GB2469465A

GB2469465A - Method and apparatus for reducing co-channel interference

Info

Publication number: GB2469465A
Application number: GB0906396A
Authority: GB
Inventors: Justin Coon; Graeme Woodward
Original assignee: Toshiba Research Europe Ltd
Current assignee: Toshiba Europe Ltd
Priority date: 2009-04-14
Filing date: 2009-04-14
Publication date: 2010-10-20
Anticipated expiration: 2029-04-14
Also published as: GB2469465B; GB0906396D0

Abstract

A method and apparatus is described for reducing co-channel interference in a wireless network or cell 54. The method allocates transmission resources (A, B, C, D) to one or more communication terminals 50 in a network or cell 54, including repeatedly sensing the communication medium, determining which resources (A, B, C, D) are to be allocated to said one or more communication terminals 50 and dynamically allocating those resources (A, B, C, D). The repeated sensing of the communication medium is carried out both spatially and spectrally. Resources are allocated such that interference to terminals in adjacent networks or cells (co-channel interference) is reduced or meets some specified criteria and the quality of the link within the desired cell is improved or meets some specified criteria.

Description

I-

Method and Apparatus for Reducing Co-channel Interference

Field of the Invention

The present invention relates generally co-channel interference in communication systems and more particularly, but not exclusively, to a method and apparatus for reducing co-channel interference in a wireless network or cell.

Background of the Invention

Many wireless networks are constructed based on a cellular' topology, whereby each user terminal or device is assigned to communicate with a subset (typically one) of one or more base stations. The geographical area that is covered by a given base station's transmission is known as a cell'.

Figure 1 is a simplified illustration of a cellular network 1 consisting of seven cells 2, each cell 2 being juxtaposed at parts of its outer circumference with at least three adjacent cells 2 to define common boundaries 4 therewith. The boundaries 4 shown in Figure 1 are idealistic as showing the maximum geographical area for the network I whereas in practice it is desirable for the adjacent cells to overlap one another as will be seen with reference to Figure 2. Each cell 2 has a respective base station 6, which is located in its central region and one or more different user terminals or devices 8, randomly distributed within the cell 2, and assigned to the base station 6.

Some examples of such cellular systems include second-generation GSM networks and third generation UMTS networks, all of which support wireless transmission of voice and data information.

In order to ensure that a wireless cellular network 1 can service a large geographical area, it is often necessary in practice to locate base stations 6 such that their respective cells 2 overlap to some degree. Such an overlapping cell arrangement is shown in Figure 2 and is required to facilitate uninterrupted mobile communication across the network 1. The overlapping cell areas are shown as regions 10. However, if the devices 8 in adjacent cells utilise the same spectrum for communication, this geographical overlapping can lead to co-channel interference (CCI) at the cell edges. This problem is exacerbated when new base stations (such as so-called femtocell' base stations), are added to the network such that their coverage area overlaps with existing cells.

An example of CCI is illustrated in Figure 3, in which the CCI occurs at the edge of two overlapping cells 12, 14 hereinafter referred to as cells A and B respectively. Cell A is provided in its central region with a base station 16, and has four users assigned to cell A, the user terminals or devices 20, 22, 24, 26 being within the geographical area of cell A. Cell B has a base station 18, set slightly offset from the cell's centre, and has three users assigned to Cell B, their terminals or devices being designated 28, 30 and 32. It will be seen that as the geographical areas of cells A and B overlap at an overlap region 34 there are two user devices, device 26 assigned to Cell A and device 32 assigned to Cell B, which both fall within this overlap region 34 and hence fall within the geographical coverage area of both Cells A and B. Consequently, whereas a desired signal can be transmitted from the base station 16 of Cell A to the desired user device 20, an interference signal is also propagated to the undesired user 32 in the overlapping region 34.

Many techniques have been proposed to mitigate CCI. A common approach is to allocate non-overlapping spectrum to adjacent cells, thereby allowing devices in these cells to communicate freely without interfering with one another. Another approach is to utilise sectorised antennas at the base stations, which allows the base station to direct its transmission to one part of the cell and, ideally, to avoid radiating into an adjacent (victim) cell (or sector within the same cell) on the same channel (i.e., using the same spectrumlfrequency). An example of the latter technique is illustrated in Figure 4.

Figure 4 illustrates the application of a sectorised transmission to avoid or minimise CCI, the reference numerals associated with the cells, base stations and user terminals or devices being identical to those of Figure 3. In Figure 4 it can be seen that the transmission from base station 16 of Cell A has been sectorised so that it is directed towards cell 20 to the exclusion of any other user terminal or device. Accordingly no interference signal is transmitted into the overlapping cell region 34 to be picked up as interference by device 32 of Cell B. Another related but different technology is opportunistic (or random) beamforming.

