TECHNICAL FIELD

The invention relates to the field of handling interference between neighboring radio access networks.

BACKGROUND

In wireless communication networks, for example, in mobile communication networks, mobile terminals and base stations transmit signals using different power levels. The used power levels depend on, for example, locations of the mobile terminal in their respective cells. Sometimes this causes interference between mobile terminals, in other words, signals associated with a mobile terminal interfere signals associated with another mobile terminal. Specifically, at an edge of a radio access network mobile terminals may cause interface to mobile terminals of a neighboring radio access network. There is a constant need for solutions that would mitigate interference between neighboring radio access networks.

SUMMARY

It is the objective of the invention to provide a solution that coordinates interference between neighboring radio access networks more efficiently.

This objective is achieved by the features of the independent claims. Further embodiments and examples of the invention are apparent from the dependent claims, the description and the figures.

According to a first aspect, there is provided a radio access network controller for a first radio access network. The radio access network controller comprises a transceiver configured to receive, from a second radio access network, a utilization parameters regarding resource allocation decisions made by a radio access network controller of the second radio access network; and a processor configured to generate an improved resource allocation data structure for the radio access network controller by taking into account the received utilization parameters. The disclosed solution improves performance by compensating for interference at the edge of the first radio access network, for example, by finding a radio resource allocation that can reduce the negative effect of interference. In an implementation form of the first aspect, the transceiver is configured to receive a radio head announcement message from the radio access network controller of the second radio access network for creating a radio head context for receiving the utilization parameters regarding resource allocation decisions made by the radio access network controller of the second radio access network, the radio head announcement message comprising radio head specification information of at least one radio head of the second radio access network. The use of the radio head announcement message enables a more compressed transmission of the utilization parameters and their values between the radio access networks, since the radio head specification information will less often be required to be changed.

In a further implementation form of the first aspect, the transceiver is configured to receive a radio resource allocation announcement message from the radio access network controller of the second radio access network, the radio resource allocation announcement message comprising the utilization parameters, and to use the radio head context to extract information from the radio resource allocation announcement message, wherein the utilization parameters comprises for each radio head at least one of: a usage frequency of a fixed beam, and a usage frequency of a transmit power level. The use of the radio allocation announcement message enables transmission of the utilization parameters between the radio access networks.

In a further implementation form of the first aspect, the transceiver is configured to send a request for the radio resource allocation announcement message to the radio access network controller of the second radio access network. By sending the request for the radio resource allocation announcement message to the second radio access network, the radio access network controller is able to determine the point of time when it wishes to receive the utilization parameters regarding resource allocation decisions made by the second radio access network in order to generate the improved resource allocation data structure. This also provides a means for the radio access network controller to synchronize and coordinate their operation towards improved resource allocation data structures.

In a further implementation form of the first aspect, the transceiver is configured to periodically receive the radio resource allocation announcement message from the second radio access network. By periodically receiving the radio resource allocation announcement message from the second radio access network, the radio access network controller is able to automatically receive the utilization parameters regarding resource allocation decisions made by the second radio access network so that it can periodically generate the improved resource allocation data structure, for example, once every night.

In a further implementation form of the first aspect, the transceiver is configured to receive a resource allocation cancellation message from the second radio access network. This enables a solution in which the previously announced resource allocations from the second radio access network can be omitted in future when making resource allocation decisions in the first radio access network. This is also useful when the second radio access network changes its operation mode from“based on resource allocation data structure” to“based on run-time measurements”.

In a further implementation form of the first aspect, the processor is configured to collect samples of the outcome of the resource allocation decisions, based on a coordination between the first radio access network and the second radio access network for interference- free resource allocation; build a resource allocation data structure for interference- free resource allocations based on the collected samples; derive interference distribution information based on the received utilization parameters; generate enhanced samples based on the built resource allocation data structure and the derived interference distribution information; and generate the improved resource allocation data structure based on the enhanced samples. This enables performing computations offline, and complex but optimal resource allocations can be run providing better performance than online resource allocation algorithms. The solution also mitigates interference between the radio access networks.

In a further implementation form of the first aspect, the processor is configured to generate, from the resource allocation data structure, a utilization parameters regarding resource allocation decisions made by the radio access network controller; and the transceiver is configured to transmit the utilization parameters to the radio access network controller of the second radio access network. This enables improving performance by compensating for interference at the edge of the second radio access network.

In an implementation form of the first aspect, the processor is configured to generate the utilization parameters based on information on resource allocation decisions collected during run-time over a time period. This allows for improved and/or optimal resource allocation decisions that may optimize performance at different time scales depending on the indicated time period. The time period may range from divisions of a second to hours and days.

In a further implementation form of the first aspect, the processor is configured to generate the utilization parameters from a resource allocation data structure.

According to a second aspect, there is provided a radio access network controller for a second radio access network. The radio access network controller comprises a processor configured to generate a utilization parameters regarding resource allocation decisions made by the radio access network controller; and a transceiver configured to transmit the utilization parameters to a radio access network controller of a first radio access network. As the first radio access network receives utilization parameters regarding resource allocation decisions made by the second radio access network, this enables improving performance by compensating for interference at the edge of the first radio access network.

