Communication system, communication unit and method of power saving therein

The present invention relates to capacity preservation within a communication unit operating in a communication system. The present invention is applicable, but not limited to, wireless link permission control and / or scheduling management in wireless communication systems, and / or processing power control in overload or flow control.

Wireless communication systems, such as cellular telephony or personal mobile wireless communication systems, are commonly referred to as multiple Base Transceiver Stations (BTSs) and often Mobile Stations (MSs). A wireless telecommunications link is provided between the plurality of subscriber units. A base station controller (BSC) is provided, and each BSC controls one or more BTS.

Wireless communication systems are distinguished from fixed communication systems such as the Public Switched Telephone Network (PSTN) primarily because mobile stations move between BTSs (and / or different service providers). And in doing so, you will encounter a fluctuating radio propagation environment.

In wireless communication systems, each BTS is associated with a specific geographic coverage area (ie, cell). A specific range defines a coverage area within which the BTS can maintain acceptable communication with the MS operating within its responsible cell. Often, these cells combine to form an expanded coverage area.

Today's communication systems, both wireless and wire-line, require the transfer of data between communication units. In this context, data also includes voice communication. Such data transfer needs to be performed effectively and efficiently in order to optimize the use of limited communication resources.

One such wireless communication system is the 3rd Generation Partnership Project (3GPP) standard, which is the Universal Mobile Telecommunication System (UMTS) radio known as UTRAN. Supports Wideband Code-Division Multiple Access (WCDMA) related to access networks. The European Telecommunication Standard Institute (ETSI) defines the 3GPP standard.

In UMTS terminology, the base station (BTS) is called node B, and the base station controller (BSC) is called a radio network controller (RNC).

Within UTRAN, many communication resources need to be managed effectively, for example. (i) Air interface (ie, CDMA power and code) resources, eg, separate interface resources for each cell.

(ii) For example, a separate resource for each node B, for example, a backhaul resource that supports a limited capacity E1 link. (iii) For example, the hardware / software resources of Node B that manage the processing power of Node B (eg, as defined by the microprocessor, back-plane networking, etc.) are achievable within the cell. It can be a constraint on data throughput.

(iv) RNC hardware / software resources. Some traditional systems require the same (or at least similar) set of QoS management algorithms to be applied to each resource. These QoS management algorithms (i) Permission control: Executed when a new cell enters the system / cell. The permission control has the purpose of determining whether or not QoS is maintained for all connections when a new cell is permitted.

(ii) Scheduling: Execute every frame. Scheduling has the purpose of ensuring that the number of data packets submitted for transmission does not exceed the capacity available in a short period of time, such as 10 msec frames.

(iii) Overload control: The above-mentioned permission control and / or scheduling mechanism is used to prepare the situation when an abnormality occurs in those functions. Actions may also include "call pre-emption," which drops low-priority calls from the system.

(iv) Flow control: This can be considered a subclass of overload control, at least related to overload control. Flow control reduces the source rate because the system is not congested.

When all four QoS control mechanisms are executed for each of many UTRAN resources, mega instruction per second (MIPS) consumes a large amount of processing power. The effect of processing is mainly felt on the wireless network controller (RNC) in the 3GPP system, and also on node B.

The inventor of the present invention has recognized and confirmed that the current QoS management algorithm treats all resources equally. Therefore, there is no consideration of their relative importance in limiting the data throughput of the network. In some, or perhaps in most cases, some of the UTRAN resources may be overdimensioned against others. The MIPS consumed when performing QoS management functions on these (relatively) overvalued resources will be wasted. This is because other resources represent bottlenecks in limiting data throughput performance.

<p> Therefore, in the field of the present invention, it is necessary to improve the QoS management methodology so as to alleviate the above-mentioned drawbacks, especially in the cellular base site resource in the wireless communication system (transmission delay is a constraint).</p>