Systems that employ opportunistic beamforming transmit a single signal via multiple antennas, applying a time-varying fading function to the transmitted signal at each antenna. Such an opportunistic beamforming system is illustrated in Figure 5 in which a time varying signal TX, transmitted by a base station is fed to two antennas 36 and 38.

The signal TX to antenna 36 has a random fading function f1(t) applied to it at a modulator 33 before it is transmitted from antenna 36 along a channel TX1 to the antenna 40 of a User k, the channel matrix to User k from antenna 36 being designated hlk(t). In similar fashion the signal TX to antenna 38 has a random fading function f2(t) applied to it at a modulator 35 before it is transmitted from antenna 38 along a channel TX2 to the antenna 40 of the User k, the channel matrix to User k from antenna 38 being designated h2k(t).

This fading function enhances the natural fading in the wireless channel, thus creating peaks and troughs in the quality of a given communication link over time. Users are then scheduled in time such that signals are transmitted only at the peaks of the users' respective link qualities.

Figure 6 is a graph of signal's link quality against time for three signals, each transmitted to a respective user. A different fading function is applied to each signal. It will be observed that each of Users 1, 2 and 3 is scheduled to receive their signal at the peaks of the signals transmitted to them.

In a multi-cell environment, opportunistic beamforming implicitly reduces CCI when the metric that defines the link quality, which is used for user scheduling, is a function of the interference from adjacent cells or devices. Specifically, this metric is typically some function of the signal-to-interference-plus-noise ratio (SINR). Therefore, when a user's SINR is high, the CCI acting upon that user is low relative to the desired signal power. It should be noted that this opportunistic nulling is a by-product of opportunistic beamforming and scheduling; in other words, the goal of opportunistic beamforming is not specifically to reduce the interference in neighbouring (victim) cells, but to enhance the link quality for an intended user.

A more thorough treatment of opportunistic beamforming, scheduling and nulling can be found in a book entitled Fundamentals of Wireless Communication, by Tse and Viswanath, Ch. 6, Cambridge University Press, 2005, the whole contents of which are incorporated herein by way of reference.

The whole of the contents of a paper by Sharma et a! N. Sharma and L. Ozarow, entitled "A study of opportunism for multiple-antenna systems", IEEE Trans Information Theory, 2005, Vol 51, no 5, ppl8O4-l8l4, is also incorporated herein by way of reference. This work by Sharma et al describes a combination of opportunistic user scheduling and antenna beamforming, of which transmit antenna selection is a special case. They do not consider the issue of interference on neighbouring networks or cells.

Zhou and Thompson devised a method that can be used to mitigate the problem of CCI in cognitive networks. The paper by Zhou and Thompson, is entitled Single-antenna selection for MISO cognitive radio,' in proc. of the 2008 lET Seminar on Cognitive Radio and Software Defined Radios: Technologies and Techniques, pp. 1-5, Sep. 2008.

The whole contents of the paper by Zhou and Thompson, is incorporated herein by way of reference. In their paper, they describe an algorithm whereby a single transmit antenna is selected by an interfering device to convey information to an intended receiver such that the interference caused to a neighbouring victim device is minimised.

This approach utilises the fading nature of the (assumed to be) independent wireless channels to obtain a gain in performance (i.e., a reduction in the interference level).

Zhou and Thomson also disclose an alternative method whereby the transmit antenna is chosen such that the ratio of the desired channel gain to the interference channel gain is maximised.

With respect to Zhou and Thompson's work, referred to above, it focuses on single-antenna selection in frequency-flat channels. They also only consider antenna selection rather than the general resource allocation problem.

WO 2004/073336 Al (Nortel Networks Ltd), the whole contents of which is incorporated herewith, describes the self-selection of radio frequency channels to reduce co-channel and adjacent channel interference in a wireless distributed network.

WO 2005/032169 A2, (Bandspeed mc), the whole contents of which is incorporated herewith, describes a method and apparatus for sector channelization and polarization for reduced interference in wireless networks.

With respect to WO 2004/073336 Al and WO 2005/032 169 A2, the use of sectorised antennas to mitigate CCI is not possible in some circumstances. In particular, if one considers the example where the base station selects the sectorised antenna for transmission that minimises interference to a particular user in an adjacent cell, but the user to which the base station transmits within its own cell is located in a direction that is similar to the aforementioned user in the adjacent cell. Tn this case, the signal quality to the desired user can be poor.

EP 1 662 834 Al (Siemens Mobile Communications), the whole contents of which is incorporated herewith, describes a dynamic channel selection based on interference and spectral usage in a multicarrier cellular communication system. With respect to EP 1 662 834 Al, existing schemes typically aim to reduce the interference to adjacent cells with little regard for the quality of the intended link. In some cases, resource allocation is carried out such that CCI is minimised and resource occupation is maximised (i.e., the number of users supported is maximised).