In an implementation form of the second aspect, the processor is configured to generate the utilization parameters based on information on resource allocation decisions collected during run-time over a time period. This allows for improved and/or optimal resource allocation decisions that may optimize performance at different time scales depending on the indicated time period. The time period may range from divisions of a second to hours and days

In a further implementation form of the second aspect, the processor is configured to generate the utilization parameters from a resource allocation data structure.

In a further implementation form of the second aspect, the transceiver is configured to transmit a radio head announcement message to the radio access network controller of the first radio access network for creating a radio head context for transmitting the utilization parameters regarding resource allocation decisions made by the radio access network controller of the second radio access network, the radio head announcement message comprising radio head specification information of at least one radio head of the second radio access network. The use of the radio head announcement message enables a more

compressed transmission of the utilization parameters and their values between the radio access networks, since the radio head specification information will less often be required to be changed. In a further implementation form of the second aspect, the transceiver is configured to transmit a radio resource allocation announcement message to the radio access network controller of the first radio access network, the radio resource allocation announcement message comprising the utilization parameters, wherein the utilization parameters comprises for each radio head at least one of: a usage frequency of a fixed beam, and a usage frequency of a transmit power level. The use of the radio allocation announcement message enables transmission of the utilization parameters between the radio access networks.

In a further implementation form of the second aspect, the transceiver is configured to receive a request from the radio access network controller of the first radio access network for the radio resource allocation announcement message; and wherein the processor is configured to generate the utilization parameters regarding resource allocation decisions made by the radio access network controller of the second radio access network in response to the request. By receiving the request for the radio resource allocation announcement message from the first radio access network, a solution is provided where the radio access network controller needs to send the radio resource allocation announcement message only in response to a specific request, thus optimizing data transmissions between the radio access networks.

In a further implementation form of the second aspect, the transceiver is configured to periodically send the radio resource allocation announcement message to the radio access network controller of the first radio access network. Thus, by periodically sending the radio resource allocation announcement message to the first radio access network, the first radio access network is kept updated about changes in interference caused by the second radio access network for example, once every night.

In a further implementation form of the second aspect, the transceiver is configured to transmit a resource allocation cancellation message to the radio access network controller of the first radio access network. This enables a solution in which the second radio access network is able to inform the first radio access network that previously announced resource allocations from the second radio access network can be omitted in future when making resource allocation decisions in the first radio access network. This is also useful when the second radio access network changes its operation mode from“based on resource allocation data structure” to“based on run-time measurements”. According to a third aspect, there is provided method comprising receiving, by a radio access network controller of a first radio access network from a radio access network controller of a second radio access network, a utilization parameters regarding resource allocation decisions made by the radio access network controller of the second radio access network; and generating, by the radio access network controller of the first radio access network, an improved resource allocation data structure for the radio access network controller by taking into account the received utilization parameters. The disclosed solution improves

performance by compensating for interference at the edge of the first radio access network.

In an implementation form of the third aspect, the method comprises receiving a radio head announcement message from the radio access network controller of the second radio access network for creating a radio head context for receiving the utilization parameters regarding resource allocation decisions made by the radio access network controller of the second radio access network, the radio head announcement message comprising radio head specification information of at least one radio head of the second radio access network. The use of the radio head announcement message enables a more compressed transmission of the utilization parameters and values between the radio access networks, since the radio head specification information will less often be required to be changed.

In a further implementation form of the third aspect, the method comprises receiving a radio resource allocation announcement message from the radio access network controller of the second radio access network, the radio resource allocation announcement message comprising the utilization parameters, and using the radio head context to extract information from the radio resource allocation announcement message, wherein the utilization parameters comprises for each radio head at least one of: a usage frequency of a fixed beam, and a usage frequency of a transmit power level. The use of the radio allocation announcement message enables transmission of the utilization parameters and values between the radio access networks. This provides knowledge about the configuration of the interfering radio access network node.

In a further implementation form of the third aspect, the method comprises sending a request for the radio resource allocation announcement message to the radio access network controller of the second radio access network. By sending the request for the radio resource allocation announcement message to the second radio access network, the radio access network controller is able to determine the point of time when it wishes to receive the utilization parameters regarding resource allocation decisions made by the second radio access network in order to generate the improved resource allocation data structure. This also provides a means for the radio access network controller to synchronize and coordinate their operation towards improved resource allocation data structures.

In a further implementation form of the third aspect, the method comprises periodically receiving the radio resource allocation announcement message from the second radio access network. By periodically receiving the radio resource allocation announcement message from the second radio access network, the radio access network controller is able to automatically receive the utilization parameters regarding resource allocation decisions made by the second radio access network so that is can periodically generate the improved resource allocation data structure, for example, once every night.

In a further implementation form of the third aspect, the method comprises receiving a resource allocation cancellation message from the radio access network controller of the second radio access network. This enables a solution in which the previously announced resource allocations from the second radio access network can be omitted in future when making resource allocation decisions in the first radio access network.

