Patent Application
20130230019
The present disclosure relates to a radio node and a method in the radio node of reducing a signal to noise and interference ratio requirement for a transmission in a scheduling interval.
The Universal Mobile Telecommunication System (UMTS) is one of the third generation mobile communication technologies designed to succeed GSM. 3GPP Long Term Evolution (LTE) is a project within the 3rd Generation Partnership Project (3GPP) to improve the UMTS standard to cope with future requirements in terms of improved services such as higher data rates, improved efficiency, lowered costs etc. The Universal Terrestrial Radio Access Network (UTRAN) is the radio access network of a UMTS and Evolved UTRAN (E-UTRAN) is the radio access network of an LTE system. In an E-UTRAN, a user equipment (UE) 150 is wirelessly connected to a radio base station (RBS) 110a commonly referred to as an eNodeB or eNB (E-UTRAN NodeB), as illustrated in
Radio Resource Management (RRM) plays a crucial role in how resources in a wireless communications system are used. In particular, RRM techniques in wireless communications systems are of high importance as they largely influence how efficiently the system is used. Two RRM functionalities, scheduling and Link Adaptation (LA), play a central role for resource allocation and have a significant influence on system performance. These two RRM functionalities work tightly together. The scheduling allocates a certain part of a spectrum, i.e. of the available frequency resources, to a certain UE during a certain amount of time. The LA computes how many bits that may be transmitted in the scheduled part of the frequency resource given operating channel conditions, a transmit power and a desired probability of a correct reception.
The scheduling and LA are used in a way that optimizes a frequency resource utilization in every cell separately. Other RRM functionalities promote the coordination between different cells, and are also very important for a good wireless communications system performance. For instance, schemes that try to mitigate and coordinate interference among different cells—commonly referred to as Inter-Cell Interference Coordination (ICIC) schemes—constitute one of the most intriguing areas in RRM. ICIC schemes try to coordinate a generated inter-cell interference between cells so that the effect of the generated interference becomes less detrimental, typically by utilizing feedback and exchanging information between neighboring radio base stations. ICIC schemes usually work on a slower basis than the scheduling and LA in order to mitigate the increased overhead and complexity arising from the extra information exchange, signaling, and processing needed for ICIC.
A main operating principle in conventional scheduling and LA is to transmit as much data bits as possible given a certain frequency resource allocation, or expressed in another way, to find a smallest possible frequency resource allocation given a certain number of data bits to transmit. At the same time, a certain probability of correct reception under the operating channel conditions should be satisfied. A commonly used criterion for the probability of correct reception is a Block Error Rate (BLER) target. The main operating principle is thus to maximize the spectral efficiency measured in bits per second and per Hz (bps/Hz) for the allocated resources. The more bits that may be transmitted over a certain part of the frequency resources over a fixed amount of time, the higher the spectral efficiency will be.
The spectral efficiency measure is without doubt a very important performance measure. However, the measure is mainly significant in case of fully loaded wireless communications systems. In other words, if the system is always fully loaded, i.e. if there is at least as much traffic to serve as the radio resources may support, then a higher spectral efficiency will lead to a better utilization of the resources as more UEs may be served. However, wireless communications systems are seldom fully or even highly loaded. Measurements from networks in operation show that only a fraction of the frequency resources are utilized most of the time and that all traffic may be served using just a portion of the available spectrum, with the exception for traffic in high density areas at peak hours. Most of the time UEs will be scheduled in a part of the frequency bandwidth only, whereas other parts of the frequency bandwidth will be free from transmissions, as illustrated in
These two limitations have consequences both on the performance in the cell itself, i.e. on the intra-cell performance, as well as on the inter-cell performance, i.e. how a cell affects its neighbors.
b illustrates required SINR as a function of frequency resource blocks for a scheduling interval corresponding to the resource allocation illustrated in
With conventional LA, an MCS of highest order, also referred to as the most efficient MCS, is chosen for a certain transmit/receive power, a desired Transport Block Size (TBS) and the resulting SINR based on the prevailing channel quality. However, the highest order of MCS typically means assigning the transport block to the smallest possible amount of resource blocks, which requires a high SINR. With a high SINR requirement, more power needs to be transmitted/received in order to reach a satisfactory performance for a given channel quality. A higher SINR requirement may thus be translated into a higher transmit power, and consequently into a higher interference to other cells.