The present invention strives to overcome the disadvantages mentioned above in

relation to the prior art.

Summary of Invention

In one aspect of the invention there is provided a method of allocating transmission resources to one or more communication terminals or devices in a network or cell, including repeatedly sensing the communication medium, both spatially and spectrally, to determine which resources are to be allocated to said one or more communication terminals or devices and dynamically allocating those resources whereby the quality of the link within a desired cell or the interference to terminals or devices in adjacent networks/cells meets some specified criteria, In one embodiment the allocation of resources is chosen whereby the interference to terminals or devices in adjacent networks or cells is reduced or meets a specified criteria and the quality of a link within a desired cell is improved or meets a specified criteria.

In a further embodiment of the invention there is provided a method of allocating transmission resources in a network or cell dynamically whereby the interference to devices in adjacent networks/cells is reduced, and the quality of the link within the desired cell meets some specified criteria. In said further embodiment the quality of the link within the desired cell may satisfy a predetermined minimum condition.

In a yet further embodiment of the invention there is provided a method of allocating transmission resources in a network or cell dynamically whereby the quality of the link within the desired cell is improved, and the interference to devices in adjacent networks/cells meets some specified criteria. In said yet further embodiment the interference to devices in adjacent networks/cells satisfies a predetermined maximum condition.

The resources to be allocated may include one or more of the following: a. one or more of a plurality of transmission frequency bands, b. scheduling users for transmission; c. one or more of a plurality of transmit antennas from which information is conveyed, and d. transmit power.

The resources may also include spreading codes and antenna polarisation.

In an embodiment the quality of a link is measured as a function of the channel gain of that link.

In a further embodiment the quality of a link is measured as a function of the maximum theoretical rate, or capacity, of that link.

In one embodiment the transmission frequency and antenna are chosen to maximise the objective function 9(6,yfL.y/t) = 6h(yf) -(1 -6)h(y)t) where 0 S �= 1 is a trade-off parameter, r is the gain of the desired channel on frequency band f corresponding to transmission from antenna, "/ is the gain of the interference channel on frequency band f corresponding to transmission from antenna, and h(x) is a function of the channel or channel gain.

The term may be a function of one or more of the channel gain, maximum theoretical rate or capacity of the link.

In a yet further embodiment the transmission frequency, antenna, and transmit power are chosen to satisfy the optimisation problem max h(P,1) st.

P1 �= P1 where is the transmit power on the fth frequency band, hy) is a channel matrix, Pf is a predetermined bound on the quality -as defined by the function -of the interference link, and 1'mc-zf is a predefined maximum transmit power limit.

The channel matrix h(xy) may be a function of one or more of the channel gain, maximum theoretical rate or capacity of the link.

In another embodiment the transmission frequency, antenna, and transmit power are chosen to satisfy the optimisation problem mn h(P1.yff) s.t I �=P1.42,1 where is the transmit power on the I th frequency band, h(x,y) is a function of one or more of the channel gain, maximum theoretical rate or capacity of the link, P; is a predetermined bound on the quality -as defined by the function y) -of the desired link, and "rr.azf is a predefined maximum transmit power limit.

The antenna responses may be deliberately chosen to have directional characteristics.

By way of example, the antenna responses may be specifically designed to overlap such that each azimuthal direction has coverage from at least two antennas and so that a subset of antennas can be selected to support transmission from multiple antennas simultaneously.

A transmitted signal for transmission to an intended user may, before transmission, be modulated with a fading function to allocate resources for transmission at a given time while simultaneously satisfying a predetermined criteria related to the intended user's link quality.

In a further aspect of the invention there is provided a communication apparatus configured to enable selective communications with one or more terminals or devices located in a spatial area around the apparatus, wherein said apparatus is configured to allocate transmission resources to one or more of said terminals or devices in said spatial area, including a sensing device to repeatedly sense the communication medium, both spatially and spectrally, and a processor for determining which resources are to be allocated to said one or more communication terminals or devices and dynamically allocating those resources, whereby the quality of the link within a desired cell or the interference to terminals on devices in adjacent networks/cells meet some specified criteria.

Further embodiment of the communication apparatus can be based on features in the dependent apparatus claims.

In a further aspect of the invention there is provided a computer readable medium storing executable instructions which, when executed on general purpose controlled communications apparatus, causes the apparatus to become configured to perform the method defined above.

In a yet further aspect of the invention there is provided a signal carrying computer receivable information, the information defining computer executable instructions which, when executed on a general purpose computer controlled communication apparatus, causes the apparatus to become configured to perform the method defined above.

