GB2454198A - A communication system comprising a communication terminal using a non-orthogonal or hybrid orthogonal/non-orthogonal uplink waveform. - Google Patents

Abstract

A communication system comprising a base station 1 and a plurality of terminals is capable of operating with a non-orthogonal uplink waveform; or a hybrid uplink waveform, comprising a non-orthogonal waveform [i.e. high speed uplink packet access (HSUPA)] with a more orthogonal waveform [i.e. single carrier frequency division multiple access (SC-FDMA)]. The invention may be applied to any combination of orthogonal with non-orthogonal waveforms, e.g. orthogonal CDMA with non-orthogonal CDMA or synchronous E-DCH with non-orthogonal CDMA. The base station further comprises a scheduler, wherein the scheduler set separate rise over thermal target levels for the non-orthogonal and hybrid waveforms. Non-orthogonal RoT typically 3 - 6 dB, orthogonal RoT typically 10-20 dB.

Description

IMPROVED UPLINK

This invention relates to systems and methods of operating a hybrid non orthogonal/orthogonal uplink.

There is a desire to introduce a greater degree of orthogonality into wideband code division multiple access (WCDMA) uplink in order to improve system performance, whilst providing the ability for co-existence with legacy user equipment (UE) in the same cell and minimising changes to UE and network functionality required for its implementation.

In accordance with a first aspect of the present invention, a communication system comprising a base station and a plurality of terminals, wherein the system is capable of operating with a non-orthogonal uplink waveform; or a hybrid uplink waveform, comprising a non-orthogonal waveform with a more orthogonal waveform; the base station further comprising a scheduler, wherein the scheduler set separate rise over thermal target levels for the non-orthogonal and hybrid waveforms.

Preferably, the scheduler operates at the non-orthogonal RoT level when there are active non-orthogonal terminals in the system, or serviced by a neighbouring base station.

Preferably, the more orthogonal waveform includes synchronisation and a common scrambling code.

Preferably, the more orthogonal waveform is an orthogonal data waveform.

Preferably, the non orthogonal waveform is a control waveform.

Preferably. the more orthogonal data waveform is one of SC-FDMA, or synchronous E-DCH uplink and the non-orthogonal waveform is a WCDMA waveform.

In accordance with a second aspect of the present invention, a system comprises a base station and a plurality of terminals at least one of which is adapted to operate a non-orthogonal uplink and another is adapted to operate a hybrid non orthogonal/orthogonal uplink; wherein a scheduler in the base station operates with first and second Rise over Thermal (RoT) targets: the first RoT target for use when there are no non-orthogonal terminals active and the second RoT target for use when there are active non-orthogonal terminals.

Preferably, the second RoT target is higher than the first.

I

Preferably. when operating the hybrid uplink. the scheduler sets power control signal to interference ratio (SIR) targets and scheduling grants such that dedicated physical control channel (DPCCH) SIR targets are met whilst the second RoT level is S in use.

Preferably, the base station checks whether any neighbour base station is servicing non orthogonal terminals before switching to operate at the second RoT level Preferably. the base station informs its neighbour base station if a non orthogonal terminal is to be activated or scheduled, in order that the neighbours can lower their RoT targets.

Preferably, the change in RoT is achieved by means of a control message common to all active terminals that raises the transmit power of at least the orthogonal part of a hybrid uplink waveform.

Preferably, the non-orthogonal terminal is a WCDMA terminal.

Preferably, the hybrid terminal is a WCDMA/SC-FDMA terminal; a WCDMA/orthogonal CDMA terminal; or an WCDMA/synchronous E-DCH terminal.

In accordance with a third aspect of the present invention, a method of allocating traffic in a terminal adapted for transmitting orthogonal and non-orthogonal waveforms in a single transmit time interval comprises determining a priority class for the traffic; and mapping high priority traffic to the ron-orthogonal waveform for autonomous transmissions and low priority traffic to the orthogonal waveform.

Preferably, subsequent transmissions, or retransmissions of the high priority traffic are scheduled on the orthogonal waveform.

In accordance with a fburth aspect of the present invention, a system comprises two transmission waveforms; a scheduled orthogonal waveform and a non orthogonal waveform whereby at least the first transmission of high priority traffic is transmitted in an unscheduled manner on the non orthogonal traffic whilst background data is transmitted on the orthogonal waveform Preferably, the orthogonal and non orthogonal waveform data are ACKJNACKed independently Preferably, a single ACK/NACK is sent based on reception of both the orthogonal and non orthogonal waveforms.

Preferably, priority is given to the non orthogonal waveform over the orthogonal waveform in TFC selection Preferably, the high priority data is mapped to the orthogonal waveform if the waveform is scheduled for a first transmission with sufficient resources to cover the S high priority PDU and otherwise is mapped to the non-orthogonal waveform.

Preferably, the orthogonal waveform is SC-FDMA and the non-orthogonal waveform is WCDMA.

Preferably, the orthogonal waveform is orthogonal CDMA and the non-orthogonal waveform is non-orthogonal CDMA.

In accordance with a filth aspect of the present invention, a system comprises a plurality of base stations and at least one user terminal; wherein one base station assigns resources to the terminal and the other base statiot receive a signal transmitted from the terminal; wherein the terminal transmits a composite of a wideband control signal and a data signal: wherein the terminal is allocated resources by the serving base station; wherein the wideband control signal, which has UE specific scrambling, indicates the resources allocated for the narrowband data signal. which has serving cell specific scrambling.

Preferably. the wideband control signal is 3GPP WCDMA and wherein the narrowband data signal one of SC-FDMA, with resources that are scheduled being time/frequency blocks: or CDMA, with resources that are scheduled being OVSF codes: or synchronous wideband CDMA, with resources that are scheduled being OVSF codes.

Preferably. the control information that indicates the resources for the data channel is located on the E-DPCCH channel of the WCDMA signal.

Preferably, the control information indicates that resources for the data channel are transmitted at the start of a transmission time interval in accordance with a sixth aspect of the present invention, a method of distributing data channel transmission time and frequency information for a terminal in soft handover, the method comprising receiving at the terminal a scheduling allocation from a serving base station: transmitting to all base stations within range control information on a non-orthogonal waveform and data on an orthogonal waveform; wherein the data channel information is transmitted on the non-orthogonal waveform using terminal specific scrambling.

In accordance with a seventh aspect of the present invention, a method of intel-cell interference management in a communication system comprising at least one base station of a cell, at least one user device and a radio network controller, the method comprising scrambling an uplink single carrier cyclic prefix (SCCP) transmission for all user devices in a cell with a cell specific scrambling code: providing scrambling code information to neighbour cells: wherein each base station scans for transmissions from users in it's cell and for neighbouring cell specific SCCP scrambling codes; measures interference due to the SCCP transmission with a given scrambling code and reports the interference and its source to the radio network controller.