In a further implementation form of the third aspect, the method comprises collecting samples of the outcome of the resource allocation decisions, based on a coordination between the first radio access network and the second radio access network for interference- free resource allocation; building a resource allocation data structure for interference- free resource allocations based on the collected samples; deriving interference distribution information based on the received utilization parameters; generating enhanced samples based on the built resource allocation data structure and the derived interference distribution information; and generating the improved resource allocation data structure based on the enhanced samples. This enables performing computations offline, and complex but optimal resource allocations can be run providing better performance than online resource allocation algorithms. The solution also mitigates interference between the radio access networks. In a further implementation form of the third aspect, the method generating, from the resource allocation data structure, a utilization parameters regarding resource allocation decisions made by the radio access network controller; and transmitting the utilization parameters to the radio access network controller of the second radio access network. This enables improving performance by compensating for interference at the edge of the second radio access network.

According to a fourth aspect, there is provided a method comprising collecting, by a radio access network controller of a second radio access network, a utilization parameters regarding resource allocation decisions made the radio access network controller of the second radio access network; and transmitting, by the radio access network controller of the second radio access network, the utilization parameters to a radio access network controller of a first radio access network. As the first radio access network receives utilization parameters regarding resource allocation decisions made by the second radio access network, this enables improving performance by compensating for interference at the edge of the first radio access network.

In an implementation form of the fourth aspect, the method comprises generating the utilization parameters based on information on resource allocation decisions collected during run-time over a time period. This allows for improved and/or optimal resource allocation decisions that may optimize performance at different time scales depending on the indicated time period. The time period may range from divisions of a second to hours and days.

In a further implementation form of the fourth aspect, the method comprises generating the utilization parameters from a resource allocation data structure.

In a further implementation form of the fourth aspect, the method comprises transmitting a radio head announcement message to the radio access network controller of the first radio access network for creating a radio head context for transmitting the utilization parameters regarding resource allocation decisions made by the radio access network controller of the second radio access network, the radio head announcement message comprising radio head specification information of at least one radio head of the second radio access network. The use of the radio head announcement message enables a more compressed transmission of the utilization parameters and values between the radio access networks, since the radio head specification information will less often be required to be changed.

In a further implementation form of the fourth aspect, the method comprises transmitting a radio head announcement message to the radio access network controller of the first radio access network for creating a radio head context for transmitting the utilization parameters regarding resource allocation decisions made by the radio access network controller of the second radio access network, the radio head announcement message comprising radio head specification information of at least one radio head of the second radio access network. The use of the radio allocation announcement message enables transmission of the utilization parameters between the radio access networks.

In a further implementation form of the fourth aspect, the method comprises transmitting a radio resource allocation announcement message to the radio access network controller of the first radio access network, the radio resource allocation announcement message comprising the utilization parameters, wherein the utilization parameters comprises for each radio head at least one of: a usage frequency of a fixed beam, and a usage frequency of a transmit power level. The use of the radio allocation announcement message enables transmission of the utilization parameters and values between the radio access networks.

In a further implementation form of the fourth aspect, the method comprises receiving a request from the radio access network controller of the first radio access network for the radio resource allocation announcement message; and generating the utilization parameters regarding resource allocation decisions made by the radio access network controller of the second radio access network in response to the request. By receiving the request for the radio resource allocation announcement message from the first radio access network, a solution is provided where the radio access network controller needs to send the radio resource allocation announcement message only in response to a specific request, thus optimizing data transmissions between the radio access networks.

In a further implementation form of the fourth aspect, the method comprises periodically sending the radio resource allocation announcement message to the radio access network controller of the first radio access network. Thus, by periodically sending the radio resource allocation announcement message to the first radio access network, the first radio access network is kept updated about changes in interference caused by the second radio access network for example, once every night.

In a further implementation form of the fourth aspect, the method comprises transmitting a resource allocation cancellation message to the radio access network controller of the first radio access network. This enables a solution in which the second radio access network is able to inform the first radio access network that previously announced resource allocations from the second radio access network can be omitted in future when making resource allocation decisions in the first radio access network.

According to a fifth aspect, there is provided a computer program comprising a program code configured to perform a method of the third aspect, when the computer program is executed on a computing device.

According to a sixth aspect, there is provided a computer program comprising a program code configured to perform a method of the fourth aspect, when the computer program is executed on a computing device.

According to a seventh aspect, there is provided a computer-readable medium comprising a computer program comprising a program code configured to perform a method according to the third aspect, when the computer program is executed on a computing device.

According to an eight aspect, there is provided a computer-readable medium comprising a computer program comprising a program code configured to perform a method according to the fourth aspect, when the computer program is executed on a computing device.

BRIEF DESCRIPTION OF THE DRAWINGS

In the following exemplary embodiments are described in more detail with reference to the attached figures and drawings, in which:

FIG. 1 A illustrates a schematic representation of a radio access network controller.

FIG. 1B illustrates a schematic representation of a radio access network controller.