In addition to a potentially higher interference, transmissions on only parts of the available resource blocks cause large fluctuations in the interference. These fluctuations would significantly affect a performance of decoders and many other functions such as LA and scheduling, since the performance is dependent on a reliable prediction of the interference.
An object is therefore to address some of the problems and disadvantages outlined above, and to allow a reduction of the SINR requirement for a transmission in a scheduling interval.
In accordance with an embodiment, a method in a radio node of a wireless communication system, of reducing a signal to noise and interference ratio requirement for a transmission in a scheduling interval is provided. The transmission is being performed in a cell served by the radio node. The method comprises estimating a frequency resource utilization in the scheduling interval, and comparing the estimated frequency resource utilization with a first threshold. When the estimated frequency resource utilization is equal to or below the first threshold, the method further comprises increasing the frequency resource utilization for the transmission, and adjusting a link adaptation for the transmission based on the increased frequency resource utilization.
In accordance with another embodiment, a radio node configured to be used in a wireless communication system, and to reduce a signal to noise and interference ratio requirement for a transmission in a scheduling interval is provided, where the transmission is being performed in a cell served by the radio node. The radio node comprises an estimating circuit configured to estimate a frequency resource utilization in the scheduling interval, and a comparator configured to compare the estimated frequency resource utilization with a first threshold. It further comprises a frequency resource allocation circuit configured to increase the frequency resource utilization for the transmission, when the estimated frequency resource utilization is equal to or below the first threshold, and to adjust a link adaptation for the transmission based on the increased frequency resource utilization.
An advantage of particular embodiments is to allow for improved LA and decoding performance due to a smoother and more predictable interference profile. A smoother interference profile is especially important for cell-edge UEs which are more affected by inter-cell interference.
a illustrates the frequency resource utilization for three UEs in a scheduling interval.
b illustrates required SINR as a function of frequency resource blocks for a scheduling interval, and a mean value of SINR.
a illustrates the frequency resource utilization for a scheduling interval according to an embodiment.
b is a state diagram illustrating the possibilities of decreasing a SINR requirement.
a-b are flowcharts of the method in the radio node according to embodiments.
a-5b and 6 illustrate schematically a radio node according to embodiments.
a-d are comparisons between prior art and embodiments in terms of required and mean SINR.
In the following, different aspects will be described in more detail with references to certain embodiments and to accompanying drawings. For purposes of explanation and not limitation, specific details are set forth, such as particular scenarios and techniques, in order to provide a thorough understanding of the different embodiments. However, it will be apparent to one skilled in the art that other embodiments that depart from these specific details exist.
Moreover, those skilled in the art will appreciate that while the embodiments are primarily described in form of a method and a node, they may also be embodied in a computer program product as well as in a system comprising a computer processor and a memory coupled to the processor, wherein the memory is encoded with one or more programs that may perform the method steps disclosed herein.
Embodiments are described herein by way of reference to particular example scenarios. Particular aspects are described in a non-limiting general context in relation to an LTE system. It should though be noted that the embodiments may also be applied to other types of radio access networks using scheduling and LA.
A problem of high SINR requirements and high interference variances due to a traditional resource allocation prioritizing a high spectral efficiency is addressed in embodiments of the invention. A frequency resource utilization in a scheduling interval is estimated and compared with a threshold. When the estimated utilization is below or equal to the threshold, the frequency resource utilization of each transmission in the scheduling interval may be increased. The increased frequency resource utilization will allow for an adjustment of the LA for the transmissions, as a higher frequency resource utilization allow for a reduced SINR requirement and thus a reduced modulation order and/or coding rate.
This disclosure thus relates to a utilization of empty portions of the available frequency resources in order to provide a lower spectral efficiency. A lower spectral efficiency will result in a higher frequency resource utilization at all times, which may e.g. allow for a decreased total transmit power compared to conventional schemes. In other words, it is proposed to decrease the spectral efficiency while ensuring that this reduced spectral efficiency does not become a bottleneck for the network performance. Based on the desired reduced spectral efficiency, a resource allocation is performed that results in a lower SINR requirement. This may in turn allow for a reduced transmit power. The purpose is to exploit a typically low or medium transmission load in cells of a wireless communications system and to relax a requirement of packing transmission data in as few frequency resources as possible.