Brief description of the drawing

Further preferred features of these aspects of the invention will now be set forth by way of the following description of specific embodiments of the invention, provided by way of example only, with reference to the accompanying drawings in which: Figure 1 is a schematic illustration of a known cellular network with different user devices assigned to different base stations, Figure 2 is a schematic illustration of a known non-ideal cellular topology with overlapping cells, Figure 3 illustrates an arrangement in which there is CCI at the cell edge, Figure 4 is an illustration of a sectorised transmission to reduce the CCI in Figure 3, Figure 5 is an illustration of opportunistic beamforming, Figure 6 is an illustration of user scheduling using opportunistic beamforming, Figure 7 is an illustration of an intended link and corresponding channel paths and interference paths, Figure 8 is an illustration of a design incorporating overlapping directional antenna patterns, Figure 9 is a graph of mean channel gains for intended and interference links when problem (1) is used to allocate one antenna of four for transmission and Figure 10 is a graph illustrating the mean of received signal power vs. upper bound on interference level for the intended link and the interference link.

Specific embodiments of the present invention will be described in further detail on the basis of the attached diagrams. It will be appreciated that this is by way of example only, and should not be viewed as presenting any limitation on the scope of protection sought.

The embodiments of the invention described below describe methods of allocating transmission resources in a network or cell dynamically (i.e., utilising temporal fluctuations in the wireless channel) such that the interference to devices in adjacent (victim) networks/cells is reduced (or at least meets some predetermined criteria) and the quality of the intended link is improved (or at least meets some predetermined criteria).

A resource is defined here as being a degree of freedom which can be utilised to convey information wirelessly to an intended receiver. For example, resources may include a. one or more of a plurality of transmission frequency bands, b. one or more of a plurality of transmit antennas from which information is conveyed, c. scheduling users for transmission, and d. transmit power.

Resources may also include spreading codes and antenna polarisation.

In general, the quality of the intended link can be measured as a function of the wireless channel gain that is derived from the transmission of a wireless signal or signals from one or more transmit antennas to the intended receiver. Alternatively, the quality of the intended link can be measured as a function of the maximum theoretical rate, or capacity, of the intended link.

In a similar manner, the amount of interference caused to a neighbouring device/cell can be measured as a function of the wireless channel gain that is derived from the transmission of a wireless signal or signals from one or more of a plurality of transmit antennas to the neighbouring device/cell. Alternatively, the amount of interference caused to a neighbouring device/cell can be measured as a function of the maximum theoretical rate, or capacity, of the interference link.

Specific embodiments of the invention will now be described with respect to the utilisation of one or more of the aforementioned resources.

Antenna Allocation Only Consider the case where the transmitter is equipped with a plurality of transmit antennas (e.g. two antennas). At an allocated time, the transmitter is scheduled to convey information wirelessly to an intended receiver, which may have a plurality of receive antennas. Simultaneously, this signal will interfere with a device in a neighbouring (victim) cell or network. An example of this scenario is depicted in Figure 7. In this example, it can be seen that there are two wireless channels 42, 44 in the intended link (corresponding to transmission from each of the transmit antennas to the intended receive antenna 46), and there are two wireless channels in the interference link to a not intended antenna 48.

As a consequence of multipath propagation, the gain of each wireless channel on a particular frequency at any given time is a random variable. With probability one, the qualities of these channels (defined and measured as stated above) will be different and can be ordered from least to greatest. Furthermore, in many cases, the group of channels corresponding to the intended link and the group of channels corresponding to the interference link are statistically independent to a close approximation. Thus, one can choose a transmission strategy whereby one or more transmit antennas are selected for transmission such that the intended link quality is improved and the interference to neighbouring cells/devices is reduced. In other words, spatial fading is exploited to reduce CCI.

In one embodiment of this invention, the transmit antenna is chosen such that the following objective function is maximised: g(&rfvRF) = -(1 -(1) where 0 �= 5 �= 1 is a trade-off parameter, YJ is the gain of the desired or intended channel on frequency band f corresponding to transmission from antenna, and Y)( is the gain of the interference channel on frequency band I corresponding to transmission from antenna. The parameter is a design parameter that allows for a trade-off in the improvements in the intended link and the reduction in interference to neighbouring cells/devices. When 1, the solitary objective is to maximise the gain of the desired link. When 5 0, the solitary objective is to minimise the gain of the interference link.

Note that these extreme cases are to some extent explained in the paper by Zhou and Thompson referred to above. Therefore, our approach is a generalisation, one of the features of novelty of which in this embodiment lies in the inclusion of the parameter and the corresponding problem formulation. By using this problem formulation, the quality of the intended link and the interference level can be adjusted easily to meet quality of service (QoS) requirements over time. It should be emphasised that, in general, the maximisation of equation (1) above will lead to different antenna selection strategies for different values of. This fact is another distinguishing feature of our approach relative to the one detailed in the paper by Zhou and Thompson.