Preferably. the communication system is UMTS LTE.

Preferably, a plurality of cell specific scrambling codes are allocated to each cell dependent upon additional parameters.

In accordance with an eighth aspect of the present invention, a communication system comprises a plurality of cells, each cell comprising a base station and a plurality of user devices, the system further comprising a radio network controller; wherein each cell is allocated at least one cell specific uplink scrambling code by the radio network controller: wherein each user device scrambles uplink transmissions with the cell specific scrambling code for transmission to the base station of the user device; wherein each base station measures incoming interference, allocates the inteiference to a group according to the scrambling code and informs the radio network controller of the cell and associated interference.

Preferably, a cell specific or common pilot sequence is embedded in every uplink transmission.

Preferably. the base station is an eNode B, or a Node B. An example of systems and methods of operating a hybrid non orthogonal/orthogonal uplink will now be described with reference to the accompanying drawings in which: Figure 1 illustrates the general concept of the improvement according to the present invention; Figure 2 is a first embodiment of the present showing a first example of a cellular layout where cells sui-rounding a cell of interest do not contain any WCDMA UEs or have a low UL load and a cell that contains a WCDMA UE; Figure 3 illustrates management of RoT and power control in the target cell for the example of Fig.2; Figure 4 shows a second example of a network layout in which a cell next to a cell of interest now becomes overloaded; Figure 5 illustrates reduction of RoT and its impact on UE receiver power levels for the layout of Fig.4; Figure 6 illustrates a physical layer model for a second embodiment of the present invention; Figure 7 illustrates a first example using the model of Fig.6, for independent HARQ and no restriction: Figure 8 illustrates a second example using the model of Fig.6, for independent HARQ with power restriction: Figure 9 illustrates a third example using the model of Fig.6, for dependent HARQ and no restriction; Figure 10 illustrates a fourth example using tl model of Fig.6. for dependent HARQ with no power restriction; Figure 11 illustrates a fifth example using the model of Fig.6, for dependent HARQ with power restriction; Figure 12 illustrates a sixth example using the model of Fig.6, for a first transmission on SC-FDMA; Figure 13 illustrates an example soft handover scenario under consideration, in accordance with a third aspect of the present invention; Figure 14 shows the structure of a terminal signal in the scenario of Fig. 13; Figure 15 is an example of resource block partitioning for the SC-FDMA data part of the signal of Fig. 14; Figure 16 shows a terminal signal at time TO for the scenario of Fig. 13; Figure 17 shows a terminal signal at time Ti for the scenario of Fig. 13; Figure 19 illustrates an unknown source of inteiference; Figure 20 illustrates a fourth aspect of the present invention, where the source of interference is known: and, Figure 21 illustrates an uplink SCCP transmission within the WCDMA band for the system of Fig.20.

One option to improve orthogonality is using a scheme whereby data is transmitted using a single carrier frequency division multiple access (SC-FDMA) waveform, in a similar manner to that of universal mobile telecommunications system (UMTS) terrestiial radio access network (UTRAN) long term evolution (LTE), whilst the layer 1 (Li) control and pilot is canied using existing WCDMA channels, as illustrated in Fig. 1. Between a transmitter and mobile terminal 2, downlink 3 operates as per release 7 downlink, but an improved release 8 uplink is provided in which a data part is sent 4 using SC-FDMA and control signals are sent 5 on the dedicated physical control channel (DPCCH) of high speed uplink packet access (HSUPA).

The use of SC-FDMA enables uplink frequency division orthogonality and also provides some of the other gain mechanisms of UTRAN LTE, for example frequency domain scheduling, inteiference coordination and multiple input/multiple output (MIMO). The use of WCDMA for the control channels enables the high speed downlink packet access (HSDPA) control channel, i.e. high speed dedicated physical control channel (HS-DPCCH) to be received at the basestation with zero functionality modifications and leaves the Release 99 power control loops unaltered, assuming that the SC-FDMA data pail follows the WCDMA power control level.

Li functionality from both LTE and WCDMA are reused and hence an Li air interface can be provided that does not require many functionality updates for a dual mode LTE/WCMDA terminal. Such an air interface allows operators an intermediate step when migrating from WCDMA to LIE, in which the operator's spectrum still provides WCDMA to legacy users, but is capable of using LIE ibr dual mode terminals in order to offer user and system benefits. Such an approach builds confidence for consumers in buying dual mode terminals and for operators in investing in LTE technology.

Although link level Li functionality is re-used, the LI, L2 and L3 management of the air interface differs from both WCDMA and LTE. For example, unlike LTE, rise over thermal (RoT) must be carefully managed by a scheduler in order to maintain coverage for WCDMA users whilst unlike WCDMA, the scheduler can manage orthogonal resources on a transmit time interval (111) by TTI basis. Hence from a system point of view aspects of both LTE and WCDMA are incorporated.

The invention relates to a system that combines a non orthogonal control waveform with an orthogonal data waveform. Examples of the invention are described with respect to a combination of SC-FDMA (Data) plus WCDMA (Control) for achieving improved uplink performance. However, the invention is not limited to this example and also applies to alternatives, such as orthogonal code division multiple access (CDMA), oi WCDMA (Control) plus Orthogonal CDMA (Data).

In a non-orthogonal WCDMA system, the radio network controller (RNC) and HSUPA scheduler and power control algorithms need to maintain a tight control of uplink interference, often quantified as so called Rise over Thermal" in order to maintain stability of the power control loops and cell coverage. Operators have generally planned their networks with a specific RoT level, and hence in order to provide coverage to legacy terminals, the RoT level must be maintained. RoT levels are typically in the range 3 to 6dB.

Typically, the coverage and RoT in a cell is planned to allow access to a reasonable data rate even at the cell edge and the RoT may be somewhat lower than is necessary for a terminal to make basic radio access (RACI-I access.

In an orthogonal system such as LTE. on the other hand, since the system is intercell interference limited, RoT can be maintained at much higher levels, typically 10-20dB. Maintaining higher RoT levels is preferable to lower RoT levels, since at low RoT levels the system performance begins to become influenced by thermal noise as well as inter system interference.

A hybrid orthogonal mon orthogonal system, such as SC-FDMA/WCDMA, must maintain stability and coverage for legacy non-orthogonal terminals. However performance of the orthogo nal waveform is limited by the RoT level. Thus, the first aspect of the present invention aims to optimise the RoT operating point based on the cell traffic level and type.

In a non-orthogonal system, such as WCDMA, RoT is tightly controlled and cannot rise above the planned level, otherwise instability may occur. In a typical deployment, there is little scope for attempting to raise the RoT beyond 7 to 10dB, as the uplink will reach its pool capacity. The RoT is controlled by the scheduler by means of adjting "grants" of Enhanced Dedicated Physical Data Channel (E-DPDCH)/DPCCH power ratio.