FIG. 2 illustrates a system comprising two radio access networks. Fig. 3 illustrates a flowchart illustrating a message exchange between a first radio access network and a second radio access network.

Fig. 4A illustrates a flowchart illustrating operation of a system comprising a first radio access network and a second radio access network.

Fig. 4B illustrates a flowchart illustrating operation of a system comprising a first radio access network and a second radio access network.

FIG. 5 A illustrates a flowchart illustrating a mode for operating a radio access network with a radio access network controller.

FIG. 5B illustrates a flowchart illustrating another mode for operating a radio access network with a radio access network controller.

FIG. 6 illustrates a flowchart illustrating building resource allocation data structures in a radio access network by a radio access network controller 100.

FIG. 7 illustrates a flowchart illustrating using shared resource allocation.

FIG. 8 illustrates a flowchart illustrating enhancing stored samples.

In the following identical reference signs refer to identical or at least functionally equivalent features.

DETAILED DESCRIPTION

In the following description, reference is made to the accompanying drawings, which form part of the disclosure, and in which are shown, by way of illustration, specific aspects in which the invention may be placed. It is understood that other aspects may be utilized and structural or logical changes may be made without departing from the scope of the invention. The following detailed description, therefore, is not to be taken in a limiting sense, as the scope of the invention is defined be the appended claims.

For instance, it is understood that a disclosure in connection with a described method may also hold true for a corresponding device or system configured to perform the method and vice versa. For example, if a specific method step is described, a corresponding device may include a unit or other means to perform the described method step, even if such unit is not explicitly described or illustrated in the figures. On the other hand, for example, if a specific apparatus is described based on functional units, a corresponding method may include a step performing the described functionality, even if such step is not explicitly described or illustrated in the figures. Further, it is understood that the features of the various exemplary aspects described herein may be combined with each other, unless specifically noted otherwise.

Although aspects and examples may be described here in terms of a cloud radio access network (CRAN), it is by way of example and in no way a limitation, and other types of radio access networks can also be used. Further, although only two neighboring radio access networks are discussed in various examples, it is evident that there may be more than one interfering neighboring radio access network.

FIG. 1A illustrates a schematic representation of a radio access network controller 100 for a first radio access network. In an aspect, the radio access network controller 100 comprises a transceiver 104 configured to receive, from a second radio access network, a utilization parameters regarding resource allocation decisions made by a radio access network controller of the second radio access network. The utilization parameters may refer, for example, to statistical information regarding resource allocation decisions made by the radio access network controller of the second radio access network. The radio access network controller 100 also comprises a processor 102 configured to generate an improved resource allocation data structure for the radio access network controller by taking into account the received utilization parameters. This enables improved performance by compensating for interference at the edge of the first radio access network. Further, the radio access network controller 100 can involve the received information from the second radio access network in the next learning phase of the resource allocation data structure, where the learning phase may be performed offline.

FIG. 1B illustrates a schematic representation of a radio access network controller 110 for the second radio access network. In an aspect, the radio access network controller 110 comprises a processor 112 configured to generate a utilization parameters regarding resource allocation decisions made by the radio access network controller 110. The radio access network controller 110 also comprises a transceiver 114 configured to transmit the utilization parameters to the radio access network controller 100 of the first radio access network. As the first radio access network receives utilization parameters regarding resource allocation decisions made by the radio access network controller 110, this enables improved

performance by compensating for interference at the edge of the first radio access network. FIG. 2 illustrates a system comprising two radio access networks 204A, 204B. The first radio access network 204A comprises a first radio access network controller 100 and multiple radio heads 202A-202F connected to the first radio access network controller 100. Similarly, the second radio access network 204B comprises a second radio access network controller 110 and multiple radio heads 206A-206F connected to the second radio access network controller 110. Each of the radio heads 202A-202F, 206A-206F provides a specific radio coverage for user nodes. FIG. 2 illustrates a simplified example in which the two radio access networks 204A, 204B have interfering sets of radio heads 208, 210. In an example, the radio access networks 204 A and 204B may refer to cloud radio access networks (CRAN).

In the following, the focus is on the downlink operation of the first radio access network controller 100. The first radio access network controller 100 determines resource allocations based on data structures determined, for example, by one or more machine- learning algorithms. The first radio access network controller 100 may first collect resource allocation instances and then periodically aggregate them into learned data structures for resource allocation. In an example, the first radio access network controller 100 may collect such resource allocations over the course of one day, and then aggregate them into a learned data structure overnight, to be utilized during the next day. It is also assumed that the second radio access network controller 110 operates in a similar fashion.