In another embodiment of this invention, the transmit antenna is chosen such that the following objective function is maximised: = sIog2(1 +P1y,) -(1 -5)1og(1 +iyff) where is the power transmitted on frequency band I. The Iog2(.) function corresponds to the Shannon capacity of the channel defined by the argument. Although the second term in this equation does not specifically refer to the interference level, its reduction implies a reduction in the interference level.

A simple method of solving problems of the form given above entails calculating the objective function for each antenna (i.e., for every 1), then choosing the antenna index that maximises the function.

Note that in order to solve problems of this nature, the transmitter must have knowledge of the various channel gains for the intended link and the interference link, which can be gleaned from feedback in frequency-division duplex (FDD) systems or from channel reciprocity in time-division duplex (TDD) systems. For example, the 3GPP Long Term Evolution (LTE) specification facilitates feedback in FDD systems with channel quality indicator (CQI) feedback, which can be taken to be a measure of the channel gain in some cases.

With reference to the above problem formulations, other functions of the channel gains can be constructed, such as functions that would minimise the mean-square error (MSE) in the received signal in the intended link. A person skilled in the art would no doubt be able to contrive any number of formulations with a similar structure to the two problems detailed above.

Frequency Allocation Only In heavily loaded systems, it may not be possible to allocate frequency bands for communication such that interference to victim devices is completely avoided. In order to maintain an acceptable Q0S, one may be required to transmit even though this will cause interference to other users. In one embodiment of this invention, the transmitter chooses to transmit from a given transmit antenna (say, ) on the frequency band that maximises the following objective function: d. d (i r\. mt 9v."YjVYfm -uYj -UJff where the variables are defined above. This approach is similar to the Antenna Allocation Only case outlined above, but the selection is done in the frequency domain.

Thus, spectral fading is exploited to reduce CCI in this case (this is also termed frequency selective fading).

Frequency and Antenna Allocation It can easily be seen that the two embodiments of resource allocation described above can be combined to yield an embodiment whereby the antenna arid the frequency band I are both chosen to maximise objective functions of the form 2,)nr) -(i -It should also be noted that some systems (such as those using orthogonal frequency division multiplexing (OFDM)) use multiple frequencies, called subcarriers, to convey information to the receiver. In such systems, the aforementioned frequency andlor antenna allocation schemes can and should be applied independently for each subcarrier or contiguous group of subcarriers.

Frequency and/or Antenna Allocation with Power Loading It is also beneficial to allow the transmitter to vary its power on a given frequency from a given antenna. This approach will not provide any advantages if it were used in conjunction with the objective functions defined in the preceding sections; however, reformulating the problem can lead to additional gains.

In one embodiment of the invention, the transmission frequency, antenna, and transmit power are chosen to satisfy the optimisation problem max mt S. fYf.i -P1 P1 <-Prncl,f where is the transmit power on the fth frequency, P1 is a predetermined bound on the quality of the interference link on the fth frequency, and rc.x.f is a predefined maximum transmit power limit on the fth frequency. The resource allocation that solves this problem maximises the gain of the intended link such that the gain of the interference link is less than or equal to P1.

A method of solving this problem is outlined below.

Si. Choose Cf. i) e F xl (the compieto s9t of freqncy bands x the complete set of antenna indfces} such that (1) has not yet been chosen.

S2. Let P, min(5,) S3. If all (. ) have been considered, choose the (f) that maximises Yii; otherwise store and the corresponding Cf. ) and go to step Si.

In another embodiment, the transmission frequency, antenna, and transmit power are chosen to satisf' the optimisation problem rnn st. Pjyf Pf _<P1 where the variables are defined above. The resource allocation that solves this problem minimises the gain of the interference link such that the gain of the intended link is at least as high as Pi Due to the structure of this problem, there may be cases where a solution is not possible.

Generally, this can result from scenarios where the intended channel gain is very poor, in which case the minimum intended link quality may not be met. However, this is a rare event in practice when there are a large number of degrees of freedom, and a simple method that can be used to solve this problem with high probability is given below.

SI. Choose Cf. iT) E F x! = {the complete set of frequency bands x the complete sot of antonia rtd ices) such that Cf. ) has not yet been chosen. (P1

S2.Let1 =tflIflTP,,.Qxf S3. If all (i) have been considered, choose the Cf. 0 that minimises J'JYJ subject to the constraints; otherwise store 1';v) and the corresponding ([ 0 if it meets the constraints and go to step Si.

It is important to note that other cost/utility functions and constraints can be formulated, which are similar to the formulations given above. A person skilled in the art would no doubt be able to devise any number of variations of the proposed formulations.