In LTE. RoT is not as critical and is typically higher and not as tightly controlled, although a slow uplink power control is present. Hence WCDMA terminals would not be able to co-exist in the same spectrum as LTE.

According to the present invention, the scheduler is able to operate at at least two RoT levels. RoT! and RoT2. RoT! is a typical non-orthogonal WCDMA RoT and is operated when there ai-e active WCDMA terminals in the cell. RoT2 is an RoT target that may be operated when WCDMA terminals are not active.

When the Node B has no active WCDMA terminals. or has not scheduled WCDMA terminals, then it operates the SCDMA control/WCDMA data terminals at a higher RoT leel. By manipulating the power control loops and the scheduling grants, the Node B can maintain signal to interference ratio (SIR) on the pilot channels whilst providing higher RoT for the data. Similarly, the SIR target at least for the control channels of the WCDMA terminals can be maintained.

RoT2 is higher than RoT!, but still sufficiently low that RACH coverage for WCDMA is maintained to the edge of the cell.

Once WCDMA terminals become active and scheduled, the Node B scheduler reverts back to Roll. Alternatively, the Node B may temporarily reduce the RoT following a RACH signature in order to enable sufficient coverage for the subsequent RACH message.

It is possible that WCDMA terminals may become active in a neighbour cell.

Thus, neighbour cells must have a mechanism for preventing a cell that is operating at RoT2 from overloading their own cells. This may be achieved through overload bits" and power control. Alternatively (and preferably), before switching from RoTI to RoT2, a protocol is defined whereby a Node B checks with its neighbours whether they have active WCDMA terminals (or alternatively, a low load). Similarly, when a cell plans to activate a WCDMA terminal and its load level is near to RoT!, the cell may indicate to its neighbour cells that they should not operate with RoT2.

Figure 2 indicates an exanple of a network deployment. Cell A represents a cell of interest, which contains no UEs operating WCDMA. Cell A requests that the surrounding cells clarify whether they can tolerate cell A operating with a higher RoT This may be done via the RNC or, in a decentralised network, directly between the nodes of each cell. In Fig. 2, cells B, C. D, E represent cells that do not contain UEs that are operating WCDMA, whilst cells F, G indicate cells that, although they contain WCDMA UEs have a low loading. Cell H is the nearest loaded cell with WCDMA UEs. but cell H is a sufficient distance from cell A that it is not likely to suffer from significant loading from an increase in RoT in the cell A. In an RNC based network, the RNC can manage the permitted RoT levels of the cells in a manner consistent with their loading. In this case, the cells can also regularly report their loading to the RNC, which manage s the RoT targets for the individual schedulers. This necessitates reporting of loading to the RNC and the RNC managing RoT targets. Alternatively, if there is no RNC present. then cells can request that each of their neighbours report their loading status before adjusting their RoT levels.

The procedure by which the RoT is raised is indicated in Fig. 3. The starting position is indicated in Fig. 3a, where the cell is operating at a RoT level 10 RoTI. The iight hand part of the figure shows the received power level from 3 UEs, split between a DPCCH power part 11 and an SC-FDMA data power part 12. The SC-FDMA parts of the UE's signals are orthogonal to one another in the frequency domain. In Fig. 3b, the cell RoT 10 is raised to RoT2. which allows the scheduled UEs more receive power at the Node B receiver 1, and hence a higher data rate by means of increasing the SC-FDMA/DPCCH ratio.

In the example of WCDMA (Control) + SC-FDMA (Data), the DPCCH is not orthogonal to the SC-FDMA and thus the increase in RoT and scheduled SC-FDMA leads to an increase in inteiference to the DPCCH channels. In order to maintain the DPCCH SIR level, the transmit power control now begins to raise the received power on the DPCCH (and hence also the SC-FDMA, whose power is set as a fixed offset from DPCCH). If the DPCCH interference to SC-FDMA is small, as may be the case in a cell loaded only with UEs operating in a dual SC-FDMA/WCDMA mode, then the increase in DPCCH power 13 may not lead to a substantial increase in SC-FDMA power 14. However the scheduler needs to redire the SC-FDMA/DPCCH power ratio in order to prevent the RoT 10 target being overshot. Although the SC-FDMA/DPCCH ratio is reduced, the received SIR on SC-FDMA is not likely to change too significantly. so the higher data rates (compared to when the system was being operated at RoT I) are maintained.

Operating at the raised RoT target enables a higher throughput from the target cell. However at a later stage, a neighbour cell E starts to serve WCDMA UEs and becomes overloaded, as indicated in Fig. 4. The overloaded neighbour cell E reports to its, neighbours (including cell A) that it is experiencing an overload (for example, via luB). Cell A is now instructed by the RNC to reduce its RoT 10 target back to RoT!.

Alternatively, in a network that does not contain an RNC, the overloaded cell E can report to its neighbours that it is overloaded, in order that for example cell A can reduce its RoT target and reduce the overload situation.

The procedure by which the cell reduces its RoT level is illustrated in Fig. 5.

The starting situation is depicted in Fig. 5a, in which the cell is operating at RoT 10 level, RoT2. In order to reduce the RoT, the scheduler reduces the SC-FDMA/DPCCH ratios, as depicted in Fig. 5b. The reduction in SC-FDMA power leads to an increase in DPCCH SIR, which leads to the power control reducing DPCCH RX power to maintain the power control target. To maintain the RoT at RoT I. the scheduler may not raise the SC-FDMA/DPCCH ratios, as indicated in Fig. 5c.

Signalling to support the described procedures may be set in standards, such as reporting of cell load and the amounts of orthogonal & non orthogonal UEs in a cell to an RNC or between Node Bs: signalling for a Node B to request its neighbours to clarify their load situation: signalling to the RNC or other Node Bs to indicate that an overload situation has atisen due to increased RoT in another cell; luB signalling to indicate which RoT target a scheduler should use from a set; scheduling grant signalling that increases/reduces the SC-FDMAIDPCCH power level for all UEs in a cell when the RoT target is to be raised/lowered.

The examples described assume a hybrid system, where a classical WCDMA system coexists with an orthogonal SC-FDMA system. Another option in synchronous E-DCH is the coexistence of the classical WCDMA system with a synchronized uplink WCDMA. By introducing synchronization and a common scrambling code, the uplink is not fully orthogonal, however it has improved orthogonality. Therefore the introduction of the second RoT target for "synchronized" users is still possible and brings similar benefits, only slightly less than in case of full orthogonality, as in the hybrid SC-FDMA/WCDMA system.

The invention to the new work scope, covers not only the coexistence between a non-orthogonal and an orthogonal system, such as WCDMA and SC-FDMA, but also coexistence of a non-orthogonal system with another system, which improves the orthogonality. rather than being fully orthogonal, as in the case of synchronous E-DCH UL.