Due to the nature of the learned data structures, the statistical behavior in terms of potentially chosen resource allocations for the set of interfering radio heads, for example, the radio heads 210, can be determined upfront as the potential choice of selected resource allocations for any radio head on the second radio access network 204B is represented by the learned data structure, and the collected input instances may determine the frequency of usage of the selected resource allocations. Thus, the first radio access network controller 100 and the second radio access network controller 110 can determine relevant resource allocations for their respective interfering radio heads from their learned data structure. In an example, the second radio access network controller 110 may determine utilization parameters, for example, used fixed beams for each radio head within the interfering set of radio heads 210 by querying the learned data structure. Further relevant information determined from the data structure may relate, for example, to the used transmit power (with respect to coordination of down-link operations). In another example, the utilization parameters can be collected during run-time of the second radio access network 204B. Utilization parameters, for example, statistical information regarding the resource allocation decisions made by the second radio access network controller 110 may be provided from the second radio access network controller 110 to the first radio access network controller 100 with respect to interference caused by the set of radio heads 210 on user terminals served by the set of radio heads 208. As an example, the statistical information may be represented as a mean, a variation, a percentage, etc. of the measured quantity. In an example, the statistical information may be represented as a percentage of the usage and an average power of a specific beam pattern in a radio head. The second radio access network controller 110 may collect beam usage information over the operation of a day, or alternatively, if using a learned structure, the second radio access network controller 110 may query the learned structure with, for example, all positions of user nodes from the previous day to derive the beam usage and power from that.

When information about the used resources of the second radio access network 204B is made available to the first radio access network controller 100, this information can be included in the next learning phase performed by the first radio access network controller 100. This allows for optimized resource allocations for the first radio access network 204A which are, for example, more robust to the variations of interference created by the neighboring radio access network 204B.

In an example, the second radio access network controller 110 shares to the first radio access network controller 100 only the radio head specification for the interfering radio heads. This is beneficial since that would minimize the signaling overhead between the controllers. In another example, the second radio access network controller 110 shares to the first radio access network controller 100 the radio head specification for all radio heads in the second radio access network 204B. The first radio access network controller 100 would then determine which radio head specification and which radio resource announcement are of interest to use. In one example, the determination may be based on the location of the radio heads of the first radio access network controller as related to the second radio access network controller. In another example, the determination may be based on a network planning map showing the radio environment and the deployment of the first and the second radio access networks allowing for an estimation of the interference. Further, the first radio access network controller 100 may in one example use collected measurements reported about neighboring cells by the user nodes being served by the first radio access network controller 100 to determine which part of the radio head specification to be used. In another example the first radio access network controller 100 can determine the radio heads of interest based on the generated resource allocation data structure.

In another example, there may be signaling between the first radio access network controller 100 and the second radio access network controller 110 defining which radio heads to include in the radio head specification and radio resource announcement messages. This would enable the first radio access network controller to communicate the selected radio heads to include in future radio resource announcement messages and thereby reduce the signaling. The signaling also facilitates the determination of the most significant set of interfering radio heads.

Yet in another example, first radio access network controller 100 and the second radio access network controller 110 may negotiate which radio heads to include in the radio head specification and radio resource announcement messages. This would enable the usage of information in both the first radio access network controller 100 and the second radio access network controller 110 to determine the selected radio heads to include in the radio resource announcement messages. Further, by reducing the set of radio heads to those which significantly contribute to the interference while excluding all others, this results in a radio head specification of a smaller size and, consequently, to a reduction of the signaling.

Fig. 3 illustrates a flowchart illustrating a message exchange between the first radio access network controller 100 and the second radio access network controller 110. The message exchange may first comprise a radio head announcement message 300 to announce involved radio heads of the second radio access network 200B. The message exchange may also comprise a radio resource allocation announcement message 302 to announce currently used resource allocations by the interfering radio access network. The message exchange may also comprise a resource allocation cancelation message 304 to indicate that the second radio access network controller 110 is not using the previously announced resource allocations any longer. In an example, the second radio access network controller 110 may send the radio head announcement message 300 to the first radio access network controller 100 for creating a radio head context for receiving utilization parameters regarding resource allocation decisions made by the second radio access network controller 110. The radio head announcement message may comprise radio head specification information of at least one radio head of the second radio access network. In the radio head announcement message 300, the second radio access network controller 110 may establish a list of interfering radio heads with the first radio access network 204A. For each interfering radio head of the second radio access network 204B, an identifier may be established together with other relevant data (for example, configuration and position information of the radio head). Due to changing propagation characteristics and network reconfigurations, the set of interfering radio heads 210 may also change over time. The same may be true due to power savings at time of low network utilization. Thus, in one example, the radio head announcement message 300 may be sent multiple times over the operation of the second radio access network 204B, for example, multiple times during a day.

In an example, the utilization parameters comprises for each radio head 206A-206E at least one of: a usage frequency of a fixed beam, and a usage frequency of a transmit power level.

In a further example, a possible format for the radio head announcement message may be as follows, assuming a total of x radio heads in the set of interfering radio heads 210.

| CRAN sending identifier | CRAN receiving identifier | RRH specification 1 | ... | RRH specification x | where“CRAN sending identifier” is a unique address of the sending CRAN,“CRAN receiving identifier” is a unique address of the receiving CRAN and“RRH specification” is the information to be provided about each element of the interfering set of radio heads 210. Possible information included in the“RRH specification” may relate to at least one of the position information of the radio head, its configuration (number of antennas, height of the antennas, tilt, type of antennas used etc.) or as an identifier for a radio head to be used in the context of the interference coordination between first radio access network 204A and the second radio access network 204B. With the resource allocation announcement message 302 statistical information about used resource allocations of the interfering second radio access network 204B may be provided to the first radio access network 204 A. The message 302 rests on the establishment of a context through the previously sent radio head announcement message 300.