User Scheduling with Antenna/Frequency Allocation One can also imagine the scenario where in addition to having multiple transmit antennas and/or frequencies from/on which to transmit, a device must also schedule transmission to a number of users. Traditionally, user scheduling can be performed opportunistically as discussed above in relation to Figure 6. This approach can easily be extended to work with antenna/frequency selection methods as well. In this case, the user, antenna and/or frequency are schedule for transmission at a given time such that some utility function is maximised, such as the user's received SITNR. However, transmission to the best' user as defined above may cause a significant amount of interference to be imposed on the victim device. Thus, the approaches described above can easily be extended to consider the channel gains or capacities associated with different users as well as different antennas and frequencies. The resulting user/antenna/frequency selection will no doubt be different, in general, to the conventional opportunistic scheduling or antenna selection solutions. Note that the work by Sharma et al referred to earlier includes a combination of opportunistic user scheduling with transmit antenna selection, however interference on neighbouring networks or cells is not considered.

Antenna Directionality The methods described above with omnidirectional antennas rely entirely upon the independence of fading from multiple transmit antennas to achieve a difference in received power at the desired and co-channel (victim) receivers. Greater average separation may be achieved by imposing different antenna directionality on some or all transmit antennas. This may be particularly effective at transmitters where there are influences that would otherwise cause a high degree of correlation between signals transmitted from different antennas, for example shadowing effects.

Overlapping directional antennas.

In modem communication systems multi-antenna transmissions are increasingly important to meet high data rate demands. Such methods are often termed MIMO (Multiple-Input-Multiple-Output) communications. To support MIMO a subset of transmit antennas should be chosen from the total of all the available transmit antennas using the methods described above. As a special case of the previous method employing directional antennas, the antenna coverage patterns of the transmit antennas can be designed specifically to achieve overlap between antennas. This can be done in such a way to increase the likelihood of strong coverage for the intending user from at least two antennas, while achieving the advantage of improved separation between the intending and victim receivers in a probabilistic sense. An example of such an antenna coverage pattern to support MIMO through subset selection is show in Figure 8. In this example, the intended receiver 50 will usually receive good coverage from Antennas A, B and C, and the victim receiver 52 would only normally receive good coverage from Antennas A and 11 A dynamic antenna allocation would tend to allocate transmission from antennas B & C more often in order to achieve good coverage to the intended user with little interference to the victim network. Note, however, that the transmit antenna selection is still performed in a dynamic sense, i.e. opportunistic over time. The propagation characteristics between each transmit and receive antenna will vary over time and thus the subset of antennas selected for transmission will vary over time. This is an important distinction between the method proposed and previous methods using sectorised antennas to reduce interference. The method proposed can achieve similar average gains expected from a sectorised approach in addition to the instantaneous gains achieved through the dynamic allocation methods described elsewhere in this specification Note also that the antenna patterns may be designed with more (or less) overlap.

An example of the gain that can be achieved by employing the Antenna Allocation Only approach when the wireless channels are Rayleigh fading can be observed in Figure 9.

In this example, one of four transmit antennas was allocated for transmission, and the problem defined in the equation (1) was used to determine which antenna should be selected. Clearly, the parameter defines a trade-off between the quality of the intended link and the level of interference. At the extreme point = , emphasis is placed only on maximising the quality of the desired link. In this case, due to the independence of the intended link and interference link channels, the average gain of the interference link is not reduced (on the graph, 0 dB is the reference value). However, the average gain of the intended link increases by 101og10(C + p0w + i)) dB where C 0.5772 is Euler's constant, PC) is the digamma function and M is the number of transmit antennas from which the selection is made. In this example, M and the increase in the average gain is log (c + ip(5)) 3.19 dB.

At the other extreme point 0, emphasis is placed only on minimising the interference level. In this case, the gain of the desired link remains 0 dB, but the average gain of the interference link is reduced by 101og10(M) dB which in this example is 101og(4) 6dB.

An example of the gain that can be achieved by employing Power Loading with Antenna Selection in Rayleigh fading channels can be observed in Figure 10. In this example, equation (2) was used to determine the power and antenna allocations. Figure depicts the mean signal strength at the receiver of the intended device as well as at the unintended receiver as a function of the upper bound on the interference level defined by P in (2). As Pj is relaxed (i.e., grows large), it can be seen in this example that the solution is identical to the solution to equation (1) when 6 1 Alternatively, when Pf is small, it can be seen that the Power and Antenna Allocation approach is able to facilitate a lower average interference level than the Antenna Allocation Only approach. This result is significantly different than that reported in the paper by Zhou and Thompson, where the lowest average interference gain is stated to be reduced by iOioj(M) dB as discussed above.

In relation to the two examples given above, conventional systems rely on sectorised antennas to reduce the interference level to an unintended user without regard for the intended user's link quality. Alternatively, conventional systems may use opportunistic beamforming to schedule transmission to users in a specific order such that some objective is achieved. In contrast, the embodiments of the invention disclosed in this specification relies on the fading nature of the channel to allocate resources for transmission at a given time while simultaneously satisfying some predetermined criteria related to the intended user's link quality as well as the interference level as seen by one or more unintended receivers.