a illustrates a scheduling interval with around 50% of the frequency resources utilized by the UEs. In a traditional scheduling, this will be the situation when the traffic load is moderate, as the aim for a high spectral efficiency leads to unutilized resources if there is not enough traffic to serve. According to embodiments, an alternative is to perform a less efficient resource allocation resulting in a higher frequency resource utilization. Thereby a lower order MCS may be chosen. All available frequency resources may be utilized to serve the same moderate traffic load that is traditionally served with only 50% of the frequency resources. Every UE may e.g. occupy twice as much frequency resources. Alternatively more than twice the frequency resources may be allocated to one UE, and less than twice for another. Any other distribution of the resources between the UEs may be performed as an optimization of the resource allocation. A resource allocation utilizing all available resources is illustrated in
As already mentioned, the MCS of highest order that satisfies a certain BLER target is chosen based on a resource allocation and a desired number of bits to transmit. In LTE the resource allocation is given as a number of Physical Resource Blocks (PRB), and the desired number of bits to transmit is given as a TBS. The chosen MCS requires a SINR level, which for a given channel and for a given UE could be translated into a required transmit power. The higher the required SINR is, the higher the required transmit power. However, if more frequency resources may be allocated to transmit the same TBS, i.e. more PRB are allocated for the same TBS which is possible when the load is low or moderate and there is no need to optimize the spectral efficiency, the LA may be adjusted to use a lower order MCS. This will in turn lead to a reduced SINR requirement, and for a specific UE and specific channel conditions, a lower transmit power per PRB is needed.
An adjustment of the LA may be performed by decreasing a code rate while still utilizing a same modulation order. However, other alternatives of adjusting the LA are also possible as may be seen in the state diagram in
In order for an eNB to know when the frequency resource utilization may be increased for transmissions in a scheduling interval, the frequency resource utilization is estimated for the scheduling interval based on conventional LA and scheduling. One way to estimate the frequency resource utilization is to use a look up table mapping a given SINR and TBS to a certain frequency resource utilization. The estimated frequency resource utilization may be compared with a first threshold, in order to decide whether an increase of the frequency resource utilization is desired or not. This first threshold may typically be pre-defined, and indicates a limit for the frequency resource utilization. When the estimated frequency resource utilization is equal to or below the threshold, the frequency resource utilization for one or more transmissions in the scheduling interval may be increased and the LA may be adjusted. The estimated frequency resource utilization for all transmissions in the scheduling interval may be used as a basis for how much more frequency resources that may be allocated compared to a conventional scheduling.
In case the estimated frequency resource utilization is above the first threshold, an increased frequency resource utilization may still be possible as long as an average cell load over time is low or moderate. It is only for the case of a continuously high load in the cell that resources must be used in the most efficient way, i.e. that the spectral efficiency must be maximized. For intermittently or occasionally high loads, i.e. bursty traffic under short or non-continuous periods of time, a spreading of the bursts in time and frequency may allow for a homogeneous resource utilization, which in turn leads to a smooth inter-cell interference profile. Therefore, when the estimated frequency resource utilization is above the first threshold indicating a high frequency resource usage for the scheduling interval, it may also be checked if the average transmission load in the cell is below a second threshold, which would mean that the present high frequency resource utilization is an exception seen over time. In order to make it possible to increase the frequency resource utilization in the scheduling interval, delay tolerant bits in the transmissions may be excluded from the transmissions in a current scheduling interval, and may be delayed to a transmission in a subsequent scheduling interval. This will allow for an increase of the frequency resource utilization in the transmissions, depending on how many delay tolerant bits that have been excluded, which will in turn allow for an adjustment of the LA.
The average cell load may be available in the radio node or may be retrieved from the network. The average cell load may e.g. be calculated as an average over different averaging periods, such as the last 60 seconds or the last few seconds. Depending on the averaging period, different results may be expected. If the averaging period is short and a measurement of the cell load is initiated at the start of a burst, the average cell load may be overestimated. Therefore, it may be better to utilize a dynamic second threshold, rather than a pre-defined one. The updates of the dynamic second threshold may be based on e.g. an amount of incoming traffic since delaying of bits was initiated. An average burst period may be computed, and when bits have been delayed and the incoming traffic burst seems to be larger than the average burst length, it may be advantageous to decrease the second threshold so that the delaying of delay tolerant bits is stopped. Otherwise there may be a risk to create a too large backlog of data for transmission.