In an example, the utilization parameters comprise for each radio head 206A-206E at least one of: a usage frequency of a fixed beam, and a usage frequency of a transmit power level.

In a further example, a possible format for the resource allocation announcement message may be as follows, assuming an announcement of resource allocation information for a total of y radio heads (not necessarily for all radio heads of the set 210):

| CRAN sending identifier | CRAN receiving identifier | RA information of RRH 1 | ... | RA information of RRH y | where“CRAN sending identifier” and”CRAN receiving identifier” are the same fields as in the radio head announcement message, while the field“RA information of RRH” may contain a data structure with the following example components:

| RRH identifier | Fixed beam identifier 1 | Usage frequency | ... | Fixed beam identifier n \ Usage frequency | Transmit power identifier 1 | Usage frequency | ... | Transmit power identifier m | Usage frequency | where“RRH identifier” is the radio head identifier provided in the previous message (the radio head announcement message),“Fixed beam identifier” is a reference to identify a used single fixed beam, which could be established, for example, through the“RRH specification” field of the previous message“radio head announcement”,“Usage frequency” is a percentage value which may represent for example how often this fixed beam is activated in the learned data structure currently active in the second radio access network 204B, and“Transmit power identifier” is a field that identifies a level of transmit power set at the radio head of interest.

In an example, the second radio access network controller 110 may send the resource allocation cancelation message 304 to the first radio access network controller 100. This message may be used to announce to the first radio access network controller 100 that the previously announced resource allocations are not valid anymore, and no update on the used resource allocations will be provided. The resource allocation cancelation message 304 may be applied if, for example, the second radio access network controller 110 switches back to a traditional scheduling mode, and does not use the previously announced resource allocations any longer.

In an example, the resource allocation cancellation message 304 may have the following format:

| CRAN sending identifier | CRAN receiving identifier | cancelation string | where“CRAN sending identifier” and“CRAN receiving identifier” are the same fields as in the two previous messages 300, 302, while the field“cancelation string” may be a special bit combination that uniquely identifies a cancelation event of the previously announced resource allocations.

In another example, a message identifier may be included in the resource allocation announcement message 302. The subsequently sent resource allocation cancellation message 304 would then refer to that message identifier.

Fig. 4A illustrates a flowchart illustrating operation of a system comprising a first radio access network having the first radio access network controller 100 and a second radio access network having the second radio access network controller 110. It is assumed here that the radio access networks have two sets of interfering radio heads, as illustrated in Fig. 2 by references 208 and 210.

At start-up time, in this example both radio access networks are operating based on some channel state information (CSI) based resource allocation algorithm, as indicated by blocks 400 and 406. Channel states together with user node positions may still be stored as training data, as indicated by blocks 402 and 408. It is also assumed here that for the collection phase, all radio heads in the interfering sets 208 and 210 alternate in the usage of time frames such that the collected training sets are essentially interference free. When enough samples have been collected, for example, over the course of one day, the radio access network controller 110 or both radio access network controllers 100, 110 determine their learned resource allocation data structures, as indicated by blocks 404 and 410. At block 412, the second radio access network controller 110 determines from its learned resource allocation data structure, for example, a usage frequency for the beams of all radio heads in the set of radio heads 210. When this information is available, the second radio access network controller 110 invokes the radio heads announcement message 300 discussed earlier in relation to Fig. 3, and establishes by this a context for the first radio access network controller 100 to interpret statistical information to be sent to the first radio access network controller 100. After establishing the context, the second radio access network controller 110 sends the resource allocation announcement message 300 and discloses by this utilization parameters, for example, the statistical information from its learned resource allocation data structures to the first radio access network controller 100. The first radio access network controller 110 then may store (as indicated by a block 414) and use this information subsequently to generate an improved resource allocation data structure with improved performance characteristics. In an example, similar information exchange can be triggered from the first radio access network controller 100 to the second radio access network controller 110. Steps 418 and 422 and blocks 420 and 424 then correspond with the previously discussed steps 300 and 302 and blocks 414 and 416, only the direction of the information exchange is different. The above discussed information exchange cycle may continue during system operation, and in an example, it may be implemented with a rather low frequency (for example, hours or even days).

Fig. 4A shows also a situation in which the second radio access network controller 110 may need to switch back to the channel state information based resource allocation mode, and the second radio access network controller 110 may send the resource allocation cancelation message 304 to the first radio access network controller 100 to announce this change in the principle interference statistics. The cancellation message is needed especially when the interference pattern changes potentially completely.