<p> According to the first aspect of the present invention, the communication system according to claim 1 is provided. According to the second aspect of the present invention, there is provided the method of reducing the capacity consumption in the system management function (for example, RNC) in the communication system according to claim 15.</p><p> According to a third aspect of the present invention, the 3GPP wireless communication system according to claim 23 is provided. According to a fourth aspect of the present invention, the wireless network controller according to claim 24 is provided.</p><p> According to a fifth aspect of the present invention, the storage medium according to claim 25 is provided. According to a sixth aspect of the present invention, the wireless network controller according to claim 26 is provided.</p><p> The invention concept of the present invention is to provide a mechanism for identifying bottleneck resources in a wireless communication system. In this regard, the provision and management of access to resources will be centered mainly on bottleneck resources. Management of other resources can be adapted accordingly to preserve processing power (in MIPS units). When applied to a 3GPP system, it can preserve the processing power at the RNC (and, to a lesser extent, node B) so that the RNC (and / or node B) operates more efficiently and effectively. Can be designed to. Thus, the improved RNC (and / or Node B) can handle as many erlangs as similar elements in traditional systems at low cost due to the low processing power required.</p><p> In summary, the invention concept of the present invention provides a mechanism for identifying resource bottlenecks in a communication system. In addition, the concept of invention also proposes a mechanism for prioritizing management algorithms primarily to center bottleneck resources. Once access to bottleneck resource performance is controlled, the level of the management algorithm MIPS is lowered (if necessary) and applied to other resources. Thus, when the bottleneck restrictions imposed by other resources make it ineffective, the processing capacity can be preserved by avoiding resource management.</p>

Hereinafter, an exemplary embodiment of the present invention will be described with reference to the accompanying drawings. In the context of the present invention, any reference to capacity preservation shall be considered to include the preservation of processor resources, eg, in terms of instruction per second (MIPS).

It is noteworthy that preferred embodiments of the present invention selectively apply the QoS algorithm to one or more system resources, but not to all system resources to the same extent. A preferred application of the present invention is a 3GPP wireless communication system architecture. In this regard, the present invention introduces the concept of "active set". An "active set" is a list of bottleneck UTRAN resources that perform a significant number, and preferably all, of the QoS management algorithms. In this regard, QoS management algorithms preferably include scheduling and authorization control.

In summary, the invention concept of the present invention provides a mechanism for identifying resource bottlenecks in a communication system. We propose a mechanism for prioritizing QoS management algorithms in order to put bottleneck resources at the center. Once the bottleneck resource performance is optimized, reduce the level of control (if necessary) and apply it to other resources. If the bottleneck constraints imposed by another resource do not provide an effect, processing power can be preserved by avoiding running management algorithms on the resource.

With reference to FIG. 1, Schematic Figure 100 shows the throughput capacity of each UTRAN resource. In this figure, the throughput of each resource can be visualized as a pipe of a given size. In the example of Figure 1, I_ub/ I_urThe backhaul resource 115 is clearly a bottleneck in the delivery of communications services. This is because this resource has the smallest diameter (data throughput) pipe when compared to the RNC or Node B hardware / software resources 105,110, or air interface resources 120.

According to a preferred embodiment of the present invention, once a "bottleneck" has been identified, efficiency improvement algorithms such as execute permission control and scheduling algorithms are applied only to this "bottleneck" resource. In this way, the "system" can preserve its processing requirements and provide the same level of service as compared to the current system.

From this, with reference to FIG. 2, a cellular telephone communication system 210 corresponding to the UMTS Air Interface is schematically shown according to a preferred embodiment of the present invention. That is, the embodiments described relate to a wideband code division multiple access (WCDMA) standard related to UTRAN.

A plurality of subscriber units 212 to 216 communicate with a plurality of nodes B 222 to 232 via the selected air interfaces 218 to 221. The number of subscriber units 212-216 and nodes B222-232 is shown in a limited number for clarity purposes only. Each node B222 to 232 contains one or more transmission / reception units, and I_ubCommunicate with the rest of the cellular system infrastructure through interface 235. Nodes B222-232 are external networks, such as the Public Switched Telephone Network (PSTN) or Internet 234, and Wireless Network Controller Stations (RNCs) 236-240, as well as any number of Mobile Switching Centers (MSCs) 242. And it can be connected via the responsible GPRS compatible node (SGSN: Serving GRRS Support Node) 244.

Each RNC236-240 controls one or more nodes B222-232. Each MSC242 (shown only one for clarity) provides a gateway to the external network 234, while the SGSN244 links to the external packet data network.

Operations and Management Center (OMC) 246 is operably connected to RNC236-240 and Nodes B222-232 (for clarity, only Node B226 and Node B228 are shown) and will be understood by those skilled in the art. Operates and manages functions within the cellular telephone communication system 210 so as to be performed.