The allocation process does not require the establishment of a link between the interfering cell and the victim cell in TDD systems since channel reciprocity combined with a channel sensing operation can be used to determine the unintended user's link quality. Moreover, in many FDD systems, such as the systems specified by the 3GPP LTE working group, means are in place for conveying information about the quality of any given link to other nodes/devices in the network. This information can be directly used in the present invention.

It will be appreciated by the person skilled in the art that the problem associated with WO 2004/073336 Al and WO 2005/032169 A2 may be overcome by the present invention by utilising the spatial and temporal fading nature of the channel rather than spatial sectorisation or beamforming to reduce interference, and the allocation is performed dynamically over time.

With respect to EP 1 662 834 the problem may be overcome by the present invention seeking to simultaneously reduce the CCI and simultaneously improve the intended link quality (or at least meet some predetermined criteria).

Claims

Claims 1 A method of allocating transmission resources to one or more communication terminals or devices in a network or cell, including repeatedly sensing the communication medium, both spatially and spectrally, to determine which resources are to be allocated to said one or more communication terminals or devices and dynamically allocating those resources whereby the quality of the link within a desired cell or the interference to terminals or devices in adjacent networks/cells meets some specified criteria.
2 A method as claimed in claim 1 wherein the allocating of transmission resources in a network or cell is performed dynamically whereby the interference to terminals or devices in adjacent networks or cells is reduced, and the quality of the link within the desired cell meets some specified criteria.
3 A method as claimed in claim 2 wherein the quality of the link within the desired cell satisfies a predetermined minimum condition.
4 A method as claimed in claim 1 wherein the allocating of transmission resources in a network or cell is performed dynamically whereby the quality of the link within the desired cell is improved, and the interference to terminals or devices in adjacent networks/cells meets some specified criteria.
A method as claimed in claim 4 wherein the interference to terminals or devices in adjacent networks/cells satisfies a predetermined maximum condition.
6 A method as claimed in any one of claims 1 to 5, wherein the transmission resources to be allocated include one or more of the following: (i) one or more of a plurality of transmission frequency bands, (ii) scheduling users for transmission; (iii) one or more of a plurality of transmit antennas from which information is conveyed, and (iv) transmit power.
7 A method as claimed in any one of claims 1 to 6 wherein the transmission resources include spreading codes andlor antenna polarisation.
8 A method as claimed in any one of claims 1 to 7 wherein the quality of a link within a cell is measured as a function of the channel gain of that link.
9 A method as claimed in claim 8 wherein the quality of a link within a cell is measured as a function of the maximum theoretical rate, or capacity, of that link.
A method as claimed in any one of claims 1 to 9, wherein the transmission frequency and antenna are chosen to maximise the objective function g(. y. -(1 -where 5 �= 1 is a trade-off parameter, vf is the gain of the desired channel on frequency band I corresponding to transmission from antenna, is the gain of the interference channel on frequency band I corresponding to transmission from antenna, and is a function of the channel or channel gain.
11 A method as claimed in claim 10 wherein the term hC) is a function of one or more of the channel gain, maximum theoretical rate or capacity of the link.
12 A method as claimed in any one of claims 1 to 9 wherein the transmission frequency, antenna, and transmit power are chosen to satisfy the optimisation problem max h(P1) s.t. h(P1yJ)�=p1 P �=Prnc where is the transmit power on the I th frequency band, y) is a channel matrix, Pi is a predetermined bound on the quality, as defined by the function h(x Y), of the interference link, and "nw.x.f is a predefined maximum transmit power limit.
13 A method as claimed in claim 12 wherein the channel matrix h y) is a function of one or more of the channel gain, maximum theoretical rate or capacity of the link.
14 A method as claimed in any one of claims 1 to 9 wherein the transmission frequency, antenna, and transmit power are chosen to satisfy the optimisation problem mm h(Pjyff) s.t.P, �=P where is the transmit power on the fth frequency band, h(xy) is a function of one or more of the channel gain, maximum theoretical rate or capacity of the link, P1 is a predetermined bound on the quality -as defined by the function -of the desired link, and rr-cxJ is a predefined maximum transmit power limit.
A method as claimed in any one of claims 10 to 14 wherein the antenna responses are chosen to have directional characteristics.
16 A method as claimed in claim 15 wherein the antenna responses are configured to overlap such that each azimuthal direction has coverage from at least two antennas and so that a subset of antennas can be selected to support transmission from multiple antennas simultaneously.
17 A method as claimed in any one of claims 1 to 16 wherein a transmitted signal for transmission to an intended user is, before transmission, modulated with a fading function to allocate resources for transmission at a given time while simultaneously satisfying a predetermined criteria related to the intended user's link quality.