In an alternative example, the radio head announcement and resource allocation

announcement messages may be transmitted as a response to receiving a request for the information. The request could be sent before a re-leaming procedure is invoked in the other radio access network in order to obtain up-to-date information. In a further example, the resource allocation announcement message may be periodically sent between the radio access networks, for example once every night, such that updated interference estimation information can be derived in the receiving radio access network.

Fig. 4B illustrates another flowchart illustrating operation of a system comprising a first radio access network having the first radio access network controller 100 and a second radio access network having the second radio access network controller 110. The flowchart of Fig. 4B is identical with the one illustrated in Fig. 4A with the exception that in Fig. 4B a radio head announcement request message 428 precedes the radio head announcement message 300, a resource allocation announcement request message 430 precedes the resource allocation announcement message 302, and a resource allocation announcement request message 434 precedes the resource allocation announcement message 422. By issuing request messages the first radio access network controller 100 and the second radio access network controller 110 are able to indicate when they require information to be sent to them.

FIG. 5A illustrates a flowchart illustrating a mode for operating a radio access network with a radio access network controller. FIG. 5B illustrates a flowchart illustrating another mode for operating a radio access network with a radio access network controller.

The first mode illustrated in Fig. 5A is operating (i.e. scheduling) on collected channel state information 500. Cellular systems are traditionally operated this way. The first mode requires that channel state information of the terminals is collected (block 502). The radio access network controller then runs an algorithm to determine suitable resource allocations. As indicated by a block 504, the set of assigned terminals to radio heads, as well as beam, filter and MCS settings per assigned terminal are selected based on the channel state information.

The second mode illustrated in Fig. 5B is operated based on position information and use machine learning operation, as indicated by a block 506. In other words, resource scheduling may be performed by a trained resource allocation data structure (acquired through, for example, machine learning processing) and takes as input positions of terminals. In other words, for the learned resource allocation data structure the positions of the terminals in the radio access network area are the input features, as indicated by a block 508. As indicated by a block 510, the output class, i.e. the output variables of the learned resource allocation data structure is the set of assigned terminals to radio heads, as well as beam, filter and MCS settings per assigned terminal. At run-time, for every scheduling instance (i.e. a transmission time interval (TTI)) the set of input features is passed to the learned resource allocation data structure that generates a resource allocation decision.

FIG. 6 illustrates a flowchart illustrating building resource allocation data structures in a radio access network by the radio access network controller 100. Initially, at run-time the system operates in the channel station information (CSI) based configuration operating some fast CSI based resource allocation algorithm, as indicated by a block 600. At this point, the radio access network controller already stores training instances, as indicated by a block 602. For these training instances, the position of user nodes is important as this information comprises the input features. However, in order to determine a high-quality output class, i.e. resource allocation, also the exact channel state information (antenna-to-antenna phase and magnitude) is stored. This information may not necessarily be used as an input feature to a machine learning process, but is used to determine the best possible resource allocation offline, which is then used as output class.

If the amount of samples is not sufficient (block 604), the processing returns back to the block 600. On the other hand, if the amount of samples is sufficient (block 604), the processing proceeds to a block 606. In the block 606, offline optimization for resource allocation is run using the samples. Then, at a block 608, machine learning based training is run on position and optimized resource allocation. As these computations are executed offline, complex but optimal resource allocations can be run, which provide in general a better performance than the online CSI based resource allocation algorithm of the CRAN.

FIG. 7 illustrates a flowchart illustrating using shared resource allocation. It is now assumed that the precise channel and applied beam with respect to the interfering radio head are known. A block 700 also assumes that resource allocation data structures have been built for optimized interference- free resource allocation. In other words, all radio heads in the interfering radio heads sets 208 and 210 alternate in the usage of time frames such that the collected training sets are essentially interference free. It is also assumed here that

CSI pilots of user nodes are also overheard by the interfering radio heads of a second radio access network comprising the radio access network controller 110. The radio access network controller 110 then sends the channel states back to the radio access network controller 100, as indicated by a message 702. The messages 300 and 302 and the block 414 have been already discussed in detail in the description of Fig. 4A, and this section is not repeated here. As a summary of the messages 300 and 302, the radio access network controller 110 uses these message to indicate to the radio access network controller 100 how often a specific radio head uses a specific beam. Based on this information as well as the channel information from the interfering radio heads 210 to user nodes of the radio access network 204A, a suitable resource allocation can be determined by the radio access network controller 100.

At a block 704, the already collected training instances are enhanced with sample information from the interference statistics received from the radio access network controller 110. More precisely, for each training instance, the interfering radio heads of the radio access network 204B are supposed to employ a beam according to their beam statistics. The impact of this can be accounted for in an accurate way by the offline exhaustive search algorithm, as the channel information from the interfering radio heads of the radio access network 204B towards the user nodes of radio access network 204A is available.

At a block 706, offline optimization for resource allocation may be run using the enhanced samples and at a block 708 machine learning training may be run on the optimized enhanced resource allocation. The output class that is determined offline will include the impact of the statistical information of the interference from the radio access network 204B, making the edge resource allocation in the radio access network 204A essentially more robust to interference. This will make the re-learning due to variations in the usage of the radio access network less complex, since the radio access network 204A does not have to collect all new training instances to determine the output class and make the radio access network 100 faster to adopt the learned resource allocation data structures.