According to a preferred embodiment of the present invention, one or more RNCs 236 to 240 are configured to include a bottleneck detection function. The functionality of the bottleneck detection function will be specifically described below with respect to the decision process of adding the identified bottleneck resources to the "active set" or excluding the identified bottleneck resources from the "active set".

In addition, there is also a scheduler that is typically executed on one or more RNCs 236-240 and schedules the transmission of data packets. The scheduler is operably coupled to the bottleneck detection function and is configured to schedule data packets according to a determined priority. That is, the bottleneck resource specified by the RNC schedules the data packet depending on whether or not it allows the data packet to pass through it.

Further, in a preferred embodiment of the present invention, there is also a permission control function / algorithm that is typically performed on one or more RNCs 236-240. The permission control / function / algorithm is operably coupled to the bottleneck detection function and is configured to allow the user requesting access according to the determined priority. That is, the permission control function / algorithm is based on whether or not the bottleneck resource specified by RNC corresponds to the transmission of the requesting user.

More generally, one or more RNCs effectively perform improved system management functions and program the RNCs in any suitable manner according to preferred embodiments of the present invention. For example, a new device can be added to a conventional communication unit (eg, RNC236). Alternatively, existing parts of a conventional communication unit can be adapted, for example, by reprogramming one or more processors in it. In this way, the required adaptation (introducing a random access detector or adapting a scheduler and / or authorization control function) can be applied to floppy disks, hard disks, and programmable read-only memory (PROM:). It can be implemented in the form of processor-executable instructions stored on storage media, such as Programmable Read Only Memory), Random Access Memory (RAM), or any combination of these or other storage media. ..

A preferred embodiment of the present invention provides improvements to the efficient use of, for example, one or more QoS management algorithms, such as bottleneck detectors and schedulers and / or authorization control functions / algorithms related to the operation of the RNC. As explained with reference, it is conceivable that these functions / algorithms may reside within other network elements. For example, the concept of invention for adapting system performance in response to detected bottleneck resources may be implemented daily or weekly. In this regard, the above-mentioned function / algorithm may be placed, for example, in OMC246, as opposed to the dynamic adaptation that occurs when the above-mentioned function / algorithm is preferably placed in RNC.

It is also possible that such aforementioned functions / algorithms may reside in other network elements or may be distributed among two or more such network elements in a wireless communication system. This is within the assumption of the invention. Furthermore, alternative wireless communication architectures can also benefit from the concept of the invention described herein, and the concept of the invention should not be considered to be limited to the specific configuration shown in FIG.

In the first embodiment of the invention, the "active set" is configured to include all UTRAN resources. The QoS algorithm is optimized to take advantage of the bottleneck detector discovery in RNC. In this first embodiment, all resources are considered to be in the "active set". The RNC determines the likelihood that each resource will be a data throughput bottleneck that limits the data throughput performance of the system. This determination is preferably made using one of the following measurements. This will be described further later.

(i) Frequency of starting the overload control function, or (ii) By measuring the utilization rate of resources. Here, it is examined how each QoS mechanism is adapted in order to correspond to the invention concept of the present invention.

<u style="single">Scheduler algorithm</u> In a preferred embodiment of the invention, the scheduler algorithm in UTRAN executes, for example, every radio frame and schedules all data packets that are aligned for transmission in the next frame.

A known scheduler operation captures the data packet at the head of the data queue and, in series, for each data packet, whether the introduction of that data packet overloads any of the many resources. Is determined. All resources are checked in the operation of known schedulers and equal importance is assigned to the resources. Resources can be, for example, code consumption, capacity consumption, bypass relay bit rate consumption, and the like.

If the scheduler determines that the introduction of a data packet overloads a particular resource, the scheduler terminates the scheduling operation. Alternatively, the scheduler operation is terminated when the data packet queue is exhausted.

One of the problems associated with this known method is that it checks each data packet unnecessarily and checks the packet consumption of each data for every resource. The inventor of the present invention has confirmed that this is wasteful, especially in resource-rich situations. For example, the downlink scheduler may have code restrictions. That is, when the code resource runs out, scheduling is stopped. When the scheduler goes down, the capacity and utilization of bypass relays are very low, which can be, for example, 50%, but each scheduled data packet is determined for the consumption of these resources. Therefore, this adds an unnecessary load to the scheduler processor.