A second aspect of the present invention relates to allocation of resources. An uplink packet scheduler generally has to have a capability for handling multiple types of uplink quality of service (Q0S). For example, a file transfer protocol (ftp) upload may require a background level QoS. whilst voice over internet protocol (V0IP) clearly requires conversational QoS and gaming QoS requirements may be even higher. Some types of traffic, e.g. gaming may have an unpredictable traffic model and require fast response from the scheduler, whilst other types of traffic, e.g. voice might require the multiplexing of many individual users at low bit rates. Furthermore, the terminals need to be able to transmit signalling radio bearers (SRBs) and other signalling information in the uplink which may arrive in an unpredictable fashion.

An orthogornl packet scheduler needs to become aware of high priority infon-nation becoming present in a terminal's buffer and allocate resources to the terminal. The scheduler needs to be able to handle such traffic without an increased reaction latency or a substantial amount of downlink signalling power.

Thus, the invention relates to a means for enabling fast response to conversational QoS and unpredictable traffic, whilst avoiding contention and minimising impact to the introduction of orthogonality.

In WCDMA, a minimum set of transport format combinatons (TFCs) are defined that the terminal is able to access regardless of whether it has a grant or not. If the terminal has no grant, then a transport format within the minimum set is used to send high priority information autonomously. If a grant is available, then the high pliolity information is sent as well as the rest of the information indicated by the grant, implying that all of the transmitted information has to be sent with high priority. A disadntage of this approach is that some system efficiency may be lost where there are low and high priority flows and the low priority flow often ends up getting multiplexed with the high priority one.

In LTE, it is difficult for the UE to make autonomous transmissions without compromising orthogonality. Hence either a significant amount of downlink (DL) signalling is required for handling traffic such as VoIP, or a so-called persistent scheduling" mechanism is required, which allows for pre-allocation of regular scheduling grants to the terminal. However the allocation of such grants may lead to inefficient resource usage and can lead to resource management problems when operating in combination with hybrid automatic repeat request (HARQ).

In the present invention the orthogonal part of the signal is used for most traffic, whilst the use of autonomous transmissions is permitted for high priority traffic.

The transmitted waveform from the terminal is allowed to carry both E-DPDCH and SC-FDMA in the same UI, ar E-DPDCH transmissions are unscheduled. Traffic is mapped to either the orthogonal SC-FDMA waveform, or the E-DPDCH waveform according to it's priority class for the first transmission. Hence, high priority traffic is mapped to E-DPDCH and can take advantage of its ability to carry out autonomous transmissions, whilst lower priority traffic is mapped to SC-FDMA in order to increase the amount of cell orthogonality.

For subsequent retransmissions. since it is known that there is high pliority traffic, the high priority traffic may be scheduled on the orthogonal SC-FDMA waveform. (Alternatively, retransmission can continue on E-DPDCH).

Transport format combination (TFC) selection is defined in a manner that ensures that the UE has sufficient power to transmit at least the E-DPDCH and HS-DPCCH traffic. HARQ may be operated either independently, or jointly for the two parts.

Description of specific examples below is made with respect to a hybrid WCDMA/SC-FDMA waveform, but more generally the invention applies to a transmission of a substantially orthogonal waveform and a non- orthogonal waveform Thus, other examples are orthogonal code division multiple access (CDMA)/non- orthogonal CDMA for data/control respectively; or synchronous E-DCH and non-orthogonal CDMA.

Implementation possibilities are described via a number of examples. Fig. 6 illustrates a physical layer model of the present invention. Background QoS traffic 20 and high priority traffic 21 are received. The background traffic 20 is always sent on the orthogonal waveform 22, in this example SC-FDMA and the high priority traffic can be sent on this waveform, but is more often sent on the non-orthogonal waveform 23, which in this example is E-DPDCH. For the example of CDMA. rather than a hybrid WCDMA/SC-FDMA, the background traffic is sent on the orthogonal CDMA waveform and the high priority traffic is sent on the non-orthogonal CDMA waveform.

In the WCDMAISC-FDMA example, in the physical layer are also enhanced dedicated physical control channel (E-DPCCH) 24; High Speed Dedicated Physical Control Channel (HS-DPCCH) 25 and dedicated physical control channel 26.

Example 1 in Fig. 7 illustrates this with independent HARQ and no power restriction. Two traffic flows, dedicated traffic channel (DTCH) DTCHI and DTCH2 are mapped to the E-DPDCH and SC-FDMA transport channels respectively. Two acknowledge/negative acknowledge (ACK/NACK) channels are provided; one relating to the E-DPDCH transport channel and the other to the SC-FDMA one. When data arrives from DTCH 1, it is queued and transmitted, or retransmitted independently of traffic on DTCH2 and vice versa. Once the first transmission is made on DTCH I. optionally, the retransmissions can be made on SC-FDMA, since the scheduler is aware of the need to schedule DTCH 1. DTCH I and DTCH2 are scheduled separate resources.

as they are ACK/NACKed separately and have diffeiing QoS levels. The UE makes a first transmission 27 on DTCH2, but this is not received correctly, so a NACK 28 is sent. Traffic 29 arrives for DId-I 1 and is transmitted on E-DPDCH. DTCH2 traffic is retransmitted 30. The Node B does not receive either correctly and sends a NACK 31, 32 for each. DTCH I is retransmitted 33 on the scheduled SC-FDMA, along with DTCH2 34. The Node B receives DTCH2 and acknowledges 36, but sends a NACK for DTCH I. Transmission of DTCH2 then continues alongside DTCH I on SC-FDMA.

Example 2 in Fig. 8 illustrates independent HARQ with power restriction. Two traffic flows. DTCHI and DTCH2 are mapped to the E-DPDCH and SC-FDMA transport channels respectively. Two ACKJNACK channels are provided; one relating to the E-DPDCH transport channel and the other to the SC-FDMA one. In this case, the UE has become power limited.

Data an-ives on DTCH2 and is queued. Whilst trarmission of DTCH2 data is ongoing, data arrives for DTCH 1. The terminal immediately starts transmitting data on DTCH 1. If DTCH2 is scheduled, the scheduler calculates the difference between the amount of power it requires for transmitting the WCDMA waveform and its maximum transmit power. If there is power available, then the SC-FDMA waveform is also transmitted using DTCH2.

If a retransmission on DTCH2 is pending, but due to transmission of WCDMA there is insufficient power to transmit using the initial TX power and the deficit is substantial, then either a first transmission may be made using a different HARQ process or the terminal may not transmit on DTCH2 in order to reduce system interference.