18 A communication apparatus configured to enable selective communications with one or more terminals or devices located in a spatial area around the apparatus, wherein said apparatus is configured to allocate transmission resources to one or more of said terminals or devices in said spatial area, including a sensing device to repeatedly sense the communication medium, both spatially and spectrally, and a processor for determining which resources are to be allocated to said one or more communication terminals or devices and dynamically allocating those resources, whereby the quality of the link within a desired cell or the interference to terminals on devices in adjacent networks/cells meet some specified criteria.
19 A communication apparatus as claimed in claim 18 wherein the apparatus is configured to allocate the transmission resources in a network or cell, the allocation being performed dynamically whereby the interference to terminals or devices in adjacent networks or cells is reduced, and the quality of the link within the desired cell meets some specified criteria.
A communication apparatus as claimed in claim 19 wherein the apparatus is configured to ensure the quality of the link within the desired cell satisfies a predetermined minimum condition.
21 A communication apparatus as claimed in claim 18 wherein the apparatus is configured to allocate the transmission resources in a network or cell, the allocation being performed dynamically whereby the quality of the link within the desired cell is improved, and the interference to terminals or devices in adjacent networks/cells meets some specified criteria.
22 A communication apparatus as claimed in claim 21 wherein the apparatus is configured to ensure the interference to terminals or devices in adjacent networks/cells satisfies a predetermined maximum condition.
23 A communication apparatus as claimed in any one of claims 18 to 22, wherein the transmission resources to be allocated include one or more of the following: (i) one or more of a plurality of transmission frequency bands, (ii) scheduling users for transmission; (v) one or more of a plurality of transmit antennas from which information is conveyed, and (vi) transmit power.
24 A communication apparatus as claimed in any one of claims 18 to 23 wherein the transmission resources include spreading codes andlor antenna polarisation.
A communication apparatus as claimed in any one of claims 18 to 24 wherein the quality of a link within a cell is measured as a function of the channel gain of that link.
26 A communication apparatus as claimed in claim 25 wherein the quality of a link within a cell is measured as a function of the maximum theoretical rate, or capacity, of that link.
27 A communication apparatus as claimed in any one of claims 18 to 26, wherein the transmission frequency and antenna are chosen to maximise the objective function (& t) _ Th(y/) (1 -s)h(yjr) where 0 �= �= 1 is a trade-off parameter, Yf is the gain of the desired channel on frequency band I corresponding to transmission from antenna, i is the gain of the interference channel on frequency band f corresponding to transmission from antenna, and h(x) is a function of a channel or channel gain.
28 A communication apparatus as claimed in claim 27 wherein the channel matrix h(x) is a function of one or more of the channel gain, maximum theoretical rate or capacity of the link.
29 A communication apparatus as claimed in any one of claims 18 to 26 wherein the transmission frequency, antenna, and transmit power are chosen to satisfy the optimisation problem flI4X h(Pjy) sjt.Pf �= Pi where is the transmit power on the fth frequency band, h(xy) is a channel matrix, Pt is a predetermined bound on the quality, as defined by the function 1(Ly), of the interference link, arid ro.zf is a predefined maximum transmit power limit.
A communications apparatus as claimed in claim 29 wherein the channel matrix h(xy) is a function of one or more of the channel gain, maximum theoretical rate or capacity of the link.
31 A communication apparatus as claimed in any one of claims 18 to 30 wherein the transmission frequency, antenna, and transmit power are chosen to satisfy the optimisation problem mm h(P1.yff) s. t. h(P1.y)) ? P; "f <P,41 where is the transmit power on the fth frequency band, h(x. y) is a function of one or more of the channel gain, maximum theoretical rate or capacity of the link, Pf is a predetermined bound on the quality -as defined by the function h(x Y) -of the desired link, and axf is a predefined maximum transmit power limit.
32 A communications apparatus as claimed in any one of claims 18 to 31 wherein the antenna responses are chosen to have directional characteristics.
33 A communications apparatus as claimed in claim 32 wherein the antenna responses are configured to overlap such that each azimuthal direction has coverage from at least two antennas and so that a subset of antennas can be selected to support transmission from multiple antennas simultaneously.
34 A communications apparatus as claimed in any one of claims 18 to 33 wherein a transmitter is configured to modulate, with a fading function, a signal before transmission to an intended user, thereby allocating resources for transmission at a given time while simultaneously satisfying a predetermined criteria related to that user's link quality.
A computer readable medium storing executable instructions which, when executed on general purpose controlled communications apparatus, causes the apparatus to become configured to perform the method claimed in any one of claims 1 to 17.
36 A signal carrying computer receivable information, the information defining computer executable instructions which, when executed on a general purpose computer controlled communication apparatus, causes the apparatus to become configured to perform the method of any of claims ito 17.