mode gap. Note again that even a single chip of overlap is not allowed.

Restrictions due to simultaneous uplink and downlink compressed mode

If both uplink and downlink compressed mode are activated at the same time, or if multiple compressed mode patterns are activated, then the restrictions due to the individual gaps exist in parallel. This means a downlink or uplink transmission is only feasible, if it is compatible with all the gaps .

From a specification point of view this doesn't need to be highlighted explicitly because it is generally the case that an implementation has to adhere to all sections of the specifications and not only to subset that may be more pleasant to adhere to. Conclusion

While the specification of the interaction of compressed mode with HSDPA operation looks quite complicated at first sight, the necessary rules can actually be incorporated quite simply into the specifications. A corresponding text proposal for the TR and the working CRs for the actual specification is provided above .

Text proposal for TR 25.858 and the working CR

It is proposed to specify the behaviour of the UE in a new section of 25.858 and the specification. The decision which chapter number to use in the TR 25.858 and which chapter in which specification is left to the editors, this will not be a crucial decision anyhow.

Operation of HSDPA during compressed mode

The Node B shall not schedule a HSDPA transmission for a UE which is using compressed mode, if part of the HS-SCCH information or the corresponding HS-DSCH TTI overlaps with a gap of downlink compressed mode.

If a part of an ACK/NACK signal overlaps with a gap of uplink compressed mode, the UE is not requested to transmit it. Instead, the UE may DTX the affected timeslot on the uplink HS-DPCCH. Further more the UE does not need to attempt to decode the transmitted packet, but is only requested to store the data of the corresponding HS-DSCH TTI in the virtual UE buffer to be able to combine them with data sent in subsequent TTIs .

The UE is not requested to transmit a Quality Indicator signalling, if a part of it overlaps with a gap of uplink compressed mode. Instead the UE may DTX the affected timeslots on the uplink HS-DPCCH.

Further remarks

In the above text, it is proposed to simply not transmit an uplink transmission (either ACK or QI) if it would overlap with a gap. As a ' further refinement, the affected transmission can also be delayed until a time, when it is possible to do the transmission. The delay can in principle take any value, e.g. the delayed transmission could start immediately after the end of the gap. However, in order to ease the implementation, it is more desirable to delay the transmission by an integer number of timeslots.

In a further embodiment, it is proposed to transmit first a delayed ACK/NACK transmission after the gap, subsequently followed by QI transmission.

In a further embodiment, ACK/NACK and/or QI transmissions are sent after the gap, possibly at a time when other ACK/NACK signals or QI transmissions would be scheduled. In this case these latter transmission are also delayed, until they can be sent.

In a further embodiment, ACK/Nack transmission is prioritised over QI transmission, i.e. if both an ACK/NACK signal is pending and a QI transmission the latter is further delayed and the ACK/Nack is transmitted.

In a further embodiment, only a single QI transmission is sent if a QI transmission has been delayed until the nominal transmission time of a later QI transmission.

In a further embodiment, a QI transmission which cannot be sent at its nominal time due to a compressed mode gap is delayed until a timeslot which is also allocated for QI transmission (note that normally no QI may be sent during this particular timeslot because QI transmission is only done intermittently) . Hence, in other words, generally the invention focuses on

a transmission method, where information is transmitted in packets, and where transmission or reception gaps are provided, whereby no transmission or reception is performed during transmission or reception gaps .

From the perspective of a base station various embodiments focus on a data transfer method as described above, whereby a packet is sent via the packet oriented channel, even in those cases where the corresponding receipt cannot be sent after a first processing time because the receipt would overlap with a transmission gap. From the perspective of a mobile station it can be seen as a data transfer method according as described above, whereby a packet is received via the packet oriented channel, even in those cases where the corresponding receipt cannot be sent after a first processing time because the receipt would overlap with a transmission gap.

From the perspective of networks a communications network is proposed, that is adapted to perform a method as described above, said communications network comprising at least one base station and one mobile station .

References

[1] Rl-02-0492 TR 25.212v4.3, section4.4 'compressed mode' with Fig. 11

[2] Rl-02-0034, Samsung, 'Interaction between compressed mode and HSDPA' , Espoo, Finland, Jan. 2002

[3] Rl-02-0199, TR 25.858 'High Speed Downlink Packet access', Espoo, Finland, January, 2002

[4] Rl-02-0356, Secretary, 'Revised minutes of TSG RAN WG1 #23 meeting', Orlando, Florida, U.S.A., Feb. 2002

[5] Rl-02-0199, TR 25.858 'High Speed Downlink Packet access', Espoo, Finland, January, 2002

The cited documents are maintained by 3GPP, the third generation partnership project, Address: ETSI, Mobile Competence Centre, 650, route des Lucioles, 06921 Sophia- Antipolis Cedex and are cited in the format used by this organisation.

Claims

1. A data transfer method wherein data are transmitted in packets between a mobile station (MS) and a base station (BS) , wherein no transmission or reception of data is performed during transmission or reception gaps, that are provided for a mobile station to tune to other frequencies.

2. The data transfer method, in particular according to claim 1,

- wherein data are transferred in parallel via a packet oriented channel in packets and a continuous channel

- between a mobile station (MS) and a base station (BS) ,

- where the transfer via the continuous channel is interrupted whereby at least one transfer gap is formed.

3. The data transfer method according to claim 1 or 2, characterised in that, the transfer gap is formed by an interruption of the reception at the mobile station.

4. The data transfer method according to any of the previous claims, characterised in that, the transfer gap is formed by an interruption of the transmission of the mobile station.

5. The data transfer method according to any of the previous claims comprising the following steps: - Sending a packet via the packet oriented channel

- Storing information obtained from the received data packet without sending a receipt after a first processing time.

6. The data transfer method according to claim 5, characterised in that a receipt is sent after a delay.

7. The data transfer method according to claim 5, comprising the following steps:

Receiving a second data packet,

Storing information obtained from the received second data packet

Performing a decoding making use of at least part of the information of each data packet.

8. The data transfer method according to any of the previous claims, where the packet oriented channel _ is a High Speed Physical Downlink Shared Channel.

9. The data transfer method according to any of the previous claims, where the certain time, the data transfer is interrupted, is determined by the base station or the mobile station.

10. The Data transfer method, in particular according to any of the previous claims, comprising the following steps: transferring information in packets, providing transfer gaps, performing data transfer only outside the transfer gaps.

11. The data transfer method according to any of the previous claims 2-4 or 8-10, wherein at least a packet is sent from the base station to the mobile station via the packet oriented channel at allowed time periods, wherein all time periods are allowed for which the packet transmission does not overlap. with a transmission or reception gap.

12. The data transfer method according to any of the previous claims 2 -4 or 8-10, wherein at least a packet is received by the mobile station regardless whether the sending of the receipt overlaps with a transmission gap.

13. A base station adapted to perform a method according to any of the preceding claims.

14. A mobile station adapted to perform a method according to any of the claims 1 to 12.

15. A communications network comprising at least one base station according to claim 13 and a mobile station according to claim 14.