The UE makes a first transmission 40 on DTCH2, but the transmission is not received correctly and a NACK 41 is sent on the Enhanced HARQ indicator channel (E-HICH). Traffic arrives for DTCH 1 and is transmitted 42 on E-DPDCH. There is insufficient power to retransmit DTCH2. so the transmission is abandoned. The Node B does not receive DTCHI correctly and sends a NACK 43. DTCHI is retransmitted 44 on the scheduled SC-FDMA. DTCH2 cannot be transmitted with a DTCHI retransmission. The Node B receives DTCHI and sends an ACK 45. Transmission of DTCH2 restarts 46.

Example 3 in Fig. 9 illustrates dependent HARQ, with no power restriction.

Two traffic flows. DTCHI and DTCH2 are mapped to the E-DPDCH and SC-FDMA transport channels respectively. However, only a single ACK/NACK channel is provided. Data is being trarmitted on DTCH2 when traffic for DTCH 1 arrives. As soon as transmission of DTCHI commences, an ACK is only sent once a HARQ process has been successfully received on both DTCHI and DTCH2. The transmit power on DTCH2 is raised sufficiently to ensure that the latency is kept at the required level. Once the first transmission is made on DTCH 1 then the scheduler is aware that retransmissions are required for both DTCH 1 and DTCH2. and so schedules sufficient orthogonal resources to make the required retransmissions. Since there is a single ACK/NACK, DTCH I and DTCH2 are transmitted together on the same resources.

The UE makes a first transmission 50 on DTCH2, but the transmission is not received correctly, so a NACK 51 is sent. A retransmission 52 is made for DTCH2 and traffic arrives for DTCH 1 and is transmitted 53. The Node B does not receive either DTCH 1 or DTCH2 correctly. so a NACK is sent 54. Retransmissions 55 are made for the DTCH 1 and DTCH2 on the scheduled SC-FDMA. The Node B does not receive DTCH 1 and sends a NACK 56, but it does receive DTCH2. The UE retransmits 57 DTCHI and DTCH2. The Node B now receives DTCHI and sends an ACK 58.

Example 4 in Fig. 10 shows dependent HARQ, with no power restriction. Two traffic flows. DTCH 1 and DTCH2 are mapped to the E.-DPDCH and SC-FDMA transport channels respectively. However, only a single ACKJNACK channel is provided. Data is being transmitted on DTCH2 when traffic for DTCH 1 arrives. As soon as transmission of DTCH 1 commences, an ACK is only sent once a HARQ process has been successfully received on DTCH I only. DTCH2 continues to transmit with a fixed number of retransmissions.

The UE makes a first transmission 60 on DTCH2, but the transmission is not received correctly, so a NACK 61 is sent on the E-HICH. A retransmission 62 is made for DTCH2 and traffic arrives for DTCHI and is transmitted 63. The Node B does not receive DTCH I correctly, so a NACK is sent 64. Retransmissions 65 are made for the DTCH 1 and DTCH2 on the scheduled SC-FDMA. The Node B does not receive DTCHI and sends a NACK 66, but it does receive DTCH2. The UE retransmits 67 DTCHI and DTCH2. because DTCH2 has a fixed number of retransrnissions. The Node B now receives DTCHI and sends an ACK 68.

Example 5 in Fig. 11 shows dependent HARQ. with power restriction. Two traffic flows, DTCH1 and DTCH2 are mapped to the E-DPDCH and SC-FDMA transport channels respectively. However, only a single ACK/NACK channel is JO provided. Data is being transmitted on DTCH2 when traffic for DTCH I arrives. The terminal calculates the amount of power required to transmit both DTCH 1 and control and then allocates the remaining power to DTCH2. If DTCH2 requires a retransmission and there is insufficient power to match the first transmission, then the DTCH2 retransmission is abandoned.

The UE makes a first transmission 70 on DTCH2, but the transmission is not received correctly and a NACK 71 is sent on the E-HICH. Traffic anives for DTCHI and is transmitted 72 on E-DPDCH. There is insufficient power for the DTCH2 retransmission, so the transmission is abandoned. The Node B does not receive DTCHI correctly and realises that DTCH2 was abandoned, so sends a NACK 73.

DTCHI is retransmitted 74 on the scheduled SC-FDMA. The Node B receives DTCHI and sends an ACK 75. Transmission of DTCH2 continues 76.

Example 6 in Fig. 12 shows a first transmission on the SC-FDMA.

Two traffic flows, DTCHI and DTCH2 are mapped to the E-DPDCH and SC-FDMA transport channels respectively. Two ACK/NACK channels are provided; one relating to tbe E-DPDCH transport channel and the other to the SC-FDMA one. When data aiiives from DTCH 1, if there is a scheduling grant for DTCH2 that is sufficiently large to accommodate the DTCH 1 PDU, then DTCH I is mapped to the SC-FDMA waveform. If not, then DTCH 1 is mapped to the WCDMA waveform. If DTCH I is mapped to the SC-FDMA waveform, then it may take priority over DTCH2 retransmissions.

The UE makes a first transmission 80 on DTCH2. The transmission is not correctly received, so a NACK 81 is sent. Traffic arrives for DTCH1 and is transmitted 82 on SC-FDMA, as there is a sufficient grant. The DTCH2 transmission is abandoned. The Node B does not receive DTC1-I 1 colTectly, so a NACK 83 is sent. A retransmission 84 is made for DTCH I and a new DTCH2 PDU is transmitted.

All of these examples allow for at least the background traffic to benefit from orthogonality. whilst unscheduled transmissions and retransmissions can be made for DTCH I. Examples 4 and 5 allow for the background traffic to be sent with a lower QoS than DTCH 1, even when DTCH 1 is active. Example 6 maximises use of the orthogonal SC-FDMA waveform.

In a third aspect of the present invention, a terminal is in soft handover with several base stations capable of receiving the hybrid uplink signal. All of the base stations are capable of receiving the WCDMA control channels and one of the base stations is assumed to have been assigned the role of "serving" base station. This base station controls the timing advance and scheduling of the terminal. Other base stations receive and decode the orthogonal SC-FDMA data part of the signal and also issue TPC commands based on the DPCCH part of the signal. However the time and frequency at which the SC-FDMA data channel is transmitted depends upon the decisions of the scheduler in the serving base station and is unknown to the non serving base stations.

In a hybrid system, where non-orthogonal WCDMA is transmitted in parallel to synchronous E-DCH, the channelisation code to be used by a UE for UL sync-E-DCH transmission will also depend on the decision of the scheduler in the serving base station and will be unknown to the non serving base stations. This means that the information about the resource allocation should be sent from the UE in UL. This information is redundant for the serving base station, but not for the non-serving base stations.