In alternative embodiments, the increase of the frequency resource utilization may be applied to certain identified UEs. In one embodiment, downgraded UEs or UEs with low-tier subscriptions may be addressed. When a certain UE has surpassed its traffic quota, or if a UE has a low-tier subscription i.e. a limit on connection speed, the UE is e.g. not allowed to have download and upload rates higher than a predetermined value. The conventional way of solving such a situation is by limiting the number of resource blocks allocated to the UE while still performing LA and scheduling in a way that maximizes the spectral efficiency, even if the UE is the only UE transmitting in a given cell. By instead limiting the download/upload rate although increasing the frequency resource usage according to embodiments, a throughput limitation would be achieved while simultaneously creating a smooth interference to other cells.
In an alternative embodiment, UEs with limited battery life may be addressed. Regardless of the amount of traffic to be transmitted, the frequency resource utilization for UEs with limited battery life may be increased allowing them to transmit their data on more resources than an optimal MCS selection would allow. The UE may then need less power to transmit its data as the LA may be adjusted which allows for a lower SINR or a lower transmit power.
In one embodiment, the increase of the frequency resource utilization for one or more transmissions of a scheduling interval, and the adjustment of the LA in the scheduling interval, is followed by a decrease of the transmit power for the scheduling interval. How much the transmit power may be decreased is dependent on how the LA is adjusted. One embodiment relates to an eNB of an LTE system. For the uplink, two alternatives to control the transmit power from the UE are possible:
For the downlink, a UE-specific RRC configuration may be used to signal the power allocation of the eNodeB.
a is a flowchart of the method in a radio node of a wireless communication system, of reducing a signal to noise and interference ratio requirement for one or more transmissions in a scheduling interval, according to embodiments of the invention. The radio node is in one embodiment an eNB in an LTE system. The transmissions are performed in a cell served by the radio node. The method comprises:
When the estimated frequency resource utilization is equal to or below the first threshold, which is the case when the load in the cell is low or medium high, the method further comprises:
b is a flowchart of the method that is performed after the comparison in 420, when the estimated frequency resource utilization is found to be above the first threshold, according to embodiments of the invention. This flowchart covers the method performed in case of bursty traffic, with high load in the cell during a short period of time. When an average transmission load in the cell is below a second threshold, the method further comprises the following:
When the average transmission load in the cell is equal to or above the second threshold, the inventive method will not be used (illustrated by the STOP sign), and the scheduling may thus be performed in a conventional way with a high spectral efficiency. This is the case when the load is high during a longer period, which will make it difficult to transmit delay tolerant bits in subsequent scheduling intervals, as the frequency resource utilization is above the first threshold in many subsequent scheduling intervals.
The second threshold may be dynamically updated. This may be done e.g. based on a comparison of the burst length with an average burst length. If the burst is longer than an 20 average burst, the second threshold may be decreased in order to control the exclusion of delay tolerant bits.
The radio node is schematically illustrated in
In
The circuits and units described above with reference to
Hence in the embodiments described, the code means in the computer program 656 of the radio node 500 comprises an estimating module 656a for, a comparator module 656b for, a frequency resource allocation module 656c for, an excluding module 656d for, and a power control module 656e for. The code means may thus be implemented as computer program code structured in computer program modules. The modules 656a-e essentially perform the steps of the flow in
Although the code means in the embodiment disclosed above in conjunction with
The examples A-D hereinafter described with reference to an LTE system and to
Example A described with reference to
The conclusion for example A is thus that a SINR of 0 dB per PRB is needed when using 20 PRB, whereas a SINR of 3.5 dB per PRB is needed when using 10 PRB. When using 10 PRB only, 10 PRB are left unused and in some sense wasted in case of a low or medium load, and the mean required SINR per PRB will in this case be higher than the mean SINR level of 0 dB valid when allocating 20 PRB instead. Several advantages may thus be observed when alocating 20 PRB instead of 10 PRB to the UE:
Example B described with reference to
Example C described with reference to
Example D described with reference to
The above mentioned and described embodiments are only given as examples and should not be limiting. Other solutions, uses, objectives, and functions within the scope of the accompanying claims should be apparent for the person skilled in the art.
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/SE10/51229 | 11/10/2010 | WO | 00 | 5/9/2013 |