FIG. 8 illustrates a flowchart illustrating enhancing stored samples. The step of enhancing the stored samples as illustrated at the block 704 of Fig. 7 can in one example be made as shown in Fig. 8.

A processed sample number N is first set to 1, as indicated by a block 802. At a block 804 a position is selected from the stored sample N. For each stored sample of position the interference distribution from the beams in the set of radio heads 210 is sampled. A random number RAND is selected at a block 806. The sampling uses the individual probability that a beam is used and the power it is using (named as Bx in Figure 8), as indicated by a block 808. At a block 810, the estimated interference from each beam is summed for the selected position, and at a block 812, the estimated interference is added to the selected stored sample. At a block 814, the new obtained sample (named as an enhanced sample in Figure 8) is stored for later usage.

At a block 816 if the amount of samples is not yet sufficient, the processing returns back to the block 806. Thus, the steps of sampling the interference from the radio heads 210 are repeated until a sufficient set of samples has been obtained to represent the interference from the reported usage of beams in the resource allocation announcement message. Then, at a block 818, the value of N is increased by one. If the value of N is not larger than the number of samples, the processing returns back to the block 804.

The illustrated steps may be repeated for all the stored samples in a database without external interference. The new enhanced samples can then be used for offline optimization and further machine learning to obtain resource allocation data structures that accommodates for the statistical interference from the radio access network 204B.

In a further example, the input features may even be extended by the quality-of-service requirements of use nodes. Some of the user nodes may relate to dependable applications such as critical machine-to -machine applications where a certain small data packet needs to be delivered within a given deadline, while the rest of the user nodes are receiving data from an elastic application (such as normal web traffic). Then, during offline computation of the optimal resource allocation such different levels of service requirement can be taken into account, ensuring that the set of critical terminals receives their data with high reliability taking the neighboring interference into account, while the expected sum goodput for the remaining user nodes is maximized.

At least one of the discussed examples provides the benefit of a better performance at an edge of a radio access network, which can be parameterized according to a specific choice of the reliability-goodput trade-off, and can be even modularized down to an individual level of the user nodes and their service requirements. The functionality described herein can be performed, at least in part, by one or more computer program product components such as software components. According to an example, the radio access network controller may comprise a processor configured by the program code when executed to execute the examples and embodiments of the operations and functionality described. Alternatively, or in addition, the functionality described herein can be performed, at least in part, by one or more hardware logic components. For example, and without limitation, illustrative types of hardware logic components that can be used include Field-programmable Gate Arrays (FPGAs), Program-specific Integrated Circuits (ASICs), Program-specific Standard Products (ASSPs), System-on-a-chip systems (SOCs), Complex Programmable Logic Devices (CPLDs), Graphics Processing Units (GPUs).

The functionality of the radio access network controller may be implemented by program instructions stored on a computer readable medium. The program instructions, when executed, cause the computer, processor or the like, to perform the steps of the encoding and/or decoding methods. The computer readable medium can be any medium, including non-transitory storage media, on which the program is stored such as a Blu-Ray disc, DVD, CD, USB (flash) drive, hard disc, server storage available via a network, a ROM, a PROM, an EPROM, an EEPROM or a Flash memory having electronically readable control signals stored thereon which cooperate or are capable of cooperating with a programmable computer system such that an embodiment of at least one of the inventive methods is performed. An embodiment of the invention comprises or is a computer program comprising program code for performing any of the methods described herein, when executed on a computer. Another example of the invention comprises or is a computer readable medium comprising a program code that, when executed by a processor, causes a computer system to perform any of the methods described herein.

Although the subject matter has been described in language specific to structural features and/or acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed as examples of implementing the claims and other equivalent features and acts are intended to be within the scope of the claims.

It will be understood that the benefits and advantages described above may relate to one example or may relate to several examples. The examples are not limited to those that solve any or all of the stated problems or those that have any or all of the stated benefits and advantages. It will further be understood that reference to 'an' item may refer to one or more of those items. The steps of the methods described herein may be carried out in any suitable order, or simultaneously where appropriate. Additionally, individual blocks may be deleted from any of the methods without departing from the spirit and scope of the subject matter described herein. Aspects of any of the examples described above may be combined with aspects of any of the other examples described to form further examples without losing the effect sought.

The term 'comprising' is used herein to mean including the method, blocks or elements identified, but that such blocks or elements do not comprise an exclusive list and a method or apparatus may contain additional blocks or elements. Although the invention and its advantages have been described in detail with reference to specific features and embodiments thereof, it is evident that various changes, modifications, substitutions, combinations and alterations can be made thereto without departing from the spirit and scope of the invention as defined by the appended claims. The specification and drawings are, accordingly, to be regarded simply as an illustration of the invention as defined by the appended claims, and are contemplated to cover any and all modifications, variations, combinations or equivalents that fall within the scope of the invention.