Attempting to decode every possible combination of resource block allocations in order to discover within which SC-FDMA resource blocks data has been transmitted would be prohibitively complex for the non serving basestations. Therefore the invention provides a mechanism for informing the non-serving basestations of the decisions of the serving basestation In WCDMA HSUPA, several basestations can receive the E-DPDCH channel in a soft handover situation. However, the E-DPDCH is always transmitted over the whole frequency band and with the same scrambling code, thus there is no need to identify the resources on which the E- DPDCH is transmitted. GB2413242 describes a method in which the UE informs all basestations of the scheduler data rate restriction, however this prior art method differs from the present invention in that (i) data rate restrictions, rather than resource allocations are indicated: (ii) it is assumed that all basestations make scheduler decisions and the UE makes a decision on a composite scheduler decision and (iii) the information is used for the management of interference and complexity at basestations and not for knowing where the allocated resources are in order to peiform decoding. Hence GB24 13242 does not address the problem of informing non-serving base stations of the decisions of the serving base station.

In UMTS terrestrial radio access (UTRA) time division duplexing (TDD), LTE, GSM. Wimax & Wireless LAN. soft handover is not possible, and therefore the problem desci-ibed above does not arise. CDMA2000 does not include an UL signal that is partially orthogonal and partially non orthogonal. and so again the problem does not arise.

Upon receiving a scheduling allocation from the serving basestation, the terminal transmits control on WCDMA and data on SC-FDMA. Also transmitted is control information that indicates where the SC-FDMA transmission can be found.

This control information is decodable by all of the basestations. For the serving basestation the data is redundant, however for the non serving basestations the information enables the basestations to decode the SC-FDMA data part.

The control information must be transmitted wideband, so the information can be transmitted within the WCDMA part of the signal, for example using the existing E-DPCCH channel. Alternatively, the control information may be transmitted early within a T1'I, in order to enable the non serving basestations to decode the control and start decoding of the SC-FDMA before the end of the transmission time.

The signal may be a composite WCDMA/SC-FDMA signal, or a composite WCDMA/CDMA signal where the WCDMA part has a UE specific scrambling code and the CDMA part has a cell specific scrambling code and the resources that are scheduled are orthogonal variable spreading factor (OVSF) codes.

Figure 13 indicates a system comprising 3 basestations, BTS I, BTS2 & BTS3 and a terminal. BTS 1 is responsible for scheduling the terminal, BTS2 and BTS3 issue UL power control commands 91. 92. 93 to the terminal and also attempt to decode the data part. Each basestation also sends an ACK/NACK 94, 95, 96 to the terminal. As soon as the terminal receives an ACK from any basestation. it is assumed that a transmission has been received successfully.

Figure 14 indicates the structure of the transmitted signal from the terminal. The wideband WCDMA control signal 100 contains a DPCCH channel 101 with pilot bits and downlink TPC bits, an HS-DPCCH control channel 102 for DL HSDPA and an E-DPCCH channel 103, which contains the UL TFCI & HARQ information and also data assignment information. Also transmitted is a SC-FDMA data channel 104. which is transmitted on a set of frequency resources, indicated by BTSI. There are 4 frequency resources across the 5MHz of the WCDMA. as indicated in Fig. 15, split into resource blocks 105, 106, 107, 108.

At time TO, BTSI allocates the 1st and 2' frequency resources 105, 106 to the terminal 90. The transmitted signal from the terminal is shown in Fig. 16. A data assignment indicator 109 indicates 1100 that the 1st and 2nd resources have been used.

The non serving basestations decode the data resource indicator and are then able to decode the SC-FDMA. At time Ti, the BTS 1 allocates the and resource blocks 106, 107 to the terminal for SC-FDMA, as indicated by the transmitted signal in Fig. 17. The data resource indicator 109 now indicates 0110 that the 2nd and 3rd resources are used.

The invention thus enables frequency domain scheduling to be combined with soft handover. Without the use of the invention, the two would not be mutually compatible, hence the system would be able to support one or the othei but not both, and hence would not benefit from the gains of both.

With the introdtion of I 6QAM to High Speed Uplink Packet Access (HSUPA) use of a Wideband Code Division Multiple Access (WCDMA) system has become necessary to sustain a single high data rate transmitter at one time in a cell. At the same time Long Term Evolution (LTE) uplink based on orthogonal Single CalTier Frequency Division Multiple Access (SC-FDMA) has been developed by 3rd Generation Project Partnership (3GPP) where simultaneous uplink transmitters are allocated with orthogonal resources and thus do not interfere with each other.

To improve High Speed Packet Access (HSPA) uplink evolution beyond ReI-7 work has been done on the orthogonalisation of the HSPA uplink, in order to allow for multiple high data rate uplink transmitters to exist at the same time in the cell thus boosting the cell capacity. Potential synergies between the LTE and the orthogonalisation of the HSPA uplink might be available if the LIE transmitter and receiver structures could be utilized.

It is desirable to improve the HSPA uplink compared to release 7 (see Figure 1, which may in turn require downlink (DL) control channels to be modified or introduced. For example, for Single Carrier Cyclic Prefix (SCCP) development, the SCCP improvement is intended to provide an increase in uplink (IlL) throughput, with gains available across the whole of the cell, rather than just to peak rate only. As far as possible, any improvement to SCCP needs to evolve from HSPA, or be similar to LTE: so as to avoid the need for a new air interface. Desired latetry performance is as good as or better than Release 7 E-DCH with a 2msec TTI and a Peak-to-Average-Power Ratio (PAPR), should better than or similar to that in Release 7. It is desirable that cost effective receiver algoiithms are possible. along with operation with at least Quadrature Phase Shift Keying (QPSK) and I6QAM. Any impacts to the DL channels needs to be minimized.

Fig. 1 illustrates an improvement to the SCCP concept. Between a transmitter and mobile terminal 2, downlink 3 operates as per release 7 downlink, but an improved release 8 uplink is provided in which a data part is sent 4 using SC-FDMA and control signals are sent 5 on the dedicated physical control channel (DPCCH) of HSUPA.

This concept involves the generation of 3rd generation partnership project (3GPP) LIE uplink data in parallel with HSUPA/enhanced dedicated channel (E-DCH) control (i.e. Dedicated Physical Common Control Channel (DPCCH), Enhanced Dedicated Physical Control Channel (E-DPCCH), Enhanced Dedicated Physical Data Channel (E-DPDCH)), summing the two streams together (after separate scrambling) and producing a composite waveform for transmission to a suitable receiver. The receiver needs to be capable of receiving the composite waveform (in the presence of noise, interference etc) and separating the two streams back into their constituent parts.

Power control can be applied to the DPCCH and the SCCP power level can be defined as an offset to the DPCCH power level in the same manner as E-DPDCH in HSUPA. This allows for HSUPA like power scheduling and hence tight control of rise over thermal (RoT in a HSUPA like manner. However this also allows for time/frequency scheduling of the data part. The following gains arise. Introduction of intra-cell orthogonality e.g. by time, frequency or code separation of users; enhanced capability to mitigate multipath effects at lower complexity than WCDMA with

I

frequency domain equalisation. due to cyclic prefix; potential for channel sensitive time/frequency scheduling; interference coordination techniques to reduce the impact of inter-cell interference: enhanced potential for inteiference cancellation compared to Release 7; and ieduction of control or pilot overhead.