An improved scheduler operation adapted according to a preferred embodiment of the present invention is shown in Flowchart 300 of FIG. First, the RNC identifies a major bottleneck resource, eg, resource "A", in step 302. Then, as shown in step 305, the scheduler operation is started by fetching the data packet at the beginning of the aligned data stream.

In a preferred embodiment of the invention, first the determination is made for the resource that is most likely to limit the data packet throughput, that is, the resource that typically reaches 100% utilization before others. That is, this (bottleneck) resource is assigned the highest priority in the scheduling decision process. The limit imposed by the bottleneck resource, called the resource "A", is determined in step 310. Of note, as in step 315, each time a data packet is added to the schedule, the scheduler evaluates the impact of each received data packet on this resource "A" only.

After this, once the resource "A" is exhausted, the process checks for the depletion of the second resource "B", as shown in step 320. Resource "B" is expected to be the next highest priority resource, the worst-to-second resource identified from a large number of resources. Therefore, the resource "B" is considered to be operating at the maximum utilization because the resource "A" is usually a restricted resource at this point. However, this relationship may not always be true, so it is preferable to check the remaining resources. If resource "B" is being used to the maximum, in step 325, data packets are excluded from the schedule until resource "B" utilization 100%. It should be noted that the consumption of resource "A" is &lt;100% at this point.

In the improved embodiment of the present invention, an intellectual determination is made as to which data packet (plurality of data packets) is excluded from the schedule. In this regard, it would be better to exclude the data packets that use the most resource "B". For example, if resource "B" is a bypass relay bandwidth, exclude data packets that consume the largest size (in bits) from the schedule.

This process continues, for example, as shown in steps 330,335, and takes in the next higher priority resource until a schedule with 100% utilization is required for all resources. The scheduler process then completes, as shown in step 345.

It is clear that the above scheduling algorithm requires only 1 / n of the process steps of a known exhaustion check algorithm. Where n is the number of resources to check. Further, as shown in the mapping table, the bottleneck detector employs "average utilization" in step 340 to order / prioritize each resource. In this way, bottleneck resources are the resources with the highest average utilization rate. In other embodiments, it is considered that a maximum, variable, or fixed load factor can also be used.

<u style="single">Permit control algorithm</u> The permission control is a process for determining whether or not a resource may be granted to the requesting communication device. If the authorization control algorithm does not check the authorization request against currently available resources in the optimal sequence, the authorization control operation may be inefficient. A suitable mechanism for performing permission control is shown in Flowchart 400 of FIG.

A suitable mechanism begins at step 402, where the RNC identifies resource "A" as the major bottleneck resource. At step 405, the RNC receives a request for a call authorization attempt. Then, if this call is allowed, it is determined whether or not resource allocation is required. For example, in step 410, the resource "A" is larger than the usable capacity of the resource "A". If the requirement for resource "A" is greater than the capacity for resource "A", then the call is not allowed, as shown in step 430. Resource "A" has previously been identified by the RNC as a bottleneck resource in terms of data throughput. As a result, resource "A" is assigned the highest priority in the authorization control process.

In step 410, if resource "A" has sufficient capacity to handle the call, in step 415, the request for a second resource, eg, resource "B", is the capacity provided by resource "B". It is determined whether or not it is larger than. If the request for resource "B" is greater than the available capacity of resource "B", then the call is not allowed, as shown in step 430.

Similarly, in step 415, if resource "B" has sufficient capacity to handle the call, then in step 420, the requirement for a third resource, eg, resource "C", is that of resource "C". Determine if it is larger than the usable capacity. If the request for resource "C" is greater than the available capacity provided by resource "C", then the call is not allowed, as shown in step 430. This process continues until all resources have been checked, at which point the call is allowed, as shown in step 425.

According to a preferred embodiment of the invention, a tracking process is introduced to count the failure rate of authorization attempts for individual resources. If the rate of authorization failure for a given resource exceeds a given threshold compared to the total number of authorization requests measured in the previous given time interval, the resource is moved further up in the list of resources to check. It is good. In this way, the resources will be checked early thereafter. Further, similar to the scheduling operation described above, the resources "A", "B" and "C" (and any other) are prioritized in order of probability of authorization failure. This ordering process is preferably based on failure count statistics.