Due to the non orthogonal control channels, introduction of orthogonality and interference coordination/cancellation benefits may not be as great as pure SC-FDMA.

However in a cell supporting HSDPA too, the non orthogonal control channels may represent a minority contribution to total cell interference.

This invention allows for re-use of existing receiver functionality for DPCCH & HS-DPCCH and HSUPA like scheduling with additional dimensions.

An estimate has been made of the gain that can be achieved by introducing orthogonality in the UL for SCCP. The first set of estimations analyses the benefit of introducing intra-cell orthogonality. The orthogonality assumed in this analysis is perfect and may be FDM or CDM. The gain of an orthogonal uplink is in the range 10% to 60%. depending on the proportion of SCCP users. Even with 100% SCCP, capacity gains of 2 to 4 times, as seen for LTE are not achieved. The reason for this is the tight power control and restriction of the rise over thermal (RoT) to 4dB; LTE operates at much higher RoT levels and becomes effectively interference limited, rather than noise limited. In fact, if the RoT level is set to 15dB (a common figure for LTE).

the spectral efficiency improvement shown with 100% SCCP exceeds 100% and hence is comparable to LIE.

Using current technology, any given cell is not able to recognise the source causing UL inteiference, unless the inteiference is caused by users served by this particular cell. An example of this is shown in Fig. 18. In each of cell 1, cell 2 and cell 3, a mobile user X, Y, Z is transmitting, X and Z to NB 1 and Y to NB3. The ongoing transmission in cells I and 3 causes a certain amount of interference in cell 2 and to each other. Cell 2 recognises only the total amount of incoming interference 110 from neighbouring cells, but is not able to distinguish which part of this interference is caused by which cell.

In an (intra-cell) orthogonalised cellular system, the inter-cell interference dictates the capacity and cell edge coverage as well as the control channel reliability.

Effective management of intercell interference requires a knowledge of the relative levels of inteiference from surrounding cells. However in existing systems such as WCDMA and LTE, it is possible for a Node B to measure and distinguish intercell and intracell interference, but not to differentiate the sources of intercell interference.

In the fourth aspect of the present invention, the problem of how a cell can detect the amount of interference level caused by t1 transmission in individual neighbour cells is addressed as illustrated in Fig. 19. The transmission from users X, Y and Z produces inteiference 111, 112, 113, which is individually distinguishable at NB2.

The problem of intra-cell interference and recognition of its source is well known in all wireless systems. In some systems, interference is avoided in the system design phase; in other systems the means to controL or to mitigate the existing interference are introduced.

GSM is an example of a system where inter-cell interference is avoided by proper frequency planning. Neighbouring cells operate in different frequency ranges.

This same frequency may be reused only with a minimum distance (reuse distance), which minimises the probability of the inteiference.

In UMTS all cells operate in this same frequency range. The influence of the proper interference management on the cell, or even network performance is not to be underestimated. Therefore soft handover was introduced. By means of soft handover, a cell may control inteifering users from other cells, well before they are able to cause extensive interference and to kill resources in t1 controlling cell.

Furthermore, the multiple access technique used in UMTS is interference sensitive, placing more demands on the system. Three power control procedures are in place, together with admission control and load control procedures to assure in real-time, a sale level of the interference within the cell. LTE inherits a frequency reuse factor 1 from UMTS. Contrary to UMTS, the UL transmission within the cell is orthogonal, so no interference is caused. Nevertheless, there is still a problem of intra-cell interference. Unlike in UMTS there is no SHO in place, so other techniques of inter-cell interference management or cancellation had to be developed.

One approach is based on an overload indicator, where each cell broadcasts an uplink load indicator (busy or not) in the downlink in a peiiodic manner. Next, each UE decodes the load indicator bits from at least one dominant interfeiing cell (based on path loss measurements). Based on the decoded load indicator, the UE appropriately reduces its allowed transmitted power spectral density. Another approach is based on backhaul signalling, where each Node-B measures received interference level and sends the information to the neighbouring Node-Bs through backbone networks.

However, none of the above mentioned radio technologies, with frequency reuse factor I, allow a cell or Node B to determine the UL interfereme caused by UEs connected to a specific neighbouring cell.

In the present invention the UL SCCP transmission is scrambled for all users in a given cell with a cell specific scramble code, as shown in Fig. 20. Within the uplink DPCCH transmission band 115, in this example of 3.84 MHz. a band 116 is allocated to a UE for SCCP transmission. Within each resource block 117, the transmission is scrambled with elements 118 of the pilot sequence. It is assumed, that the special pilot sequence is embedded in every SCCP transmission, so that after descrambling the sequence is recognised. Information about the scrambling code may be either hard-coded during the network planning, or delivered from the RNC to all neighbour cells in a semi-static way. The pilot sequence may be either cell specific, or even SCCP specific -equal to all SCCP users in the whole network. Every NB or cell scars, not only for its own user's transmission, but also for the neighbouiing cell specific SCCP scrambling codes. The measurements on the interference caused by SCCP transmission with the given scrambling code are measured, summarised, and reported. Based on this information, the NB suffering from interference recognises which amount of inteiference comes from which cell. A focused action against the interferer may be undertaken, rather than all neighbours blindly taking responsibility for the interference.

In an implementation of the invention, eveiy cell gets at least one (UL) scrambling code allocated for UL SCCP transmission, the SCCP user scrambles their UL transmission with the cell specific code before sending it to the NB; every UL SCCP transmission has an embedded pilot sequence. which is known and recognised by each and every neighbouring NB after descrambling procedure; the pilot embedded into the UL transmission may be either cell specific or even a common sequence throughout the whole network; every NB makes measurements of the incoming interference; using the cell specific scrambling code the measured interference is divided into groups, at least one for every neighbouring cell (if one cell specific scrambling code is allocated to the cell; if more cell specific scrambling codes are allocated, e.g. according to the used data rates, more groups per cell may be created); the information about the neighbouring cell and interference created by it is sent to the RNC for a more targeted RRM.

The invention may also be applied to LTE. as this has no mechanism for recognising cell-specific interference. For LTE, the pilot is already embedded in the UL transmissioa but cell-specific scrambling needs to be introduced. The measurement results can be reported between eNBs, and in some cases between eNBs and remote radio resource managing units.

The invention allows for more efficient iriteiference management procedures by scrambling the UL SCCP transmission for all users in the given cell with the cell specific scramble code. It is assumed, that the special pilot sequence is embedded in every SCCP transmission, so that after descrambling the sequence is recognized.

Incoming interference, identified by the given scrambling code is measured and reported to RNC for better inter-cell interference management.