In this way, the number of checks required for a call authorization attempt that ultimately fails is minimized. That is, the resource check hierarchy is configured so that the authorization control process is more likely to fail in the first step of checking the resource "A".

This algorithm has the advantage of being extremely effective during periods of high load, when interruptions occur regularly, and when the RNC processor is already under heavy load stress. According to the second embodiment of the present invention, the active set is regarded as a subset of UTRAN resources. In this regard, a reduction in MIPS is achieved by performing QoS management only on the resources in the active set, i.e. one or more resources identified as bottleneck UTRAN resources. In the second embodiment, it is worth noting that the overload detection and reaction mechanism is eligible for all UTRAN resources and is always in the "active" mode of operation.

Also, in this second embodiment, the active set of UTRAN resources is most preferably configured adaptively so that the resources can be dynamically added to or excluded from the active set. Preferably, the cell settings are configured to include all UTRAN resources in the active set.

Then, if one or more overload alerts are issued corresponding to a UTRAN resource at a given time interval immediately before, or in a given number of scheduling / authorization control events immediately before, that particular UTRAN resource is activated. It is possible to add it to the set.

Similarly, if a given UTRAN resource limit is not recorded as one of the reasons that prevents packets to be scheduled or calls to allow, it is possible to exclude that particular resource from the active set. Again, this determination is performed at a given time interval immediately before or at a given number of previous scheduling and / or authorization control events. In addition, it is preferable that at least one UTRAN resource remains in the "active set" list. In this case, all relevant QoS mechanisms (permission control, scheduling, flow control, overload control) are applied to one UTRAN resource remaining in the active set list.

<tables num="1"><img file="JP4348367B2_D0001.tif" /></tables> Performs all QoS management functions when the UTRAN resource is in the active set. It should be noted that in an actual system, there are far more UTRAN resources than the limited number shown in Table 1.

In the improved mechanism of the second embodiment of the present invention, the active set is associated with each of the QoS management mechanism, permission control, and scheduling. Certain QoS mechanisms are "executed" only for those UTRAN resources in the active set. Preferably, each QoS management mechanism is configured to operate on their respective time scale. For example (i) The permission management function can be configured to manage the average number of resources on a relatively long time scale (eg, in seconds).

(ii) The scheduler can manage schedule resources on a shorter time scale (for example, about 10 msec). It is also conceivable to enable different overload control mechanisms to be activated on different time scales. For a few resources (assuming an air interface in this example), the pipe size of the ideological resource undergoes relatively large fluctuations on the short time scale, but moderately on the longer time scale. It may be constant. So, for example, it may be important to perform air interface scheduling, but it may not be necessary to perform air interface permission control.

<tables num="2"><img file="JP4348367B2_D0002.tif" /></tables> Table 2 shows three active sets for this improvement over the second embodiment, one for each QoS mechanism. Again, each system has far more UTRAN resources than the limited number shown in Table 2.

Alternatively, a set of QoS mechanisms that manage resources on a given time scale can be defined. In this case, for each time scale, UTRAN resources to which the applicable QoS mechanism can be applied can be specified, as shown in Table 3 below.

<tables num="3"><img file="JP4348367B2_D0003.tif" /></tables> Table 3 shows three active sets, one for each QoS management time scale, and applies different QoS mechanisms to each time scale.

In this improvement of the second embodiment, when an overload control alarm is issued to a UTRAN resource, the resource can be added to the active set list according to the QoS mechanism / QoS management time scale. Conceivable. Preferably, the measurements are made over the corresponding QoS management time period. Similarly, if a limit on a UTRAN resource is not recorded as one of the reasons that prevents packets to be scheduled or calls to be allowed, then the resource is active for a given QoS mechanism / QoS management time scale. Can be excluded from the set list. Again, it is conceivable to make this decision at a given time interval immediately before, or at a given number of scheduling and / or authorization control events immediately before.

In addition, the aforementioned mechanisms that add or exclude resources from the active list are overloaded for that particular UTRAN resource during the last time interval or a given number of scheduling and / or authorization control events. No load alarm should be issued. In addition, it is preferred that at least one UTRAN resource remains in the "active set" list that matches the QoS management mechanism or QoS management mechanism time scale.