Claims (34)

1. A communication system comprising a base station and a plurality of terminals, wherein the system is capable of operating with a non-orthogonal uplink waveform; or a hybrid uplink waveform. comprising a non-orthogonal waveform with a more orthogonal wavefo rm; the base station further comprising a scheduler, wherein the scheduler set separate rise over thermal target levels for the non-orthogonal and hybrid waveforms.

2. A system according to claim 1, wherein the scheduler operates at the non-orthogonal RoT level when there are active non-orthogonal terminals in the system, or serviced by a neighbouring base station.

3. A system according to claim 1 or claim 2, wherein the more orthogonal waveform includes synchronisation and a common scrambling code.

4. A system according to any preceding claim, wherein the more orthogonal waveform is an orthogonal data waveform.

5. A system according to any preceding claim, wherein the non orthogonal waveform is a control waveform.

6. A system according to at least claim 4, wherein the more orthogonal data waveform is one of SC-FDMA, or synchronous E-DCH uplink and the non-orthogonal waveform is a WCDMA waveform.

7. A system comprising a base station and a plurality of terminals at least one of which is adapted to operate a non-orthogonal uplink and another is adapted to operate a hybrid non orthogonal/orthogonal uplink; wherein a scheduler in the base station operates with first and second Rise over Thermal (RoT) targets; the first RoT target for use when there are no non-orthogonal terminals active and the second RoT target for use when there are active non-orthogonal terminals.

8. A system according to claim 7. wherein the second RoT target is higher than the first.

9. A system according to claim 7 or claim 8, whereby when operating the hybrid uplink, the scheduler sets power control signal to interference ratio (SIR) targets and scheduling grants such that dedicated physical control channel (DPCCH) SIR targets are met whilst the second RoT level is in use.

A system according to any of claims 7 to 9. wherein the base station checks whether any neighbour base station is servicing non orthogonal terminals before switching to operate at the second RoT level.

11. A system according to any of claims 7 to 10. wherein the base station inforim its neighbour base station if a non orthogonal terminal is to be activated or scheduled, in order that the neighbours can lower their RoT targets.

12. A system according to any of claims 7 to 11. wherein the change in RoT is achieved by means of a control message common to all active terminals that raises the transmit power of at least the orthogonal part of a hybrid uplink waveform.

13. A system according to any of claims 7 to I 2. wherein the non-orthogonal terminal is a WCDMA terminal.

14. A system according to any of claims 7 to 13, wherein the hybrid terminal is a WCDMA/SC-FDMA terminal; a WCDMA/orthogonal CDMA terminal; or an WCDMA/synchronous E-DCH terminal.

15. A method of allocating traffic in a terminal adapted for transmitting orthogonal and non-orthogonal waveforms in a single transmit time interval; the method comprising determining a priority class for the traffic: and mapping high priority traffic to the ron-orthogonal waveform for autonomous transmissions and low priolity traffic to the orthogonal waveform.

16. A method according to claim 15. wherein subsequent transmissions, or retransmissions of the high priority traffic are scheduled on the orthogonal waveform.

17. A system comprising two transmission waveforms; a scheduled orthogonal waveform and a non orthogonal waveform whereby at least the first transmission of high priority traffic is transmitted in an unscheduled manner on the non orthogonal traffic whilst background data is transmitted on the orthogonal waveform

18. A system according to any ofclaims 15 to 17, whereby the orthogonal and non orthogonal waveform data are ACKJNACKed independently

19. A system according to any of claims 15 to 17, whereby a single ACK/NACK is sent based on reception of both the orthogonal and non orthogonal waveforms.

20. A system according to any of claims 15 to 19, whereby priority is given to the non orthogonal waveform over the orthogonal waveform in TFC selection

21. A system according to any of claims 15 to 20, wherein the high priority data is mapped to the orthogonal waveform if the waveform is scheduled for a first transmission with sufficient resources to cover the high priority PDU and otherwise is mapped to the non-orthogonal waveform.

22. A system according to any of claims 15 to 21, wherein the orthogonal waveform is SC-FDMA and the non-orthogonal waveform is WCDMA.

23. A system according to any of claims 15 to 21, wherein the orthogonal waveform is orthogonal CDMA and the non-orthogonal waveform is non-orthogonal CDMA.

24. A system comprising a plurality of base stations and at least one user terminal: wherein one base station assigns resources to the terminal and the other base stations receive a signal transmitted from the terminal; wherein the terminal transmits a composite of a wideband control signal and a data signal: wherein the terminal is allocated resources by the serving base station: wherein the wideband control signal.

which has UE specific scrambling. indicates the resources allocated for the narrowband data signal. which has serving cell specific scrambling.

25. A system according to claim 24, whereby the wideband control signal is 3GPP WCDMA and wherein the narrowband data signal one of SC-FDMA, with resources that are scheduled being time/frequency blocks; or CDMA. with resources that are scheduled being OVSF codes; or synchronous wideband CDMA, with resources that are scheduled being OVSF codes.

26. A system according to claim 24 or claim 25, wherein the control information that indicates the resources for the data channel is located on the E-DPCCH channel of the WCDMA signal.

27. A system according to any of claims 24 to 26, whereby the control information indicates that resources for the data channel are transmitted at the start of a transmission time interval

28. A method of distributing data channel transmission time and frequency information for a terminal in soft handover, the method comprising receiving at the terminal a scheduling allocation from a serving base station; transmitting to all base stations within range control information on a non-orthogonal waveform and data on an orthogonal waveform; wherein the data channel information is transmitted on the non-orthogonal waveform using terminal specific scrambling.

29. A method of inter-cell interference management in a communication system comprising at least one base station of a cell, at least one user device and a radio network controller, the method comprising scrambling an uplink single carrier cyclic prefix (SCCP) transmission for all user devices in a cell with a cell specific scrambling code: providing scrambling code information to neighbour cells: wherein each base station scans for transmissions from users in it's cell and for neighbouring cell specific SCCP scrambling codes: measures interference due to the SCCP transmission with a given scrambling code and reports the interference and its source to the radio network controller.

30. A method according to claim 29. wherein the communication system is IJMTS LTE.

31. A method according to claim 29 or claim 30, wherein a plurality of cell specific scrambling codes are allocated to each cell dependent upon additional parameters.

32. A communication system comprising a plurality of cells, each cell comprising a base station and a plurality of user devices, the system further comprising a radio network controller: wherein each cell is allocated at least one cell specific uplink scrambling code by the radio network controller; wherein each user device scrambles uplink transmissions with the cell specific scrambling code for transmission to the base station of the user device; wherein each base station measures incoming interference, allocates the inteiference to a group according to the scrambling code and informs the radio network controller of the cell and associated interference.

33. A system according to claim 32. wherein a cell specific or common pilot sequence is embedded in every uplink transmission.

34. A system according to claim 32 or claim 33, wherein the base station is an eNode B, or a Node B.