In yet another improvement of the second embodiment, it may be possible to reduce or eliminate the dependence on the overload control alarm activated as a mechanism for changing the active set. An alternative approach could be to make regular measurements of each load state of the resource and make addition or deletion decisions based on the current load state. Such a mechanism has the advantage of reducing the occurrence of (undesirable) overload.

Consider the case where there is only one active set, just for the sake of simplicity. When the cell is set up (ie, when node B is powered on), all UTRAN resources are in the active set. Then, the load is regularly measured for each of the UTRAN resources. You can also take the average of the load measurements, or, for example, the xth percentile (x)^th Percentile) can also be considered. Either way, in this example, the load measurements are expressed as a percentage of the total UTRAN resource capacity.

That is, for example, if the load on the UTRAN resource is less than Threshold_1 for a predetermined time period T_1, for example, this UTRAN resource is excluded from the active set. Further, for example, if the load of a UTRAN resource is greater than, for example, Threshold_2, during a predetermined time period T_2, the UTRAN resource is added to the active set list.

It is also within the scope of the present invention that any combination of the above-mentioned invention concepts can be adopted. For example, it is conceivable that an active set can be provided for each QoS mechanism or each QoS mechanism time scale. In this regard, for example, at regular intervals, all UTRAN resources are determined for a particular load condition measured on a given time scale (eg, 10,100 or 1000 msec). If a given load criterion (threshold) is met, the UTRAN resource is added to or excluded from the active set list.

In addition, whenever an overload on a resource is identified, it may be possible to immediately add the resource to the active set. It is also within the scope of the present invention to be able to "offline" decisions about a particular QoS management mechanism to apply to a given resource. In this case, the judgment may be encoded as an OMC parameter. Offline dimensional and / or trial and error experience and / or expert systems can also be used as part of this process.

Although the preferred embodiments of the present invention have been described above in connection with the UTRAN 3GPP system with reference to the boundary identifier, the concept of the invention is, for example, other remote communication systems including a core network or a backbone network, wireless. Or it is considered to be equally applicable to wire lines.

For completeness, it is worth clarifying how to actually take advantage of the reduced complexity (capacity per MIPS) requirement. However, those skilled in the art will appreciate that the concept of the invention described herein can be utilized in a number of other ways and therefore the concept of the invention is not limited to the mechanisms described below.

When a wireless communication network is currently in place, the RNC needs to have a processing power that is approximately equal to the processing power that would be needed to handle the worst-case situation. In this regard, the RNC needs to be configured to be sufficient for all UTRAN resources. For this reason, some inefficiencies usually occur in the initial network installation. This is because it is expected that RNCs are usually underutilized in some respects. In addition, the inefficiency of the RNC processor increases as the network load increases and more nodes B are added.

Therefore, it can be seen that at least the following advantages can be obtained from the improved QoS management methodology for executing the bottleneck detection algorithm as described above. (i) By monitoring the load on the RNC processor resources, the MOC can determine whether or not to add extra RNC processor resources.

(ii) The rate at which RNC cards must be added to handle higher network loads is reduced. (iii) When part of the QoS processing is executed on node B (for example, hardware / software permission control), even with this technique, if the node B hardware / software resource is not a barrier resource, signaling and call setting delay Decreases.

Although the specific and preferable embodiments of the embodiment of the present invention have been described above, those skilled in the art can easily apply modifications and modifications to the preferred embodiments, and they fall under the concept of the present invention. That is clear.

As described above, a communication system and a method for reducing capacity consumption in the communication system have been provided, which significantly reduce the drawbacks of the above-mentioned prior art.

<figref num="1">The schematic diagram of an example of the throughput capacity of each UTRAN resource determined and responded to according to a preferred embodiment of the present invention.</figref><figref num="2">FIG. 3 is a block diagram of a 3GPP cellular radio communication system configured to correspond to various invention concepts of preferred embodiments of the present invention.</figref><figref num="3">A flowchart of a scheduler that schedules data packet transmission according to the number of resources used for transmission according to the invention concept of a preferred embodiment of the present invention.</figref><figref num="4">A flowchart of permission control for transmission of a data packet according to the number of resources used for transmission according to a preferred embodiment of the present invention.</figref>